Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
INTRODUCTION TO THE LAW OF PRESIDENTIAL ELECTIONS: ISSUES IN THE WAKE OF FLORIDA 2000 REUBIN O’D. ASKEW* In November and December 2000, the country focused its whitehot gaze at our state. National media, political pundits, and lawyers poured into Tallahassee as one of American history’s momentous events unfolded. Although the stakes were high and the debate was often heated, Florida’s three governmental branches helped lead the nation to a peaceful resolution of the closest presidential contest in its history. Besides the 2000 election’s obvious political importance, it raised issues of legal significance in the selection of the President. On March 23, 2001, some of the country’s most noted scholars and legal practitioners gathered at the Florida State University College of Law to address many of these legal and constitutional issues, including the roles of state and local election practices in presidential selection, the Fourteenth Amendment, the role of state constitutions, the Twelfth Amendment, and the Electoral College selection process. The articles in this issue memorialize the presentations made that day.1 I attended the symposium and was proud to partake in the day’s debates, which were sometimes serious, sometimes lively, but always respectful. I am pleased to see that some of what was debated among the symposium’s participants, and between the participants and the crowd, has been incorporated into the twenty-one articles printed within. The information presented in this issue helps us understand what happened here in Florida after the 2000 election, and teaches us how to apply the lessons we learned to future elections throughout the country. * Governor, State of Florida, 1971-79. 1. Editor’s note: a webcast of the symposium can be viewed at http://www.law.fsu.edu/ events/symposia/election_law/webcast.php. Professor Richard D. Friedman, Judge Richard A. Posner, and Hayward H. Smith did not attend the symposium. xiii BUSH V. GORE AS AN EQUAL PROTECTION CASE RICHARD BRIFFAULT* I. INTRODUCTION ..................................................................................................... II. COUNTING AND RECOUNTING THE FLORIDA BALLOTS IN THE 2000 PRESIDENTIAL ELECTION .................................................................................... A. The Protest Phase ......................................................................................... B. The Contest Phase......................................................................................... III. BUSH V. GORE: THE OPINIONS ............................................................................ A. The Per Curiam Opinion.............................................................................. B. Justice Souter ............................................................................................... C. Justice Breyer................................................................................................ D. Justices Stevens and Ginsburg .................................................................... IV. EQUAL PROTECTION AND THE FLORIDA SUPREME COURT’S MANUAL RECOUNT ORDER .................................................................................................................. A. The Equal Protection Clause and the Vote .................................................. B. Equal Protection and the Administration of Elections: Setting the Standard for Federal Judicial Intervention................................................ C. The Uncertain Constitutional Status of Undervote Ballots........................ D. Geographic Discrimination and the Undervote .......................................... E. Fundamental Fairness and the Recount Order .......................................... V. CONCLUSION: FEDERALISM AND EQUAL PROTECTION IN BUSH V. GORE ........... 325 330 331 337 341 342 343 344 344 345 345 349 356 362 368 372 I. INTRODUCTION In Bush v. Gore,1 the United States Supreme Court applied the Equal Protection Clause to the mechanics of state election administration. The Court invalidated the manual recount of the so-called undervote—that is, ballots that vote-counting machinery had found contained no indication of a vote for President—which the Florida Supreme Court had ordered to determine the winner of Florida’s vote for presidential electors in the 2000 presidential election.2 The United States Supreme Court reasoned that the principles it had previously articulated in applying the Equal Protection Clause to the vote were violated by the Florida court’s failure to assure consistency between and within Florida’s counties in the determination of whether particular undervote ballots constitute legally valid votes.3 The Court correctly determined that the Equal Protection Clause applies to the state and local procedures affecting the casting and * Vice Dean & Joseph P. Chamberlain Professor of Litigation, Columbia University School of Law. This paper benefited from the comments of the official commentators— Steve Bickerstaff, Heather Gerken, and Spencer Overton—of an earlier draft presented at the Florida State University College of Law’s symposium on the Law of Presidential Elections: Issues in the Wake of Florida 2000, as well as from the thoughtful criticisms of Mike Dorf, Sam Issacharoff, and Rick Pildes. 1. 531 U.S. 98 (2000). 2. See Gore v. Harris, 772 So. 2d 1243 (Fla. 2000), rev’d sub nom. Bush v. Gore, 531 U.S. 98 (2000). 3. Bush v. Gore, 531 U.S. at 104-07. 325 326 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 counting of ballots but was, for the most part, wrong to find that the Florida Supreme Court’s order denied Florida voters equal protection. In particular, the inconsistencies in counting undervotes, which the Florida court’s order appeared to tolerate and which so disturbed the United States Supreme Court, did not constitute an equal protection violation. Equal protection ought to apply to the nitty-gritty of local election practices because those practices can have the effect of disenfranchising voters and discriminating among identifiable groups of voters. Such practices can negate the right to vote and the right to an equally weighted vote—rights long protected by the Equal Protection Clause. However, with virtually every local administrative decision having the potential to burden some voters relative to others, the application of the Equal Protection Clause to election rules and procedures could effectively federalize an area which has long been the domain of state and local government. Decentralization of election administration reflects important political values, including the opportunities for local participation and decisionmaking concerning contestable political issues, as well as protection from centralized political manipulation and abuse. Decentralization necessarily entails variation in election practices across the different local units charged with administering the procedures for casting and counting ballots. Subjecting all interlocal differences in election rules and procedures to close constitutional scrutiny could eliminate meaningful decentralization of election administration. This is not to say that decentralizing election administration to the local level is an inherently wise policy. State legislative or administrative measures addressed to the selection of voting machinery, ballot design, the process of obtaining absentee ballots, or the standards for conducting manual recounts could certainly improve our system of casting and counting votes. However, given the political values that support decentralization, I would suggest that the mix of state and local decisionmaking in election administration is primarily a matter for political, not judicial determination. To be sure, judicially imposed centralization would be appropriate when certain practices are necessarily required or precluded by constitutional principles. The presumption of universal adult citizen suffrage and the one person, one vote rule for weighting ballots are constitutional principles that ended alternative state or local rules concerning the availability of the franchise and the apportionment of legislative representation. Comparable constitutional principles might prohibit certain state or local election administrative practices that consistently burden the vote or discriminate among voters. But not all questions concerning election administration can be resolved by reference to 2001] EQUAL PROTECTION 327 constitutional principles, and not all state or local rules that affect the casting and counting of ballots violate constitutional norms. The political tradition of decentralized election administration and the values that support it, combined with the absence of constitutional rules for answering many questions of election procedure, suggest the need for an equal protection standard that both protects fundamental voting rights and respects local variations in rules and procedures. The Supreme Court apparently agrees. Even as it applied equal protection to the details of election administration, Bush v. Gore underscored the need to constrain equal protection review when it “limited” its “consideration . . . to the present circumstances, for the problem of equal protection in election processes generally presents many complexities.”4 This Article will examine the equal protection issues presented in Bush v. Gore. Part II will review the political and legal struggle over counting and recounting the Florida presidential vote. Part III will summarize the equal protection analysis of the recount issues undertaken by the Justices in Bush v. Gore. Part IV will then examine the Supreme Court’s treatment of the equal protection issues created by the Florida Supreme Court’s order in the context of a more general effort to determine an appropriate role for federal court equal protection review of state election procedures. Drawing on a series of lower federal court cases decided prior to Bush v. Gore that dealt with constitutional challenges to local election practices, I will suggest that federal constitutional intervention in state election administration should be limited to cases of “patent and fundamental unfairness”5 in which the state or local practice undermines the integrity of the election itself. “[O]rdinary dispute[s] over the counting and marking of ballots,”6 even those involving administrative errors that result in distinctions among voters, should not be treated as raising equal protection issues justifying federal court action. Applying that standard, the Florida Supreme Court’s manual recount order did not violate the Equal Protection Clause because it would not have caused fundamental unfairness in the Florida election. The Florida court’s manual recount order would not have led to the exclusion of any voters;7 it did not unconstitutionally 4. Id. at 109. 5. Griffin v. Burns, 570 F.2d 1065, 1077 (1st Cir. 1978) (arguing that in cases of “patent and fundamental unfairness” due process may be violated). 6. Id.; see also Duncan v. Poythress, 657 F.2d 691, 704 (5th Cir. Unit B 1981) (stating that no constitutional question is presented by “garden variety challenges to the manner in which ballots are counted”). 7. See Bennett v. Yoshina, 140 F.3d 1218, 1227 (9th Cir. 1998) (holding that election procedure will likely be held unconstitutional on substantive due process grounds only if “significant disenfranchisement” results from a change in election procedure). 328 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 favor any group of voters over any other group of voters;8 and it did not unsettle any of the expectations, strategies, or voting plans of any of the participants in Florida’s electoral process.9 This is not to say that the Florida Supreme Court’s order was flawless. The Florida court violated equal protection principles by ordering the tabulation of recount results from some counties that may have included votes obtained from the manual recount of overvote ballots—that is, ballots that the vote-counting machinery rejected because they contained two or more presidential votes per voter— without providing for a manual recount of overvote ballots elsewhere in the state. Although the Florida court could have constitutionally limited the recount to the undervote, it was a mistake to mingle votes retrieved from overvote ballots in some counties with a recount limited to undervote ballots in the other counties. However, the Florida court’s failure to provide specific guidelines for the determination of what constituted a valid ballot—Bush v. Gore’s principal concern—was not unconstitutional. The lack of such guidance could have led to uncertainties in assessing ballots and might have resulted in variations among counties and canvassing teams in the standards for counting ballots. But variations in the definition of a valid undervote ballot in a manual recount would not have posed a threat to fundamental fairness. The manual recount would not have led to the exclusion of any ballots that were constitutionally required to be counted. Nor would it have led to a departure from any generally accepted standard for determining which undervote ballots are valid votes. Bush v. Gore itself highlighted the minimal constitutional protection accorded to undervote ballots when the Supreme Court effectively excluded all votes that might have been gleaned from a statewide inspection of the undervote ballots from Florida’s final tally. Presumably, the failure to count undervote ballots was not a constitutional violation. Moreover, the intercounty or intracounty variations in standards for determining whether an undervote ballot contained a legally valid vote were not ordered by state law and would not have reflected a state-level decision to prefer certain parts of the state or voters who 8. Cf. Welch v. McKenzie, 765 F.2d 1311 (5th Cir. 1985) (rejecting challenge to election based on claimed irregularities, errors, and fraud in the distribution of absentee ballots given district court’s finding that, although numerous violations of state election laws had occurred, there was no evidence of racially discriminatory intent); Bell v. Southwell, 376 F.2d 659 (5th Cir. 1967) (invalidating an election due to racially discriminatory practices in administration). 9. Cf. Partido Nuevo Progresista v. Barreto Perez, 639 F.2d 825, 828 (1st Cir. 1980) (upholding the counting of ballots containing marks outside the spaces and squares designated by law for marking preferences: “no party or person is likely to have acted to their detriment by relying upon the invalidity of ballots with marks outside the ballots’ drawn rectangles”). 2001] EQUAL PROTECTION 329 backed particular candidates or voters associated with particular parties over others. Thus, any variations in counting standards would not have amounted to an unconstitutional discrimination among Florida voters. To be sure, differences in standards for assessing undervote ballots may be troubling. But the Florida Supreme Court’s apparent willingness to tolerate variations may have been required by the court’s need to abide by the special legal imperatives for resolving disputes concerning presidential elections articulated by the United States Supreme Court in Bush v. Palm Beach County Canvassing Board.10 In order to honor the state legislature’s constitutional prerogative of writing the rules for the selection of presidential electors and the legislature’s presumptive interest in benefitting from the federal “safe harbor” law providing congressional deference to state resolutions of disputes concerning electors, the Florida court may have been unable to spell out more precise standards than those found in existing statutes and case law. Rather than create a problem of fundamental unfairness, the Florida court’s order would have increased the fairness of the Florida vote. Unlike the county-level manual recounts conducted prior to the certification of the Florida results, the court-ordered statewide recount would not have been biased in favor of a particular candidate. Moreover, the manual recount would have provided a partial remedy for the intercounty disparities in the percentage of votes that resulted in undervote ballots—disparities closely associated with the intercounty differences in the quality of Florida’s voting machinery. In short, instead of limiting voting rights and discriminating among voters, the Florida Supreme Court’s order promoted voting rights and the equal treatment of voters. Indeed, one striking consequence of the United States Supreme Court’s decision is that, unlike any other case in which the Equal Protection Clause was used to vindicate the right to vote, Bush v. Gore produced a smaller electorate marked by greater intercounty discrepancies than would have been the case had the Court stayed its hand. In Part V, I will conclude by touching on the relationship between the Court’s equal protection analysis and its commitment to federalism. A central premise of our federal system is that many important questions are left to smaller units rather than bigger ones, even though—indeed, perhaps, because—that will create a multiplicity of different approaches. That is the philosophy of federalism which has been so central to the jurisprudence of the Justices who composed the Bush v. Gore majority and embraced the application of the Equal Protection Clause to the manual recount order. Bush v. Gore’s con10. 531 U.S. 70 (2000). 330 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 cern about interlocal variations in election administration suggests a surprising discomfort about the very values of local decisionmaking and interlocal variation which are at the heart of federalism itself. II. COUNTING AND RECOUNTING THE FLORIDA BALLOTS IN THE 2000 PRESIDENTIAL ELECTION On November 7, 2000, the people of the United States went to the polls to vote for the next President and Vice President, or rather, as they were reminded over the next few days, to choose the electors who would vote for President and Vice President of the United States. On November 8, 2000, they awoke to find that although Vice President Al Gore enjoyed a narrow lead in the popular vote—his lead ultimately grew to 540,000 votes or about one-half of one percent of the total vote cast—there was no electoral vote winner. With 270 electoral votes necessary to win, Gore had carried states and the District of Columbia casting a total of 267 electoral votes.11 His Republican opponent, Texas Governor George W. Bush, had carried states that would cast 246 electoral votes for him. Still in the balance, and with 25 electoral votes, the key to the election, was Florida. On the morning after the election, Bush led in Florida by 1,784 votes. Florida law required the ballot counting machines to count the ballots again if the winner’s margin over the next-best candidate totaled less than one-half of one percent of the vote.12 Bush’s margin over Gore was about three-hundredths of one percent of the vote. The machine recount, which was completed by Friday, November 10, reduced Bush’s lead to a mere 327 votes,13 although his lead was likely to grow once the overseas absentee ballots, which historically had favored Republicans, were included. At no point after the machine recount did Bush’s margin over Gore ever exceed one thousand out of the nearly six million votes cast. With the candidates so close, the legal issues over the next five weeks were dominated by Gore’s efforts to obtain a recount.14 For11. Ultimately, Vice President Gore received just 266 electoral votes when one District of Columbia elector who was pledged to Gore spoiled her ballot. 12. FLA. STAT. § 102.141(4) (2000), amended by 2001 Fla. Laws ch. 40, § 41, at 148-49. 13. THE NEW YORK TIMES, 36 DAYS: THE COMPLETE CHRONICLE OF THE 2000 PRESIDENTIAL ELECTION CRISIS 29 (2001) [hereinafter 36 DAYS]. 14. Recount-related issues were not the only legal questions growing out of the Florida presidential vote that state and federal courts addressed in November and December 2000. Voters in Palm Beach County unsuccessfully challenged that county’s unusual and apparently confusing “butterfly ballot,” which, they contended, caused many Gore voters to mistakenly vote for Pat Buchanan. Fladell v. Palm Beach County Canvassing Bd., 772 So. 2d 1240 (Fla. 2000). The actions of election officials in Seminole and Martin Counties, which enabled Republican party workers to add voter identification numbers to requests for absentee ballots, led to challenges to the legality of absentee ballots in those counties. Jacobs v. Seminole County Canvassing Bd., 773 So. 2d 519 (Fla. 2000); Taylor v. Martin County Canvassing Bd., 773 So. 2d 517 (Fla. 2000). Absentee ballot issues may ultimately 2001] EQUAL PROTECTION 331 mally, this challenge consisted first of a protest phase—that is, challenges to county-level election results—prior to the formal certification of the statewide results; and, then, a contest phase, or a challenge to the certified statewide result. The recount struggle can also be analyzed in terms of the different legal issues that dominated its different stages. Initially, these concerned primarily questions of timing, authority, and discretion. Could Florida’s Secretary of State waive the statutory deadline for the submission of county-level election results and include late-filed results from county canvassing boards that had undertaken manual recounts? Was she required to do so? Did the statutory authorization to undertake a manual recount on evidence of an “error in the vote tabulation” apply to instances where the vote-counting machinery had worked as designed but had failed to count imperfectly marked ballots? Did the Florida courts have the authority to require the Secretary of State to accept late-filed returns? In early December, the legal issues began to shift from the powers and duties of boards and courts to the equal treatment of voters in different counties, the standards for determining whether an imperfectly marked ballot is a legal vote, and the interplay of these two questions. These were the issues that either shaped or came directly before the Supreme Court in Bush v. Gore. However, the earlier issues played an important role in the development of the recount struggle and also helped frame the equal protection questions that took center stage in Bush v. Gore. A. The Protest Phase Like most states, Florida uses a highly decentralized procedure for conducting elections, counting votes, and challenging vote counts.15 Elections are conducted by county supervisors of elections, and the votes are counted by county canvassing boards composed of each county’s supervisor of elections, a county court judge, and the chair of the board of county commissioners.16 The county canvassing board certifies the results and, for elections involving state or federal offices, transmits them to the state. The state Elections Canvassing Commission, composed of the Governor, the Secretary of State, and the Director of the Division of Elections, certifies the returns and dehave been crucial in resolving the presidential election. David Barstow & Don Van Natta, Jr., How Bush Took Florida: Mining the Overseas Absentee Vote, N.Y. TIMES, July 15, 2001, at A1. 15. Florida’s election laws are codified at FLA. STAT. chs. 101-02 (2000). After the 2000 presidential election, the Florida Legislature made significant amendments to the election code. See 2001 Fla. Laws ch. 40. This Article considers the election statutes as they existed before the 2001 amendments. 16. FLA. STAT. §§ 102.131(1), 102.141(2) (2000) (amended 2001). 332 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 clares the winner based on the county results.17 This process is required by law to be concluded within seven days after the election, or, in 2000, by 5 p.m. on Tuesday, November 14—with the exception of overseas absentee ballots which, as a result of a consent decree, may be received until ten days after the election.18 Florida law also provides that protests of election returns may be submitted to the county canvassing boards.19 Any candidate or voter can protest the returns of an election as erroneous,20 any candidate or political party whose candidates’ names appeared on the ballot may ask a county board for a manual recount, and the county board is authorized to undertake a manual recount.21 If a manual check of at least three precincts involving one percent of the total votes cast in the county “indicates an error in the vote tabulation which could affect the outcome of the election,” the county canvassing board is authorized, but not required, to “manually recount all ballots.”22 To do so, the county canvassing board appoints counting teams composed of voters who are members of at least two political parties, who then inspect the ballots by hand.23 “If a counting team is unable to determine a voter’s intent in casting a ballot, the ballot shall be presented to the county canvassing board for it to determine the voter’s intent.”24 Between November 9 and November 11, the Democratic Party filed protests and requested manual recounts in Broward, MiamiDade, Palm Beach, and Volusia Counties. Broward, Miami-Dade, and Palm Beach are the three most populous counties in the state. Gore carried Broward and Palm Beach by substantial margins.25 Gore led by much smaller margins in Miami-Dade, the state’s most populous county, and in Volusia, a much smaller county. In Broward, Miami-Dade, and Palm Beach Counties, the machine counts found a significant number of presidential ballots that contained no presidential preference—nearly 30,000 such ballots in Miami-Dade and Broward each, and another 15,000 in Palm Beach.26 In 17. Id. § 102.111(1). 18. See id. 19. Id. § 102.166. 20. Id. § 102.166(1). 21. Id. § 102.166(4)(a), (c). The request for a manual recount must be filed prior to the time the county canvassing board certifies the result for the office in question, or within seventy-two hours of election day, whichever is later. Id. § 102.166(4)(b). 22. Id. § 102.166(5). 23. Id. § 102.166(7)(a). 24. Id. § 102.166(7)(b). 25. 36 DAYS, supra note 13, at 285. Gore received his highest county percentage in Broward, and his third highest percentage of the county vote in Palm Beach. Gore’s number two county, Gadsden, cast well under 20,000 votes, compared with the nearly one million votes cast in Broward and Palm Beach together. 26. Id. 2001] EQUAL PROTECTION 333 terminology that became famous as the recount struggle continued, these no-preference ballots consisted of undervotes—ballots which, according to the vote-counting machinery, contained no vote for any presidential candidate—and overvotes—ballots in which the voter voted for more than one candidate and thus invalidated the ballot. The undervote ballots became the principal focus of the political and legal battle over the recount. Although some undervote ballots may have reflected the decision of voters to skip the presidential election and focus on other races, Gore’s forces alleged that in many cases voters had attempted to cast a presidential vote, but due to problems with the voting machinery, their preferences had failed to register.27 Indeed, the percentage of a county’s ballots containing undervotes was associated with the type of voting machinery the county used, suggesting that problems with the voting machinery were at least as important a factor as voter preferences in explaining why some ballots that bore voters’ markings had not been read by the machines as containing votes.28 Broward, Miami-Dade and Palm Beach Counties were among the two dozen Florida counties that used Votomatic punch card ballots.29 In those counties, a voter casts a ballot by placing a punch card into a holder and then uses a stylus to make a hole in the ballot card corresponding to the voter’s preference. A machine “reads” the light coming through the hole in the ballot card and records it as a vote. A ballot can be counted by the machine only if the hole is punched through cleanly, and the “chad”—or the material occupying the space to be punched out to make the hole—is cleanly detached. Counties using Votomatic punch card ballots had five times the undervote rate as counties using optical scan ballots, in which voters mark their choice with a pencil next to the name of their preferred candidate.30 The high undervote rate was apparently attributable to problems characteristic of the punch card mechanism. If the punch card and the cardholder are improperly aligned, the punch tool may fail to punch out the hole fully. The rubber or plastic strips that help hold the card in place may age and become too stiff to allow the paper to be punched out of the hole, that is, the strips prevent the chad from passing through, creating a dented or “dimpled” chad but not a detached one. When these problems occur, the voter’s attempt to vote may leave a mark on the ballot which is not read by the votecounting machinery but is detectable as a vote to a human votecounter. 27. Compl. to Contest Election at 7-8, Gore v. Harris, No. CIV-00-2808 (Fla. 2d Cir. Ct. Nov. 27, 2000), available at http://election2000.stanford.edu/CV-00-2808a.pdf. 28. 36 DAYS, supra note 13, at 189-91. 29. Id. 30. Id. 334 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 The combination of large voting populations, large margins for Gore, and high undervote rates due to the use of Votomatic punch card ballots made Broward and Palm Beach Counties particularly attractive targets for Democratic efforts to obtain a manual recount. Although Gore’s margin in Miami-Dade was smaller, the county’s large population and high undervote rate indicated that it, too, might provide Gore with an appreciable net gain relative to Bush. The fourth county in which a manual recount was sought, Volusia, differed from the others. Volusia used optical scan equipment and, thus, had only a small undervote. But the Volusia count had been marked by a malfunction of the electronic ballot tabulating machine in one precinct, making a manual recount appropriate.31 Initially, only the Palm Beach and Volusia county canvassing boards voted to undertake manual recounts. Palm Beach County undertook the one percent sample recount and found sufficient new votes to constitute “an error in the vote tabulation which could affect the outcome of the election.”32 Broward and Miami-Dade undertook the sample recounts but decided that countywide manual recounts were not warranted.33 Nor was it clear whether any of the manual recount findings would be included in the certified results. On November 13, Florida Secretary of State Katherine Harris announced that she would adhere to the November 14 statutory deadline for certifying the election results—excluding the overseas absentee ballots— even if manual recounts were still pending. She contended that in the absence of an Act of God, such as a hurricane, she had no authority to count any returns received after the November 14 deadline.34 On November 14, Judge Terry Lewis of Florida’s Second Circuit Court in Tallahassee found that Secretary Harris had the discretion to include late-filed manual recount returns in the statewide results, notwithstanding the statutory deadline to certify the election.35 Indeed, Judge Lewis suggested that such authority to waive the deadline was actually necessary to prevent discrimination against the most populous counties.36 Secretary Harris’ refusal to accept manually recounted returns submitted in good faith after the seven-day deadline would mean . . . that only in sparsely populated counties could a Canvassing Board safely exercise what the Legislature has clearly 31. See Siegel v. LePore, 120 F. Supp. 2d 1041, 1046 (S.D. Fla. 2000), aff’d, 234 F.3d 1163 (11th Cir. 2000). 32. FLA. STAT. § 102.166(5) (2000) (amended 2001). 33. 36 DAYS, supra note 13, at 65 (Broward), 72 (Miami-Dade). 34. McDermott v. Harris, 2000 WL 1693713, at *1 (Fla. 2d Cir. Ct. Nov. 14, 2000). 35. Id. 36. Id. at *2. 2001] EQUAL PROTECTION 335 intended to be an option where the Board has a real question as to the accuracy of a vote. . . . It is unlikely that the Legislature would give the right to protest returns, but make it meaningless because it could not be acted upon in time.37 In response to Judge Lewis’ ruling, Secretary Harris invited the Broward, Miami-Dade, and Palm Beach canvassing boards— Volusia’s had completed its recount by the statutory deadline—to submit statements of “facts and circumstances” that would justify her acceptance of late-filed amended returns. After the boards filed their statements, she rejected their reasons, concluding that only proof of voter fraud, substantial noncompliance with statutory election procedures, an Act of God, or similar “extenuating circumstances” such as “an electrical power outage, a malfunction of the transmitting equipment, or a mechanical malfunction of the voting tabulation system”—none of which had been alleged by the counties—justified waiver of the statutory deadline.38 On November 17, Judge Lewis sustained the Secretary’s action as an acceptable exercise of her discretion.39 The Florida Supreme Court, however, agreed to take the case and enjoined the Secretary and the state Elections Canvassing Commission from certifying the results of the presidential election pending the court’s decision on the merits. Following the state supreme court’s order, the Miami-Dade canvassing board voted to join Broward and Palm Beach Counties in conducting a full manual recount. As the court prepared for a full hearing, the counties canvassed the overseas absentee ballots, with final but unofficial figures boosting Bush’s lead to 930 votes. On November 21, a unanimous Florida Supreme Court reversed Judge Lewis and found that Secretary Harris was required to accept late-filed returns. In Palm Beach County Canvassing Board v. Harris, the court rejected the Secretary’s argument that the statutory standard of “error in the vote tabulation” referred only to machine failures to include machine-readable results.40 The court, thus, confirmed that the manual recount was authorized by statute when the sample recounts detected a discrepancy between the machine totals and the sample manual recount results.41 Emphasizing the fundamental importance of the right to vote under the Florida Constitution,42 the court held that the Secretary could ignore the late-filed re- 37. Id. 38. Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d 1220, 1226-27 n.5 (Fla. 2000), vacated sub nom. Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000). 39. McDermott v. Harris, 2000 WL 1714590, at *1 (Fla. 2d Cir. Ct. Nov. 17, 2000). 40. 772 So. 2d at 1229-30. 41. Id. 42. Id. at 1236-37. 336 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 sults of a county manual recount only when the results are submitted “so late that their inclusion will compromise the integrity of the electoral process” by precluding the ability of either a candidate or voter to contest the certification of the election results or preventing the State of Florida from completing its count in time to participate fully in the presidential election.43 The court required the Secretary to accept all amended county canvassing board certifications filed by 5 p.m. on Sunday, November 26.44 In the five days between the Florida Supreme Court’s order and the deadline it imposed, Broward County completed its recount. The Miami-Dade County canvassing board began a recount focused on the approximately 10,000 ballots that contained no presidential preference; then reversed itself and ordered a recount of all of the county’s nearly 700,000 ballots; then reversed itself again and, besieged by an intense and intermittently violent crowd of Republican demonstrators, voted that since it could not complete the full recount within the time allotted by the Florida Supreme Court it would not undertake a recount at all.45 The Florida Supreme Court unanimously refused a request by the Gore campaign to compel MiamiDade to resume the recount.46 Palm Beach County undertook its recount but found itself running out of time as the evening of November 26 approached. Palm Beach County requested an extension until 9 a.m. on November 27—a time that the Florida Supreme Court had indicated was also acceptable.47 Secretary Harris rejected the request. Ultimately, Palm Beach completed its recount a little after 7 p.m. on the night of November 26, but Secretary Harris refused to include the returns in the certified results.48 With only the Broward County recount results included, George W. Bush, with his lead reduced to 537 votes, was certified as the winner. Following the Florida Supreme Court’s order to the Secretary to accept late-filed county results but before the expiration of the court’s deadline for completion of the recounts, Bush sought United States 43. Id. at 1237. 44. Id. at 1240. 45. See 36 DAYS, supra note 13, at 133-35. 46. Id. at 142. 47. The court specified 5 p.m., Sunday, November 26, as the deadline for the county canvassing boards’ submissions of manual recount results “provided that the office of the Secretary of State, Division of Elections is open in order to allow receipt thereof. If the office is not open for this special purpose on Sunday, November 26, 2000, then any amended certifications shall be accepted until 9 a.m. on Monday, November 27, 2000.” Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d at 1240. Plainly, the Secretary’s acceptance of Palm Beach County’s recount results on the morning of Monday, November 27 would have been consistent with both the spirit and the letter of the Florida Supreme Court’s order. 48. 36 DAYS, supra note 13, at 164-66, 171-72. 2001] EQUAL PROTECTION 337 Supreme Court review of the Florida court’s decision. On November 24, while the recounts were underway, the Supreme Court granted certiorari with respect to two of the questions Bush raised49— whether the Florida court’s order was inconsistent (1) with Article II, Section 1, Clause 2 of the United States Constitution, which provides that presidential electors shall be appointed by each state “in such Manner as the Legislature thereof may direct;” or (2) with 3 U.S.C. § 5, which requires Congress to accept a state’s resolution of a dispute concerning the selection of presidential electors provided, inter alia, that the state’s resolution is pursuant to “laws enacted prior to” election day and is completed not later than six days before the day set for the Electoral College to vote.50 On December 4, the United States Supreme Court in Bush v. Palm Beach County Canvassing Board,51 vacated the Florida Supreme Court’s order. The United States Supreme Court expressed concern that the Florida court’s reliance on the state constitution’s protection of the right to vote in interpreting the state legislative scheme for election protests was in tension with the federal constitutional provision giving the state legislature exclusive power to direct the appointment of presidential electors.52 The Court also noted the relevance of 3 U.S.C. § 5, observing that “a legislative wish to take advantage of the ‘safe harbor’ would counsel against any construction of the [Florida] Election Code that Congress might deem to be a change in the law.”53 Consequently, the United States Supreme Court vacated the Florida Supreme Court’s action and remanded the case to the Florida court to clarify whether it had been appropriately mindful of Article II, Section 1 and 3 U.S.C. § 5 in its analysis of the Florida Election Code. On remand, the Florida Supreme Court reiterated its earlier decision, taking care this time to ground its reasoning solely on the text of the relevant Florida statutes and its “perception of legislative intent.”54 B. The Contest Phase With Bush certified as the statewide winner, Gore moved under Florida law to contest the certification. Unlike his earlier protests of 49. Bush v. Palm Beach County Canvassing Bd., 531 U.S. 1004 (2000). 50. 3 U.S.C. § 5 (1994). 51. 531 U.S. 70 (2000). 52. Id. at 76-77. 53. Id. at 78. 54. Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d 1273, 1291 (Fla. 2000). The initial Palm Beach decision had been unanimous. The remand was on a 6-1 vote, with Chief Justice Wells dissenting solely based on his opposition to “issuing a new decision while the United States Supreme Court has under consideration Bush v. Gore . . . .” Id. at 1292. Bush v. Gore had been argued prior to the Florida Supreme Court’s decision on remand; the United States Supreme Court’s decision was issued the next day. 338 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 the Broward, Miami-Dade, Palm Beach and Volusia returns, which involved requests of individual county canvassing boards to recheck the tabulations within their counties, the contest was a judicial proceeding, brought in circuit court, to challenge the result of the entire election. The statutory grounds on which Gore relied, however, were similar to the grounds for his protests—“[r]eceipt of a number of illegal votes or rejection of a number of legal votes sufficient to change or place in doubt the result of the election.”55 Gore alleged five instances of the rejection of legal votes or the inclusion of illegal votes which were sufficient to “change or place in doubt the election”: (1) Secretary Harris’ failure to include the results of the Palm Beach County recount, completed just hours after the November 26 deadline; (2) An additional 3,300 undercount ballots which the Palm Beach County canvassing board had examined but declined to treat as legal votes; (3) The results of the partial manual recount undertaken in Miami-Dade County before the Miami-Dade canvassing board had voted to abandon the recount; (4) An additional nine to ten thousand Miami ballots which the Miami-Dade County canvassing board had set aside as undervote ballots but had never reviewed; and (5) Votes identified in the machine recount of Nassau County’s votes that were not included in the certified Nassau result. With respect to Nassau, the statutory machine recount had reduced the county total by 218 votes and clipped Bush’s lead by 51 votes. Although Nassau County originally certified the machine recount figures as the official result, the county canvassing board subsequently voted to rescind the certification and, instead, certified the election-night count, thereby adding to Bush’s lead. Even though this had occurred after November 14, Secretary Harris accepted the results and included them in Bush’s 537-vote margin of victory.56 Following a two-day trial, Judge N. Sanders Sauls of Florida’s Second Circuit ruled on December 3, 2000, that, as a matter of law, in order to prevail in an election contest the challenger must demonstrate a “reasonable probability” that the ballots in question would change the statewide result.57 The court determined that Gore had failed to meet the reasonable probability standard. As a result, Gore 55. FLA. STAT. § 102.168(3)(c) (2000). 56. See generally Gore v. Harris, 772 So. 2d 1243, 1247-48 (Fla. 2000), rev’d sub nom. Bush v. Gore, 531 U.S. 98 (2000); Compl. to Contest Election at 3-4, Gore v. Harris, No. CIV-00-2808 (Fla. 2d Cir. Ct. Nov. 27, 2000), available at http://election2000.stanford.edu/ CV-00-2808a.pdf; 36 DAYS, supra note 13, at 173-74. 57. Gore v. Harris, 2000 WL 1770257 (Fla. 2d Cir. Ct. Dec. 4, 2000), rev’d, Gore v. Harris, 772 So. 2d 1243 (Fla. 2000), rev’d sub nom. Bush v. Gore, 531 U.S. 98 (2000). 2001] EQUAL PROTECTION 339 could not obtain review of the ballots which he claimed would give him enough votes to prevail. The court also determined that the specific decisions of the Miami-Dade, Nassau, and Palm Beach canvassing boards with respect to the inclusion or exclusion of ballots must be sustained unless they constituted an abuse of discretion. The court found that no such abuse of discretion was shown. Gore appealed to the Florida Supreme Court, and on December 8 a divided court ruled in his favor. In Gore v. Harris,58 six of the seven Florida Supreme Court justices ruled that the circuit court had applied erroneous legal standards. They determined that the “abuse of discretion” standard was far too deferential to the county canvassing boards’ decisions,59 and they found that, due to amendments to the Election Code enacted in 1999, an election contest plaintiff need prove only a reasonable possibility, not probability, of success in order to compel the counting of uncounted ballots.60 A four-justice majority then found that an “undisputed showing of the existence of some 9000 ‘undervotes’ [in Miami-Dade County] in an election contest decided by a margin measured in the hundreds [provided] a threshold showing that the result of an election has been placed in doubt, warranting a manual count . . . .”61 The Florida Supreme Court, however, went well beyond Gore’s request for relief and held that, given the statewide nature of the presidential election, “it is absolutely essential in this proceeding and to any final decision, that a manual recount be conducted for all legal votes in this State . . . in all Florida counties where there was an undervote and, hence, a concern that not every citizen’s vote was counted.”62 The focus of the contest litigation in Miami-Dade and Palm Beach—and of the earlier protest litigation in those two counties and Broward County—was largely the result of decisions of the Gore campaign to target their efforts on the counties where a recount was likely to generate the most Democratic votes. But, the court reasoned, the “election should be determined by a careful examination of the votes of Florida’s citizens and not by strategies extraneous to the voting process.”63 The court, thus, required a statewide recount of the undervote.64 The court remanded the case to the circuit court with di58. 772 So. 2d 1243 (Fla. 2000), rev’d sub nom. Bush v. Gore, 531 U.S. 98 (2000). 59. Id. at 1252 (per curiam), 1271 (Harding, J., dissenting). 60. Id. at 1256 (per curiam), 1271 (Harding, J., dissenting). 61. Id. at 1256. Two of the Florida Supreme Court dissenters agreed with the majority in rejecting the abuse of discretion and reasonable probability of success standards. See id. 1270-71 (Harding, J., dissenting). 62. Id. at 1253. 63. Id. 64. Id. Justices Harding and Shaw agreed with the majority concerning the legal standards for a contest but determined that Gore had failed to show “by a preponderance of the evidence, that the outcome of the statewide election would likely be changed” by the 340 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 rections to order the county supervisors of elections and county canvassing boards to conduct a manual recount of the undervotes in all counties that had not previously done so.65 The Florida Supreme Court only briefly addressed a question that had beset the canvassing boards of the three counties that had undertaken manual recounts—“what, under Florida law, may constitute a ‘legal vote?’”66 Florida law defined a ballot to be validly cast if the intent of the voter could be discerned. The recounts that the county canvassing boards had conducted during the protest phase were marked by sometimes heated debates over what constituted sufficient evidence of intent to vote for a candidate: Did a ballot have to have some of its chad detached? Was piercing of the chad, so that light could penetrate, sufficient? Was piercing even necessary, or would an indentation or dimpling of the chad next to a candidate’s name be sufficient to indicate intent to cast a vote? Even the dimpled-chad standard was not entirely straightforward, with some observers arguing that dimpled chads could evidence intent only if all the different offices listed on the ballot were marked by dimpled chads. Others contended that a dimpled presidential chad even without a pattern of dimpled voting was enough.67 Perhaps mindful of the United States Supreme Court’s warning in Bush v. Palm Beach County Canvassing Board against encroaching on the legislature’s federal constitutional prerogative to set the rules for the selection of presidential electors or threatening the federal statutory safe harbor by judicial creation of a postelection law concerning the review of undervote ballots, the Florida court stuck closely to Florida’s statutes and case law, which had defined a “legal vote” as one in which there is a “clear indication of the intent of the voter.”68 The court did reject Gore’s claim concerning the 3,300 ballots examined during the Palm Beach County recount but not included in the county vote total. Gore contended that Palm Beach County’s failure to apply the most expansive application of the inrelief he had sought. Id. at 1272 (Harding, J., dissenting). Justices Harding and Shaw agreed with the majority that any recount would have to be statewide, but the dissenters found that Gore had failed to show that it was reasonably likely he would prevail on a statewide recount. Id. Moreover, they disagreed with the majority’s decision to limit the recount to the undervote, indicating that all no-vote ballots—overvotes as well as undervotes—would have to be manually counted. Id. at 1272-73. They expressed doubt that this could be accomplished by the federal “safe harbor” date of December 12. Id. Only Chief Justice Wells agreed with the circuit court that an abuse of discretion standard applied. He also determined that a fair and accurate statewide recount could not be conducted in the limited time available, and he called for the conclusion of the ballot counting process. Id. at 1266-70 (Wells, C.J., dissenting). 65. Id. at 1262. 66. Id. at 1256. 67. See, e.g., 36 DAYS, supra note 13, at 130, 145-46, 152-53, 159. 68. Gore v. Harris, 772 So. 2d at 1257. 2001] EQUAL PROTECTION 341 tent-of-the-voter standard—which would have treated a dimpled presidential chad as a vote without regard to whether other preferences on the ballot were also marked by dimpled chads—was a legal error. The Florida Supreme Court’s rejection of Gore’s Palm Beach claim, however, was not a decision on the merits concerning how to apply the intent of the voter standard but simply a finding that Gore had “failed to introduce any evidence to refute the Canvassing Board’s determination that the 3300 ballots did not constitute ‘legal votes.’”69 Completing its disposition of Gore’s specific contest claims, the Florida Supreme Court ruled against Gore with respect to the Nassau County vote when it affirmed the circuit court’s decision upholding Nassau County’s use of the original machine count—rather than the machine recount—in determining that county’s vote.70 The court, however, also ruled that the Palm Beach County manual recount— which the county canvassing board had completed but which Secretary Harris had refused to include in her certified count—and the additional votes that had resulted from the partial recount conducted by the Miami-Dade County Canvassing Board had to be immediately included in the statewide total.71 This cut Bush’s lead to less than 200 votes.72 The effect of the Florida Supreme Court’s order was short-lived. The following day the United States Supreme Court stayed the Florida court’s mandate.73 Three days later, the United States Supreme Court reversed, holding that various aspects of the Florida court’s order violated the Equal Protection Clause.74 III. BUSH V. GORE: THE OPINIONS Bush had raised an equal protection argument in his petition for writ of certiorari challenging the Florida Supreme Court’s Palm Beach County Canvassing Board v. Harris decision, but the United States Supreme Court had focused only on the Article II and 3 U.S.C. § 5 questions and declined to grant certiorari on the equal protection question.75 Bush had also raised equal protection arguments in separate efforts to bar county-level recounts prior to the Gore v. Harris 69. Id. at 1260. 70. Id. 71. Id. 72. There was some dispute as to the size of Gore’s net gain in Palm Beach County. Depending on the count, Bush’s lead after the Florida Supreme Court decision was either 154 or 193. 73. See Bush v. Gore, 531 U.S. 1046 (2000). 74. Bush v. Gore, 531 U.S. 98 (2000). 75. See Bush v. Palm Beach County Canvassing Bd., 531 U.S. 1004 (2000). 342 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 contest order, but two Florida federal district courts and the en banc Eleventh Circuit Court of Appeals had rejected his arguments.76 Equal protection, however, dominated the Supreme Court’s review of the Florida Supreme Court’s manual recount order.77 All nine Justices discussed the equal protection question, with six Justices, and possibly seven, finding an equal protection violation. The Court’s reliance on equal protection has potentially enormous significance. Article II, Section 1 and 3 U.S.C. § 5 apply only to presidential elections. Equal protection principles, however, apply to all elections— federal, state, and local. Although the Court self-consciously “limited” its “consideration” “to the present circumstances,”78 Bush v. Gore broke new ground in applying equal protection to state and local election administration and procedure. The case could subject a wide range of state and local election practices to federal constitutional review. A. The Per Curiam Opinion The Bush v. Gore per curiam opinion, joined by Chief Justice Rehnquist and Justices O’Connor, Scalia, Kennedy, and Thomas, found four equal protection problems in the Gore v. Harris order. First, and most importantly, the Florida Supreme Court permitted inconsistent treatment, both among counties and, apparently, among counting teams within counties, in the determination of which ballots would count as legal votes.79 The per curiam reasoned that by not providing more detailed guidance than the “intent of the voter” standard, the Florida court’s manual recount order would lead to the differential treatment of similarly marked ballots in different counties. Indeed, by accepting recount totals from counties that had already conducted their recounts using apparently differing standards, the Florida court had “ratified this uneven treatment.”80 Second, the Florida court’s order also accepted recount results from some counties that had not limited their recounts to undervotes but had also apparently included overvotes.81 As a result, valid votes found on overvote ballots in those counties would be included in the final tally, but comparably valid votes found on overvote ballots cast elsewhere in the state would not be. Third, Gore v. Harris had certified a partial 76. See Siegel v. LePore, 120 F. Supp. 2d 1041 (S.D. Fla. 2000), aff’d, 234 F.3d 1163 (11th Cir. 2000) (en banc); Touchston v. McDermott, 120 F. Supp. 2d 1055 (M.D. Fla. 2000), aff’d, 234 F.3d 1133 (11th Cir. 2000) (en banc). 77. See Bush v. Gore, 531 U.S. 98 (2000), rev’g Gore v. Harris, 772 So. 2d 1243 (Fla. 2000). 78. Id. at 109. 79. See id. at 107. 80. Id. 81. Id. 2001] EQUAL PROTECTION 343 manual recount result from Miami-Dade County. “The Florida Supreme Court’s decision thus gives no assurance that the recounts included in the final certification must be complete.”82 In other words, some ballots might be accepted and included in the final tabulation, but other comparable ballots might not be counted and thus not included in the final result. Although the failure to complete the recount would be due to a lack of time rather than a desire to exclude any voters, “[t]he press of time does not diminish the constitutional concern. A desire for speed is not a general excuse for ignoring equal protection guarantees.”83 Finally, the Court criticized “the actual process” in which the manual recount would be undertaken.84 The order in Gore v. Harris did not specify who would recount the ballots. The county canvassing boards were forced to pull together ad hoc teams [comprised] of judges from various Circuits who had no previous training in handling and interpreting ballots. Furthermore, while others were permitted to observe, they were prohibited from objecting during the recount.85 Although these concerns sound more in due process than in equal protection, the per curiam did not refer to the Due Process Clause and voiced these process concerns in the framework of its equal protection discussion. Presumably, the lack of a proper process would make discrepancies in recount standards more likely to occur and less likely to be corrected. B. Justice Souter Justice Souter’s very brief treatment of the equal protection issue focused exclusively on the question that was the primary focus of the per curiam—the use of different standards in the determination of whether a ballot ought to be counted.86 It is true that the Equal Protection Clause does not forbid the use of a variety of voting mechanisms within a jurisdiction, even though different mechanisms will have different levels of effectiveness in recording voters’ intentions; local variety can be justified by concerns about cost, the potential value of innovation, and so on. But evidence in the record here suggests that a different order of disparity obtains under rules for determining a voter’s intent that have been applied (and could continue to be applied) to identical types of ballots used in identical brands of machines and exhibiting identical physical characteristics. . . . I can conceive of no le82. 83. 84. 85. 86. Id. at 108. Id. Id. at 109. Id. Id. at 134 (Souter, J., dissenting). 344 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 gitimate state interest served by these differing treatments of the expressions of voters’ fundamental rights. The differences appear wholly arbitrary.87 Justice Souter’s opinion did not address whether the inclusion of some overvotes, the certification of incomplete returns, or the Florida Supreme Court’s recount process presented constitutional concerns. C. Justice Breyer Although counted by commentators as one of seven Justices who found the Gore v. Harris order violated the Equal Protection Clause, Justice Breyer’s position was far more equivocal. On the one hand, Justice Breyer joined Justice Souter’s opinion, including the portion of Justice Souter’s opinion that found Bush’s equal protection argument “meritorious.”88 On the other hand, Justice Breyer also joined Justice Stevens’ dissenting opinion, in which Justice Stevens rejected the claim that the Florida Supreme Court’s manual recount order violated equal protection principles.89 In his own separate opinion, Justice Breyer observed that the inconsistencies in counting undervote ballots “implicate principles of fundamental fairness,” and that “in these very special circumstances, basic principles of fairness should have counseled the adoption of a uniform standard to address the problem.”90 But many rulings implicate principles of fairness without violating them, and to say that one course of action is better than a second is not to say that the second is unconstitutional. In his own opinion, Justice Breyer never stated that the Florida Supreme Court’s action violated equal protection or that the Constitution requires a uniform standard for evaluating undervote ballots.91 D. Justices Stevens and Ginsburg Justice Stevens, in an opinion joined by Justices Ginsburg and Breyer, denied that the Florida court’s failure to spell out more detailed operational standards for the application of the intent of the voter standard created an equal protection problem. Noting that “we have never before called into question the substantive standard by 87. Id. (citations omitted). 88. Id. at 133-35 (Souter, J., dissenting). By contrast, Justices Stevens and Ginsburg joined Justice Souter’s opinion except for the part of his opinion finding an equal protection violation. Id. at 129. 89. Id. at 125 (Stevens, J., dissenting). 90. Id. at 145-46 (Breyer, J., dissenting) (emphasis added). 91. The portion of Justice Breyer’s opinion analyzing the equal protection question was joined by Justice Souter but not by Justices Stevens and Ginsburg, suggesting perhaps that his Supreme Court colleagues thought Justice Breyer was agreeing with Justice Souter even if Justice Breyer’s language did not go quite that far. 2001] EQUAL PROTECTION 345 which a State determines that a vote has been legally cast,”92 Justice Stevens concluded there was “no reason to think that the guidance provided [by that standard] is any less sufficient—or will lead to results any less uniform—than, for example, the ‘beyond a reasonable doubt’ standard employed everyday by ordinary citizens in courtrooms across this country.”93 He concluded that the concern over different standards in different counties was “alleviated—if not eliminated—by the fact that a single impartial magistrate will ultimately adjudicate all objections arising from the recount process.”94 He also suggested that one implication of the Court’s decision is that “Florida’s decision to leave to each county the determination of what balloting system to employ—despite enormous differences in accuracy— might run afoul of equal protection.”95 In her separate opinion, Justice Ginsburg, joined by Justice Stevens, was even more dismissive of the equal protection claim. Noting that “we live in an imperfect world, one in which thousands of votes have not been counted,” she could not “agree that the recount adopted by the Florida court, flawed as it may be, would yield a result any less fair or precise than the certification that preceded that recount.”96 IV. EQUAL PROTECTION AND THE FLORIDA SUPREME COURT’S MANUAL RECOUNT ORDER A. The Equal Protection Clause and the Vote There is no federal constitutional right to vote. The Constitution of its own force enfranchises no one. Article I, Section 2 sets the tone by looking to state law for the determination of who may vote in federal elections when it provides that the electorate for selecting members of the House of Representatives shall consist of “Electors in each State [who] shall have the Qualifications requisite for Electors of the most numerous Branch of the State Legislature.” The Seventeenth Amendment makes the same provision for the electorate that chooses United States Senators. And, as we were so forcefully reminded by both Bush v. Palm Beach County Canvassing Board97 and Bush v. Gore,98 the Constitution gives the people no vote in the presidential 92. Id. at 125 (Stevens, J., dissenting). 93. Id. 94. Id. at 126. 95. Id. 96. Id. at 143 (Ginsburg, J., dissenting). 97. 531 U.S. 70, 76-77 (2000). 98. 531 U.S. at 104 (“The individual citizen has no federal constitutional right to vote for electors for the President of the United States unless and until the state legislature chooses a statewide election as the means to implement its power to appoint members of the Electoral College.”). 346 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 election at all. As for the voting rights provisions of the Fifteenth, Nineteenth, Twenty-fourth and Twenty-sixth Amendments, none of these confer a right to vote in federal, state, or local elections. Rather, each is phrased in the negative, eliminating a qualification that a state or locality might otherwise have utilized to determine who may exercise the franchise but not requiring that anyone actually be enfranchised. For most of American history, constitutional doctrine joined with constitutional text in denying the existence of a general constitutionally protected right to vote. In 1875, in Minor v. Happersett,99 the Supreme Court emphatically rejected the argument that the right to vote is a right of citizenship. In that case, Minor, a woman, argued that, by virtue of the Fourteenth Amendment’s declaration, “[a]ll persons born or naturalized in the United States, and subject to the jurisdiction thereof” are citizens of the United States and of the states in which they reside. Minor was a citizen of Missouri and therefore asserted that she was entitled to vote in Missouri federal and state elections, notwithstanding the fact that Missouri law limited the franchise to men. The Supreme Court agreed that she was a citizen and thus a member of the “political community,”100 but found that citizenship had no relevance to the question of whether she was entitled to vote. Voting was simply not an attribute of citizenship. The scope of the franchise was a matter entirely for state determination, subject only to the Fifteenth Amendment’s preclusion of racial discrimination in voting. Although Minor’s specific determination that a state could limit the vote to men was overturned by the Nineteenth Amendment, Minor’s central premise, that the Constitution provides no general protection for the right to vote, was still good law almost eighty-five years later. In 1959, in Lassiter v. Northampton County Board of Elections,101 the Court, noting that “[t]he States have long been held to have broad powers to determine the conditions under which the right of suffrage may be exercised,”102 upheld the constitutionality of a state literacy test. To be sure, the Court had become vigorous in detecting and invalidating state suffrage laws that violated the Constitution’s specific ban on racial discrimination, and the Court had repeatedly recognized Congress’s power to regulate the suffrage in federal elections. But apart from the voting criteria explicitly condemned by specific provisions of the Constitution, there was no federal constitutional protection of the right to vote. 99. 100. 101. 102. 88 U.S. (1 Wall.) 162 (1875). Id. at 165. 360 U.S. 45 (1959). Id. at 50. 2001] EQUAL PROTECTION 347 All that changed during the voting rights revolution of the 1960s and early 1970s. Beginning with the legislative apportionment cases,103 and then turning to restrictions on the franchise itself, the Supreme Court transformed the constitutional status of the vote. Voting became a fundamental right, with laws infringing that right subject to strict judicial scrutiny. In short order, the Court invalidated such longstanding limitations on suffrage such as the poll tax,104 property tax payment requirements,105 and durational residency requirements.106 The Court also established one person, one vote as the constitutional ground rule for political representation in elected bodies. Although the Court adhered to the traditional federal constitutional formula of invalidating criteria for voting rather than creating an affirmative entitlement to vote, the holdings and reasoning of the Court effectively established such an entitlement. Inclusion and equality are the twin hallmarks of the new jurisprudence of voting rights. Once a state or locality provides that an election is used to fill a public office or to answer a governmental question, then all adult citizens who are residents of the jurisdiction are presumptively entitled to vote in that election, and all voters must have equally weighted votes. Voting was transformed from a matter of legislative grace into a fundamental aspect of citizenship. A defining characteristic of citizenship is the opportunity to participate in political decisionmaking. “Most citizens can achieve this participation only as qualified voters through the election of legislators to represent them. Full and effective participation by all citizens . . . requires, therefore, that each citizen have an equally effective voice in the election o f” public officials.107 The assumption that citizenship means the right to vote was most clearly underscored in the way the Court in Reynolds v. Sims nonchalantly equated citizenship with the suffrage in the phrase “[a] citizen, a qualified voter.”108 The extension and protection of the right to vote was a critically important official public statement of the voter’s status as a citizen. Disenfranchisement and malapportionment were particularly troubling not simply because they interfered with political participation but because they expressed a state’s determination that the excluded and the underrepresented were less than full citizens. 103. See, e.g., Reynolds v. Sims, 377 U.S. 533 (1964). 104. See Harper v. Virginia Bd. of Elections, 383 U.S. 663 (1966). 105. See Hill v. Stone, 421 U.S. 289 (1975); City of Phoenix v. Kolodziejski, 399 U.S. 204 (1970); Cipriano v. City of Houma, 395 U.S. 701 (1969). 106. See Dunn v. Blumstein, 405 U.S. 330 (1972). 107. Reynolds, 377 U.S. at 565. 108. Id. at 568. 348 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 The Court’s notion of the right to vote as a signifier of full citizenship may also explain the handful of Court-approved exclusions from the franchise. Residency is the state and local equivalent of citizenship; it determines whether one is a member of a particular state or local political community.109 As a result, nonresidents have no right to vote where they do not reside.110 Felons may be denied the franchise in part because conviction of a felony has long had the metaphoric status of loss of political citizenship.111 Finally, the Court found that certain bodies, such as quasi-proprietary special districts, are not political communities. As such they lacked citizens, and citizen entitlements to suffrage and equally weighted votes do not apply to voting arrangements in such special districts.112 In addition to tightly linking voting to citizenship, the Court justified its new protection of the franchise instrumentally. The vote is a critical tool that enables citizens to protect their rights and interests. “[T]he right to exercise the franchise in a free and unimpaired manner is preservative of other basic civil and political rights.”113 Beyond rights, the franchise is the key means by which those affected by government actions can make their interests known and their voices heard.114 Universal adult citizen suffrage is thus necessary to legitimate government as representative of the people. Only when all adult citizens are free to vote can we assume that elected officials are representative of and accountable to the people as a whole. Finally, the Court defended its intervention in an area traditionally left to the states in comparative institutional terms. Because the current distribution of the franchise and the current weighting of votes determines who holds political power, the political process could not be counted on to correct laws that result in disenfranchisement and underrepresentation: The presumption of constitutionality and the approval given “rational” classifications in other types of enactments are based on an assumption that the institutions of state government are structured so as to represent fairly all the people. However, when the challenge to the statute is in effect a challenge of this basic as- 109. “An appropriately defined and uniformly applied requirement of bona fide residence may be necessary to preserve the basic conception of a political community, and therefore could withstand close constitutional scrutiny.” Dunn, 405 U.S. at 343-44. 110. See Holt Civic Club v. City of Tuscaloosa, 439 U.S. 60 (1978). 111. See Richardson v. Ramirez, 418 U.S. 24 (1974). 112. See, e.g., Ball v. James, 451 U.S. 355 (1981); Salyer Land Co. v. Tulare Lake Basin Water Storage Dist., 410 U.S. 719 (1973). See generally Richard Briffault, Who Rules at Home? One Person, One Vote and Local Governments, 60 U. CHI. L. REV. 339, 359-84 (1993). 113. Reynolds, 377 U.S. at 562. 114. See Kramer v. Union Free Sch. Dist. No. 1, 395 U.S. 621, 630-33 (1969). 2001] EQUAL PROTECTION 349 sumption, the assumption can no longer serve as the basis for presuming constitutionality.115 Judicial intervention was necessary—and strict judicial scrutiny of restrictions on the franchise was called for—because the self-favoring biases of political insiders, who were elected by a limited electorate or under voting rules that overrepresented some groups, made departures from universal adult resident citizenship and equally weighted votes both inherently suspect and unlikely to be corrected by the political branches without judicial intervention. The overall thrust of the Court’s application of the Equal Protection Clause to the vote was the expansion of the franchise. In theory, the inequality created by laws that unequally enfranchised some and disenfranchised other similarly situated people could have been remedied by disenfranchising the enfranchised. But that result never occurred. In the right to vote cases, the Court’s focus was less on inequality per se and more on extending the franchise to groups of adult resident citizens who had previously been excluded from voting. B. Equal Protection and the Administration of Elections: Setting the Standard for Federal Judicial Intervention As the Supreme Court noted in Bush v. Gore, constitutional protection of the right to vote goes well beyond “the initial allocation of the franchise.”116 Equal protection of the vote applies to the weighting of votes in the election of officials;117 the design of systems of representation, including the drawing of legislative districts and the selection of single- or multi-member districts; and the rules, such as those regulating the ability of candidates and parties to get on the ballot, that determine the range of options available for casting one’s vote.118 Equal protection could reasonably be extended to the state and local rules that govern the casting and counting of ballots. Certainly if a state law mandated the use of two different kinds of election machinery, with different error rates, in different parts of the state, the voters in the area required to use the machine with the higher error rate—so that a higher percentage of their votes were legally ignored—could reasonably contend that their votes were unconstitutionally underweighted compared to the votes of residents with better machinery. Similarly, if the state imposed a fee for submitting an absentee ballot, that fee could be challenged by voters unable to get to the polls as a form of wealth tax on voting and would be subjected to strict judicial scrutiny. So too, if county officials provided assis115. 116. 117. 118. Id. at 628. 531 U.S. 98, 104 (2000). Gray v. Sanders, 372 U.S. 368 (1963). Williams v. Rhodes, 393 U.S. 23 (1968). 350 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 tance to Republican party workers, but denied comparable assistance to Democrats in filling out absentee ballot applications, that would plainly be subject to constitutional challenge as partisan discrimination affecting the vote.119 Election administration practices, however, may present a more complicated problem for equal protection analysis than state laws affecting the formal right to vote or the weighting of representation in a multi-member elected body. Whereas adult citizen suffrage and one person, one vote are well-established constitutional norms, there are no constitutionally mandatory standards or widely accepted requirements for many aspects of the election process. To use just some of the examples that surfaced during the Florida recount struggle, there are multiple types of voting machinery, different types of ballot design, and a variety of requirements and procedures for casting absentee ballots and for protesting and contesting elections. To subject all of these diverse practices and rules to equal protection review could ultimately result in a federal court-ordered, nationwide standardization of the mechanics of elections. In many of these cases, the basis for a judicial determination of one constitutionally mandatory election procedure would be far from clear. Moreover, judicial standardization would undermine, if not end, the longstanding American tradition of decentralized control of elections—a tradition going back at least to the late eighteenth century when, according to Alexis de Tocqueville, Massachusetts local officials prepared the voting lists for state elections and transmitted the results of the local poll to state officials.120 Indeed, during much of American history, local governments set their own voting rules for participation in local elections. As a result, a person could be qualified to vote in a colonial or state election but not in a local election or vice versa.121 Of course, much as “it is revolting to have no better reason for a rule of law than that so it was laid down in the time of Henry IV,”122 tradition alone cannot justify a practice inconsistent with current values. Decades of history did not save malapportionment and durational residency requirements when the Supreme Court began to 119. Cf. Jacobs v. Seminole County Canvassing Bd., 773 So. 2d 519, 523 (Fla. 2000) (The court found that the county supervisor of elections provided special assistance to the Republican party in submitting requests for absentee ballots, but that “there was no evidence that such a request was made by the Democratic party” and no evidence of a denial of requested assistance to Democrats. “Thus, there was no adequate showing that there was disparate treatment of Republicans as opposed to any other individuals or groups with regard to the ballot request forms.”). 120. See ALEXIS DE TOCQUEVILLE, DEMOCRACY IN AMERICA 73 n.14 (J.P. Mayer ed., 1969). 121. See ALEXANDER KEYSSAR, THE RIGHT TO VOTE: THE CONTESTED HISTORY OF DEMOCRACY IN THE UNITED STATES 6, 20-21, 30, 186-87, 200-01 (2000). 122. Oliver Wendell Holmes, The Path of the Law, 10 HARV. L. REV. 457, 469 (1897). 2001] EQUAL PROTECTION 351 apply an equal protection analysis to voting and representation laws in the 1960s and 1970s. Local decisionmaking, however, does continue to reflect and reinforce important contemporary political values. Local decisionmaking curbs abuses of power by the upper level government; builds democracy; increases the satisfaction of citizen preferences; and facilitates innovation, experimentation and political learning. De Tocqueville stressed the first two points in Democracy in America, when he focused on the political benefits of “Administrative Decentralization in the United States.”123 Given the absence of institutional curbs on state power, de Tocqueville found that the practice of decentralization played an important role in preventing state tyranny.124 Moreover, de Tocqueville expressed the concern that the equality and individualism that accompany democracy can make it difficult for people to cooperate and thus make them easy targets for despotism. By giving Americans an interest and an opportunity to participate in self-government—and thus a “habit and taste” for working together concerning public matters—local government strengthened Americans’ commitment to their own freedom. In his view, local government was a sort of “primary school” of democracy.125 Decentralized decisionmaking enabled people to become “citizens”; that is, active participants in self-governance who, due to that participation, would become committed to maintaining and defending self-government. Decentralization thus serves the same interests as voting itself. Decentralized power as a break on centralized tyranny is still an important theme in contemporary arguments for decentralization. It may also be directly relevant to the conduct of manual recounts. As Judge Middlebrooks observed in rejecting a federal court challenge to the Florida 2000 presidential recounts then underway in Broward, Miami-Dade, Palm Beach, and Volusia Counties, [r]ather than a sign of weakness or constitutional injury, some solace can be taken in the fact that no one centralized body or person can control the tabulation of an entire statewide or national election. For the more county boards and individuals involved in the electoral regulation process, the less likely it becomes that corrup- 123. DE TOCQUEVILLE, supra note 120, at 87. 124. See id. at 89. 125. John Stuart Mill used a similar metaphor, noting that participation in local government “may be called the public education of the citizens.” JOHN STUART MILL, CONSIDERATIONS ON REPRESENTATIVE GOVERNMENT 288 (Prometheus Books 1991). Thomas Jefferson also emphasized democracy’s dependence on the training in citizenship provided by local self-government: “[B]y giving to every citizen, personally, a part in the administration of the public affairs,” “and in the offices nearest and most interesting to him,” local government “will attach him by his strongest feelings to the independence of his country, and its republican constitution.” Thomas Jefferson, Letter to Samuel Kercheval: July 12, 1816, in THOMAS JEFFERSON: WRITINGS 1395, 1399-1400 (1984). 352 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 tion, bias, or error can influence the ultimate result of an election.126 Certainly, in the Florida recount fight it was striking how Secretary Harris’ decisions consistently favored the interests of her party’s candidate—whom she herself had served as state campaign co-chair during the election. In Florida, decentralization served to reduce the ability of one party to make all the administrative decisions affecting the recount. Administrative decentralization may also be of value in building democracy and increasing the ability of government to satisfy divergent preferences. Justice O’Connor emphasized these benefits of decentralized decisionmaking when she argued that “decentralized government . . . will be more sensitive to the diverse needs of a heterogeneous society; it increases opportunity for citizen involvement in democratic processes; [and] . . . it makes government more responsive . . . .”127 These concerns seem apposite in the election administration setting. Election administration involves the juggling of multiple values: facilitating voting, promoting innovation, achieving accurate and timely tabulations, assuring fairness to candidates and parties, controlling costs, relying on volunteers to do the nitty-gritty of administration,128 and permitting opportunities for challenges while also providing for finality of results. Different states and localities may balance these values differently and their different conclusions can affect the rules they adopt. Decentralization is particularly valuable where there is no one right answer, reasonable people disagree, and those disagreements correlate with residence in a particular local government. People in rural areas may set the balance in one direction, and people in urban areas may set it differently. People in areas dominated by one ethnic group, social class, age group or first-time voters may have one preference, while other areas with different demographics might feel differently. Where there is considerable intrastate disagreement on what the right rule ought to be—and no particular rule is constitutionally required—a uniform, statewide rule could have the unfortunate effect of forcing a lot of people to live under a rule they oppose. Decentralized decisionmaking, with local rules responsive to local 126. Siegel v. LePore, 120 F. Supp. 2d 1041, 1052 (S.D. Fla. 2000), aff’d, 234 F.3d 1163 (11th Cir. 2000). 127. Gregory v. Ashcroft, 501 U.S. 452, 458 (1991). 128. See, e.g., Bodine v. Elkhart County Election Bd., 788 F.2d 1270, 1272 (7th Cir. 1986) (“[E]lections are generally conducted by volunteers, rather than trained professionals. While this may be a positive aspect of the electoral system it inevitably leads to errors of widely differing degrees of severity.”) (citing Hennings v. Grafton, 523 F.2d 861 (7th Cir. 1975)); Duncan v. Poythress, 657 F.2d 691, 701 (5th Cir. Unit B 1981) (“[E]rrors are inevitable in a society which relies upon volunteers to conduct most elections.”). 2001] EQUAL PROTECTION 353 preferences, increases the number of people likely to be satisfied with the rule under which they are governed. Decentralization and variation may also promote innovation. It may be easier to try out new forms of voting machinery, new types of ballot design, technological developments like electronic voting, or experiments like mail-in voting, two-day voting, or Sunday voting in one or a handful of jurisdictions rather than in the state or nation as a whole. Equal protection could jeopardize such innovation by requiring the immediate statewide application of innovations in voting before any difficulties in administration or unintended consequences have been studied and resolved. Alternatively, innovation could be discouraged if the constitutionalization of questions of administration result in liability for jurisdictions whose innovations misfire and unintentionally burden voting. This does not mean that local control of all aspects of election administration is necessarily desirable. The Florida presidential recount revealed to the nation many of the shortcomings of local control—variations in the quality of election machinery that relate to differences in local resources; ill-considered judgments like Palm Beach County’s butterfly ballot; the close relationships between election administrators and local party workers that affected applications for absentee ballots; and the lack of consistent standards for resolving questions concerning the validity of disputed ballots. State legislative determination of ballot design and more precise standards for resolving counting disputes, state financing of high-quality machinery for all counties, and vigorous state oversight of local election officials’ cooperation with party workers could certainly promote a more equitable, reliable, and effective electoral process. But, given both the tradition of decentralized administration and the values that support it, the determination of whether particular administrative questions are resolved at the state or at the local level is primarily a matter for the state political process, not federal constitutional law. Certainly, local election administration may be marred by errors, irregularities and minor violations of state election laws that infringe the voters’ rights. Election machines can be improperly programmed,129 break down,130 or be installed too late for some voters to use.131 Ballots and absentee ballot applications can be mishandled.132 The states, however, have rules and procedures for addressing these 129. See, e.g., Gamza v. Aguirre, 619 F.2d 449 (5th Cir. 1980). 130. See, e.g., Hennings v. Grafton, 523 F.2d 861 (7th Cir. 1975). 131. See, e.g., Gold v. Feinberg, 101 F.3d 796 (2d Cir. 1996). 132. See, e.g., Welch v. McKenzie, 765 F.2d 1311 (5th Cir. 1985); Pettengill v. Putnam County R-1 Sch. Dist., 472 F.2d 121 (8th Cir. 1973). 354 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 “garden variety” election disputes.133 In the absence of a showing that these measures are insufficiently attentive to voting rights or that an election dispute raises more serious issues, there is little basis for federal judicial intervention. Most problems of election administration, including rules—or violations of rules—that burden the vote in individual cases simply do not raise the kinds of concerns that triggered the Supreme Court’s intervention to protect voting rights. Random machine breakdowns, unpatterned mishandling of ballots, and intermittent irregularities in absentee ballots do not undermine the political rights of affected voters, signify that they are second-class citizens, or threaten their ability to advance their rights and interests in the political process. It will often not be known precisely which voters’ ballots were lost due to mechanical problems or election day errors, and voters whose ballots were lost in a particular election or for a particular race are not entirely excluded from the electorate but may be able to vote in other elections or for other races in the same election. State remedies may be able to catch and correct errors or prevent their recurrence in the future. The voters affected by administrative mistakes are as much members of the polity, and able to vindicate their interests over time, as voters whose votes were counted. Of course, to the extent that maladministration is intentional, recurring, predictable, and targeted at particular groups of voters, with state remedies failing to correct the problem, a constitutional problem is presented. Such administrative practices compromise the integrity of the election. The voters whose ballots are consistently not counted are burdened with a reduced opportunity to participate, are treated disrespectfully by their states, and suffer a reduction in their political power. They may be, for voting purposes, second-class citizens. In their cases dealing with problems of election administration, the lower federal courts have repeatedly distinguished between ordinary, or “garden variety,” election irregularities and election practices that reach “the point of patent and fundamental unfairness.”134 In cases involving such standard problems as “the malfunctioning of voting machines and innocent human errors,”135 the federal courts have deferred to state and local control of election administration and have avoided finding that election irregularities—including violations of state election laws and actions interfering with the ability 133. Duncan, 657 F.2d at 701. 134. Griffin v. Burns, 570 F.2d 1065, 1077 (1st Cir. 1978). 135. Duncan, 657 F.2d at 701. Accord Gold, 101 F.3d at 796; Bodine v. Elkhart County Election Bd., 788 F.2d 1270 (7th Cir. 1986); Gamza v. Aguirre, 619 F.2d 449 (5th Cir. 1980). 2001] EQUAL PROTECTION 355 to cast a ballot or have it properly counted136—violated the constitutional right to vote. On the other hand, election administrative actions involving racial discrimination,137 intentional and widespread disenfranchisement,138 or changes in election rules on which voters had reasonably relied to their detriment in deciding whether and how to vote,139 have been found to constitute the kind of fundamental unfairness that rises to the level of a constitutional violation. The lower federal courts’ focus on fundamental unfairness not only holds together protection of the vote with respect for state and local control over elections, but it is also consistent with the Supreme Court’s approach in other settings for determining whether state laws burden voting rights. In the legislative apportionment context, for example, the Court has extended the right to an equally weighted vote from its original use in invalidating districts of unequal population to partisan gerrymandering, recognizing that such gerrymandering can diminish the effectiveness of the votes of political minorities.140 But the adoption of a districting plan with the intent—and effect—of giving one party a higher percentage of legislative seats than its percentage of the popular vote (and the concomitant reduction in the percentage of seats relative to votes for the other parties) is not enough to sustain an action for unconstitutional partisan gerrymandering. Rather, the Court has held that in order to violate the Equal Protection Clause the gerrymander has to be so severe that it “will consistently degrade a voter’s or a group of voters’ influence on the political process as a whole.”141 In other words, intentional gerrymandering alone is not unconstitutional; only an intentional gerrymander that consistently undermines the votes of a partisan group over time is unconstitutional. Similarly, the Court has recognized that state laws that limit the ability of third parties and independents to win a place on the ballot burden the rights of voters to cast effective votes.142 But the Court has held constitutional state laws requiring independents to demonstrate some substantial level of support in order to be listed on the ballot;143 limiting the ability of primary election losers to run as inde- 136. See, e.g., Gold, 101 F.3d at 796; Bodine, 788 F.2d at 1270; Welch, 765 F.2d at 1311; Hennings v. Grafton, 523 F.2d 861 (7th Cir. 1975); Pettengill, 472 F.2d at 121. 137. See, e.g., Bell v. Southwell, 376 F.2d 659 (5th Cir. 1967). 138. See Duncan, 657 F.2d at 691. 139. See Roe v. Alabama, 43 F.3d 574, 581-82 (11th Cir. 1995); Griffin, 507 F.2d at 1065. 140. See Davis v. Bandemer, 478 U.S. 109, 118-27 (1986). 141. Id. at 132. 142. See Williams v. Rhodes, 393 U.S. 23, 30 (1968). 143. See Jenness v. Fortson, 403 U.S. 431 (1971). 356 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 pendents in the general election;144 and refusing to count write-in ballots.145 The right to vote includes a right to have choices, but laws limiting the range of choices do not unconstitutionally burden the right to vote. So long as a state’s laws provide minor parties and independents with a reasonable opportunity to get on the ballot, the state can enforce some laws that keep candidates and parties off the ballot even though that has the effect of narrowing the range of choices and, thus, constraining the right to vote. “Election laws will invariably impose some burden upon individual voters.”146 Only unduly burdensome limits on the ability of new parties and candidates to get on the ballot violate the right to vote. Where a state provides “adequate ballot access,” laws such as a ban on write-in voting are constitutional even though they impose a “burden on voters’ rights to make free choices and to associate politically through the vote.”147 Did the Florida Supreme Court’s order, with its tacit approval of varying county-level standards for determining whether an undervote ballot contained a legally valid vote, create a situation of fundamental unfairness for Florida’s presidential voters? That determination involves consideration of the constitutional status of the undervote ballots, the cause of the variation in standards, and the justifications for the Florida court’s action. The next three sections of this Part take up those issues. C. The Uncertain Constitutional Status of Undervote Ballots Are undervote ballots votes entitled to the full constitutional protection available to votes? The answer is not clear, but probably “no.” Conceivably, there could be three kinds of undervotes. First, there are ballots that show no markings whatsoever with respect to any of the presidential candidates. Such a ballot reflects the voter’s choice not to vote for President. The voter might have come to the polls to vote in another race but decided not to vote in the presidential election. There is no dispute that such ballots should be treated as nonvotes. Second, some undervotes might be attributable to machine error. Such a ballot would have been properly marked, with the chad cleanly detached, but for some reason the vote-counting machinery failed to record a vote. There was no question that a voter who cast such a ballot is entitled to have that ballot counted as a vote. The crux of the conflict in Bush v. Gore was a third type of undervote: a ballot that reflects some marking of the ballot card next to a presi144. See Munro v. Socialist Workers Party, 479 U.S. 189 (1986); Storer v. Brown, 415 U.S. 724, 735-36 (1974). But cf. Anderson v. Celebrezze, 460 U.S. 780 (1983) (invalidating an early filing deadline for presidential candidates). 145. See Burdick v. Takushi, 504 U.S. 428 (1992). 146. Id. at 433. 147. Id. at 438-39. 2001] EQUAL PROTECTION 357 dential candidate’s name but not enough to detach the chad, thereby causing the ballot-counting machinery to treat the ballot as not including a choice of candidate for President. Under Florida law and the law of many states, such imperfectly marked ballots may be valid where the markings reflect the intent of the voter to cast a ballot. But not all markings on a ballot demonstrate the intent of the voter to cast that ballot. Some markings are just stray marks. Others may be indications of a tentative disposition to cast the ballot followed by a final decision not to. In other words, some but not all imperfectly marked ballots are votes, while some imperfectly marked ballots are not votes. There is no federal constitutional or statutory standard for determining what counts as a valid ballot, and Bush v. Gore declined to establish one. Although some states, by legislative or judicial ruling, have adopted relatively specific standards, there is no consistency across the states. Different state standards include requiring the chad be detached in two corners,148 requiring only that some light penetrate the ballot,149 or accepting dimples or indentations without any detachment or penetration at all.150 Many states have no clearly articulated standard more precise than the intent of the voter.151 In Florida, prior to the 2000 election, neither the legislature, nor the Secretary of State—as the state’s chief administrative officer with responsibility for elections—nor the courts had spelled out criteria for the determination of when an undervote ballot demonstrates the intent of the voter to cast a vote. Constitutional protection of the vote does not require the broadest possible definition of a validly cast undervote ballot. The determina148. See, e.g., In re Issue 27 Election of November 4, 1997, 693 N.E.2d 1190, 1191 (Ohio Com. Pl. 1998) (holding that a ballot will be counted only if marked by a “hanging chad,” that is, a chad “attached by two or less corners”); cf. Duffy v. Mortenson, 497 N.W.2d 437, 439-40 (S.D. 1993) (finding that a ballot must be counted where two of the four corners of the chad were detached). 149. See, e.g., IND. CODE ANN. § 3-12-1-9.5(c) (Michie 2000) (“A chad that has been pierced, but not entirely punched out of the card, shall be counted . . . .”); id. § 3-12-1-9.5(d) (“A chad that has been indented, but not in any way separated from the remainder of the card may not be counted . . . .”). 150. See, e.g., TEX. ELEC. CODE ANN. § 127.130(d)(3) (Vernon 2000) (A ballot may be counted where “an indentation on the chad from the stylus or other object is present and indicates a clearly ascertainable intent of the voter to vote”); Delahunt v. Johnston, 671 N.E.2d 1241, 1243 (Mass. 1996) (“[T]he presence of a discernible impression made by a stylus” counts as a clear indication of a voter’s intent.). 151. See, e.g., Pullen v. Mulligan, 561 N.E.2d 585, 610 (Ill. 1990). Although the legislature certainly has the power to provide a mandatory standard for marking punch card ballots, as it did for the marking of paper ballots, no such standard has been set out in the Election Code. We would be usurping the power of the legislature if we were to infer such a standard in the Election Code and then conclude that the legislature intended such standard to be given a mandatory construction. Id. 358 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 tion of whether an imperfectly marked ballot contains a vote involves the balancing of a variety of competing concerns. These include assuring that all votes are counted, avoiding the mistaken tabulation of nonvotes, and maintaining the political neutrality of the process. Respecting the intent of the voter entails both counting all ballots intentionally cast and not counting ballots not intended to be cast for a candidate. Moreover, some standards, like the inclusion of dimpled or indented ballots, are less precise than the requirements of detachments or penetration by light. So there is a tension between the benefits of a more inclusive standard in assuring that all intentionally cast ballots are counted and the dangers of giving a greater role to the subjective perceptions of the individual ballot counters. In the right to vote cases, universal suffrage was a presumption of democratic theory and quickly became a constitutional benchmark. The denial of the right to vote was an unusual deviation from the suffrage norm that had to be justified. Although in the legislative apportionment context, there was no universal practice of equipopulous representation, the one person, one vote principle quickly emerged because of its close correspondence with notions of majority rule and equal voting rights and its relative ease of administration. But there is no comparable baseline in state electoral practices or democratic theory for deciding how to count imperfectly marked ballots. The Constitution does not require a specific test for counting undervotes, nor does it necessarily prefer a more inclusionary approach over a more restrictive one. Presumably, each of the tests employed by the states and by the individual county canvassing teams in Florida— partially detached chad, pierced chad, and dimpled chad—were constitutional. In any event, the Court made no statement that any of the divergent standards, including the most restrictive, was invalid. Thus, it would not have been an unconstitutional infringement on the right to vote if any counting team—or all the counting teams— had adopted a relatively restrictive approach to applying the intent of the voter standard and had excluded most of the imperfectly marked ballots, even if they contained discernable marks. Indeed, three members of the Court—constituting a majority of those who signed the per curiam opinion—indicated their view that as a matter of Florida law none of the imperfectly marked ballots should be counted as valid votes. Chief Justice Rehnquist, joined by Justices Scalia and Thomas, pointing to the directive given to Florida voters to “AFTER VOTING, CHECK YOUR BALLOT CARD TO BE SURE YOUR VOTING SELECTIONS ARE CLEARLY AND CLEANLY PUNCHED AND THERE ARE NO CHIPS LEFT HANGING ON THE BACK OF THE CARD,” as well as to the specific provisions of the Florida election contest law, concluded that ballots not counted by properly functioning machinery because the bal- 2001] EQUAL PROTECTION 359 lots were imperfectly marked should not be counted at all.152 Florida’s Secretary of State had taken the same position, arguing in effect that the only undervotes that should be counted are those that a properly functioning machine would have registered. Implicitly, Chief Justice Rehnquist and Justices Scalia and Thomas must have concluded that failure to count all imperfectly marked ballots was constitutional. The concurring Justices actually used the term “improperly marked,” suggesting that the voters were somehow at fault for the failure to register a proper vote. For the concurring group at least, a complete exclusion of imperfectly marked ballots does not deny equal protection of the laws to the voters whose ballots were not counted since those voters had no right to have their ballots counted at all. The concurring opinion aside, there was nothing in the per curiam opinion that indicated that a voter who casts any particular sort of imperfectly marked ballot had any substantive entitlement to have that ballot treated as a valid vote. Indeed, the per curiam Justices, like their concurring brethren, also appeared to conclude that undervote ballots were of minimal constitutional status. As the per curiam opinion observed in finding that the Florida Supreme Court’s certification of Miami-Dade County’s partial recount results violated equal protection, “[t]he press of time does not diminish the constitutional concern. A desire for speed is not a general excuse for ignoring equal protection guarantees.”153 If equal protection guarantees applied to imperfectly marked ballots, then presumably even “the press of time” would not have justified the failure to review them. Yet the effect of the per curiam decision was to terminate the recount of Florida’s undervote ballots because of the impending federal statutory safe harbor deadline—in other words, “the press of time.” If “the press of time” does not excuse ignoring equal protection guarantees but does require the termination of the statewide manual recount of undervote ballots, then, logically, those ballots are not entitled to much protection under the Equal Protection Clause. Imperfectly marked ballots appear to fall into a constitutional gray area. They might be votes and they might not be votes. That does not mean that such ballots must be ignored. A state or locality may choose to go beyond the constitutionally mandated right to vote 152. Bush v. Gore, 531 U.S. 98, 119 (2000). That was also the view of three members of the U.S. Court of Appeals for the Eleventh Circuit who dissented from that court’s affirmance of the district court’s refusal to preliminarily enjoin, on equal protection grounds, the manual recounts under way before Gore v. Harris. See Touchston v. McDermott, 234 F.3d 1133, 1141-45 (11th Cir. 2000) (Tjoflat, J., dissenting, joined by Birch and Dubina, JJ.). 153. Bush v. Gore, 531 U.S. at 108. 360 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 and enfranchise those, such as noncitizens154 or nonresidents,155 who do not have a constitutional entitlement to vote but have a sufficient tie to the polity that it is reasonable to include them. Such an extension of the franchise does not unconstitutionally “dilute” the votes of the members of the electorate who are constitutionally entitled to vote.156 As a result, even if Florida did not have to count the undervote ballots, the state could do so. Any undervote ballots found to reflect an intent to vote for a presidential candidate and added to the tally would not have unconstitutionally diluted the votes recorded by machines. 157 The uncertain constitutional status of the undervote ballot does not mean that the Equal Protection Clause does not apply when some imperfectly marked ballots are reviewed and counted and some are not. Certainly, it would be unconstitutional for a canvassing board to count only those undervote ballots marked for a Democrat while ignoring those marked for a Republican. But the status of the undervote ballots—the fact that they could be ignored and that no particular criterion for counting such a ballot is either constitutionally mandated or clearly implicated by constitutional protection of the vote—affects the sense of whether variations in the standards for assessing undervote ballots creates an equal protection violation. The only voters in Florida’s 2000 presidential election who were arguably denied equal protection by the manual recount order were those who cast imperfectly marked ballots that might have been treated as valid in a county applying a liberal “intent of the voter” standard but whose ballots would not have been found valid in the voter’s own county. Such a voter has no constitutional entitlement to have her ballot reviewed at all and certainly has no entitlement to have her ballot reviewed under a liberal intent standard. The state could have adopted a restrictive, detached chad standard that would 154. See, e.g., Jamin B. Raskin, Legal Aliens, Local Citizens: The Historical, Constitutional and Theoretical Meanings of Alien Suffrage, 141 U. PA. L. REV. 1391, 1460-67 (1993). 155. See, e.g., May v. Town of Mountain Vill., 132 F.2d 576, 582 (10th Cir. 1997). 156. See Briffault, supra note 112, at 398-99. 157. Plaintiffs in Siegel v. LePore and Touchston v. McDermott who challenged the manual recounts ordered by the local canvassing boards in Broward, Miami-Dade, Palm Beach, and Volusia Counties were voters in other parts of the state who claimed that the inclusion of manually recounted ballots would “dilute” their votes. Those claims were consistently rejected by the federal courts that heard them. See Touchston v. McDermott, 120 F. Supp. 2d 1055 (M.D. Fla 2000), aff’d, 234 F.3d 1133 (11th Cir. 2000) (en banc); Siegel v. LePore, 120 F. Supp. 2d 1041 (S.D. Fla. 2000), aff’d, 234 F.3d 1163 (11th Cir. 2000) (en banc). Judge Middlebrooks, the trial judge in Siegel, was particularly critical of the vote dilution theory, noting that a recount “strives to strengthen rather than dilute the right to vote by securing, as near as humanly possible, an accurate and true reflection of the will of the electorate.” Siegel, 120 F. Supp. 2d at 1050. The Supreme Court denied the petitions for writ of certiorari in both cases. See Touchston v. McDermott, 531 U.S. 1061 (2001); Siegel v. LePore, 531 U.S. 1005 (2000). 2001] EQUAL PROTECTION 361 have treated all dented, dimpled, or pierced chads as no-votes. It is difficult to see the constitutional harm to those voters whose ballots may constitutionally be treated as no-votes and might still be treated as no-votes under a consistent statewide standard if similar ballots are tabulated in other parts of the state. In the absence of the adoption of a standard intentionally discriminating against a group or groups of voters, a variation in the standards for counting undervote ballots does not create an equal protection problem. Indeed, in other cases the Supreme Court has suggested that when a state’s voting rules go beyond what the Constitution requires in vindicating the right to vote, the distinctions the state draws in providing more generous rules are subject to rational basis review rather than the strict scrutiny ordinarily applied to discriminations affecting the vote. This is true even if the result is that some voters are benefited while others are not. Thus, in McDonald v. Board of Election Commissioners,158 the Court applied only rational basis review and upheld an Illinois law that made absentee ballots available to voters who could not go to the polls for medical reasons as well as to voters who were away from their home county, but not to voters unable to go to the polls because they were jailed in their home counties. Noting that Illinois had no obligation to provide absentee ballots at all, the Court treated the absentee ballot law as an instance of the legislature’s traditional authority to take reform “one step at a time, addressing itself to the phase of the problem which seems most acute to the legislative mind.” . . . [A] legislature need not run the risk of losing an entire remedial scheme simply because it failed, through inadvertence or otherwise, to cover every evil that might conceivably have been attacked.159 The Court then sustained the limited availability of absentee ballots.160 In short, in the absence of a showing that the variations in standards for evaluating undervote ballots were intended to benefit the voters in some counties or burden those in others—or that the inconsistencies were intended to benefit or burden other identifiable 158. 394 U.S. 802 (1969). 159. Id. at 809 (citation omitted). 160. Id. at 809-11. McDonald’s specific ruling—upholding the constitutionality of the denial of absentee ballots to pretrial detainees incarcerated in their home counties—was overruled five years later in O’Brien v. Skinner, 414 U.S. 524 (1974). McDonald had focused on the fact that the detainees had failed to demonstrate that, without absentee ballots, they were unable to vote. In O’Brien, the plaintiffs made such a showing and the Court found that their inability to get to the polls was due entirely to their detention by the state. Thus, denial of absentee ballots was tantamount to denial of the vote. Of course, in Gore v. Harris, there was no claim that the state had prevented the voters who cast imperfectly marked ballots from casting proper ballots. 362 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 groups of voters or particular candidates—variations in the standards for reviewing undervote ballots would not create an equal protection claim. Of course, a finding of geographic discrimination was central to Bush v. Gore’s holding. Yet, as I will indicate in the next section, any geographical variations in counting standards that might have occurred were not unconstitutional discriminations. D. Geographic Discrimination and the Undervote The Bush v. Gore per curiam sought to tie its reversal of the Florida court’s manual recount order to an important theme in the Supreme Court’s voting rights jurisprudence—the prohibition of geographic discrimination. Indeed, the revolution in voting rights jurisprudence of the 1960s began with the problem of state laws favoring voters in some areas over voters in other areas, rather than with the right to vote per se. As the per curiam explained: An early case in our one-person, one-vote jurisprudence arose when a State accorded arbitrary and disparate treatment to voters in its different counties. Gray v. Sanders, 372 U.S. 368 (1963). The Court found a constitutional violation. We relied on these principles in the context of the Presidential selection process in Moore v. Ogilvie, 394 U.S. 814 (1969), where we invalidated a county-based procedure that diluted the influence of citizens in larger counties in the nominating process.161 Gray v. Sanders162 involved Georgia’s use of the county unit rule in primary elections that selected the Democratic nominees for statewide office. Under the unit rule, votes were counted within counties; the candidate who got a plurality of votes within that county won the “county unit” votes allocated to that county. The county unit votes were then tabulated to determine the primary winner. The allocation of the county unit votes overrepresented the least populous counties. The Supreme Court held that such a weighting of the primary vote to favor the least populated counties—in practice, the rural counties—was unconstitutional.163 Moreover, the Court indicated that even if the malapportionment had been cured and each county had received its fair share of the unit votes, the winner-takeall aspect of unit voting, which in effect disregarded the votes of all but the winning candidate in each county, would have rendered the system unconstitutional.164 161. 162. 163. 164. Bush v. Gore, 531 U.S. 98, 107 (2000). 372 U.S. 368 (1962). Id. at 379. Id. at 381 n.12. 2001] EQUAL PROTECTION 363 Moore v. Ogilvie165 also involved a state law that formally favored less populous counties. Illinois law provided that in order to be placed on the ballot, independent candidates for President had to secure, inter alia, petition signatures from 200 voters from each of fifty counties. Given the enormous variation in county populations in the state, “[t]his law . . . discriminates against the residents of the populous counties of the State in favor of rural sections.”166 The intercounty variations in the Florida Supreme Court’s manual recount order differed from the geographic discriminations in Gray and Moore in three important ways. First, unlike the county unit rule and the county signature requirements, the intercounty differences in assessing imperfectly marked ballots were not commanded by the state. They were the results of local decisionmaking. The state supreme court had not required a more inclusive rule in some counties and a more restrictive rule in others. If voters in some counties were disfavored relative to voters in other counties, that was the result of the actions of decisionmakers in the voters’ own counties, not the state supreme court. This is very significant. Whereas in Gray and Moore the voters in the disfavored counties would have had to persuade a legislature composed of representatives of both favored and disfavored counties to change the rules—which was not likely to succeed and was likely, instead, to result in the perpetuation of the discrimination against populous areas—in Florida any county utilizing a more restrictive undervote assessment rule could, on its own, change that rule to a more inclusive one. Second, the variations in Florida were unpatterned whereas in Gray and Moore the state laws favored a specific interest group, rural voters, over another group, the residents of populous urban areas. In Gore v. Harris, not only did the state supreme court not determine which county must use a restrictive rule and which must use an inclusive rule, but there was also no claim or evidence that the differences in ballot assessment standards mapped on to any “independently identifiable group or category . . . .”167 As a result, the variation in counting standards would have been comparable to the West Virginia constitutional provision requiring bond issue ballot propositions to receive 60 percent of the vote in order to pass. Although that provision gave greater weight to the votes of voters in a 40.1 percent to 49.9 percent minority than to voters in a 50.1 percent to 59.9 percent majority, the Supreme Court in Gordon v. Lance upheld the supermajority voting rule, finding that “[u]nlike the restrictions in our 165. 394 U.S. 814 (1969). 166. Id. at 819. 167. Gordon v. Lance, 403 U.S. 1, 5 (1971). 364 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 previous cases, the West Virginia Constitution singles out no ‘discrete and insular minority’ for special treatment.”168 Indeed, the variations in Florida’s manual recount standards may have actually been less troubling than the variations in undervote rates attributable to the use of different kinds of election machinery. To the extent that Florida’s counties chose between punch card and optical scan technologies based on cost, and to the extent that those choices correlated with each county’s per capita tax base, then the differences in the ability to cast an effective vote would begin to correlate with local wealth. Although not exactly a poll tax, such a relationship between wealth and voting ability, if proven, would surely raise a constitutional question. By contrast, there was no apparent connection between local wealth and the choice of manual recount standard. Finally, unlike the county unit rule in Gray, the Florida Supreme Court’s order would not have resulted in the disregarding of any counted ballots. In Gray, votes cast for a candidate who failed to carry the voters’ county were effectively ignored in the final decision (much as the votes for a presidential candidate who fails to carry a state are effectively ignored in the selection of presidential electors). Under the Florida court’s order, all ballots that had been counted would have been tabulated in the final total. To be sure, the order in Gore v. Harris would have resulted in the differential treatment of comparable undervote ballots depending on where they had been cast. But unlike the two cases cited by the per curiam, Gray and Moore, there was no state-mandated discrimination, and indeed, no discrimination with respect to a constitutionally protected vote. Nor did the recount order expressly or impliedly favor one candidate or one set of voters over another candidate or set of voters. The lack of any partisan bias in the recount order can be favorably contrasted with the county-level recounts undertaken during the protest phase of the Florida election battle and the Florida court’s order to include the results of the Miami-Dade and Palm Beach county recount results in the candidates’ vote tallies. The county-level recounts potentially injected political bias in the Florida statewide tally. Those recounts resulted from the decisions of the Gore campaign to target counties which had given Gore majorities. That action was not unreasonable. Under Florida law, protests are filed and decisions whether or not to undertake a recount are made at the county level.169 In the absence of any legislative provision for a statewide protest, it would have been logistically difficult for a candidate to 168. Id. 169. FLA. STAT. § 102.166 (2000) (amended 2001). 2001] EQUAL PROTECTION 365 seek a statewide precertification recount. Gore would have had to file in sixty-seven separate counties, and even if he had done so there was no guarantee that a statewide recount would have resulted because each county had the discretion to decline to undertake a recount. Indeed, if the preliminary sampling recount had failed to detect an appreciable discrepancy between machine and manually counted ballots in a particular county,170 that county canvassing board probably could not have ordered a recount even if there were some undervote ballots in the county that might generate legally valid votes. Thus, Gore reasonably targeted his recount efforts on the counties that he thought would generate the most additional votes for him. But this meant that the protest period recount was tilted in favor of the Democratic candidate and Democratic voters. To be sure, the fact that Gore asked for a recount did not mean that a county canvassing board was required to grant his request. Indeed, Miami-Dade County, after initially granting the request, rescinded its vote and terminated its recount.171 But the impetus for the recounts came from the Democrats, thus skewing the recount’s focus in the Democrats’ direction. A selective manual recount focused on particular counties or territorial subdivisions of the state is not necessarily unconstitutional. A selective recount could be justified by a combination of factors, including time constraints precluding a complete statewide recount and the utilitarian benefits of focusing the limited recount opportunity on those areas which, due to large populations and high percentages of undervotes linked to the use of particular voting machinery, potentially provide the most undervotes. But in the 2000 Florida election’s protest phase, the selective recounts were driven in large measure by partisan strategies and certainly had predictable partisan consequences.172 Indeed, one of the appealing features of the 170. Id. § 102.166(4) (amended 2001). 171. 36 DAYS, supra note 13, at 133-35. The Florida Supreme Court unanimously rejected Gore’s request that Miami-Dade be compelled to resume the recount. Gore v. MiamiDade County Canvassing Bd., 780 So. 2d 913 (Fla. 2000); see also 36 DAYS, supra note 13, at 142. 172. The Bush campaign advanced this argument in its unsuccessful effort to obtain a preliminary injunction against county level recounts. The Eleventh Circuit’s per curiam opinion did not expressly consider the argument that the selective county recounts were unconstitutional. Rather, the court found that Bush had failed to establish one of the requisites for a preliminary injunction—substantial likelihood of irreparable injury. The court reasoned that, at the time of the court’s decision, Bush was “suffering no serious harm, let alone irreparable harm,” since he had been certified the winner of Florida’s electoral votes notwithstanding the inclusion of some manually recounted ballots. Siegel v. LePore, 234 F.3d 1163, 1177 (11th Cir. 2000) (en banc). Chief Judge Anderson, concurring specially, reached the merits of the question of whether the recounts were unconstitutional. He determined that due to the facts that both candidates had the opportunity to seek recounts, the county canvassing boards had complete discretion to reject a recount request, and the recounts themselves were “untainted by partisan manipulation,” the selective manual re- 366 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 Florida Supreme Court’s decision in Gore v. Harris is its correction of the partisan skew that resulted from the selective recounts undertaken in the protest phase. The role of partisan strategy in the selection of ballots for a partial manual recount explains why the portion of the Florida Supreme Court’s order directing the inclusion of the Palm Beach and partial Miami-Dade recount results in the candidates’ vote totals violated equal protection principles, as Bush v. Gore held. According to the United States Supreme Court, the Palm Beach and Miami-Dade recounts were not limited to the review of undervote ballots but also included valid votes found through the review of overvotes—that is, ballots treated as void because they reflected votes for more than two candidates. Overvotes that turned out to be valid votes involved ballots in which a voter punched out a candidate’s chad and also wrote in the same candidate’s name. Although such ballots would be rejected by a machine as votes for two different candidates, they plainly indicated the intent of the voter to vote for just one candidate. It would be constitutional to focus a manual recount on undervotes and exclude overvotes—as the Florida Supreme Court did. As strict scrutiny ought not apply to distinctions in no-vote ballots which constitutionally could be completely ignored, the decision of a court or election administration panel to focus a recount on undervotes and to exclude overvotes should be subject to the rational basis test, not strict scrutiny. Given the data indicating considerable positive correlation between a county’s use of punch card machinery and its undervote rate173—and the lack of a similar correlation with respect to overvotes—it would be reasonable to infer that, at least in punch card counties, some significant fraction of undervotes consisted of failed efforts to register a preference. On the contrary, overvotes were more likely to involve either voter mistakes or efforts to register two preferences. As a result, it would be reasonable to conclude that undervote ballots are more likely than overvote ballots to include legal votes.174 With the presidential recount subject to severe time constraints, a state could reasonably seek to reconcile its legiti- counts were not unconstitutional. Id. at 1182-86 (Anderson, C.J., concurring specially). Three members of the court dissented, finding, inter alia, that the selective recounts amounted to partisan discrimination. Id. at 1202-09. (Birch, J., dissenting, joined by Tjoflat and Dubina, JJ.) 173. 36 DAYS, supra note 13, at 191. 174. The discovery that some overvote ballots consisted of punched-out and written-in votes for the same candidate suggests that there may have been more valid votes on overvote ballots than was generally assumed. Nevertheless, given the differences between undervote and overvote ballots and the problems with the punch card machinery associated with some of the undervote ballots, it would have been reasonable, even if mistaken, to assume that undervote ballots would be a greater source of valid votes—and, thus, the principal target for a time-constrained manual recount—than overvote ballots. 2001] EQUAL PROTECTION 367 mate goals of counting as many valid ballots as possible and meeting the statutory deadlines for participating in the presidential selection process by focusing on the undervote. So, too, it would not necessarily be unconstitutional to combine a statewide recount of the undervote with a partial recount of the overvote. Again, the combination of tight time constraints, the effort to maximize the vote count, and the use of nondiscriminatory criteria for selecting those counties in which to undertake the overvote partial recount would mean that a recount consisting of a statewide count of the undervote and a partial count of the overvote could be reasonable. But if the counties in which the manual recount of the overvote took place were selected to advance partisan interests, then the fairness of the resulting vote count would be tainted. That is apparently what happened in Florida. Palm Beach and Miami-Dade Counties were selected for recounts by the Democrats because they were carried by the Democratic candidate. If, as a result of the Florida Supreme Court’s order, votes were obtained from the recount of overvotes in those two counties but from no other counties in the state, that would have biased the result in favor of Gore. Thus, the United States Supreme Court was right to determine that the Florida court’s inclusion of votes obtained from the recount of the overvote ballots in those counties was unconstitutional.175 The selective protest phase recounts and the inclusion of the overvote recount results from the counties selected by the Democrats thus present serious issues of partisan bias. But the Florida Supreme Court’s order directing a statewide manual recount of undervote ballots did not have a similar partisan tilt. There was no reason to assume that the lack of consistent statewide standards was intended to favor one candidate and that candidate’s voters over the other candidate and his voters, or that it would have had the effect of doing so. There was no evidence that geographic variation was a disguised form of partisan manipulation.176 175. Presumably, had Gore prevailed on the recount, Bush could have raised a similar argument with respect to the inclusion in the statewide result certified by Secretary Harris of votes obtained from recounting the overvote in heavily Democratic Broward County. 176. The Bush v. Gore per curiam also expressed concern that the Florida Supreme Court’s certification of a partial total from Miami-Dade gave “no assurance that the recounts included in a final certification must be complete.” 531 U.S. 98, 108 (2000). The implication is that it would be unconstitutional to certify a partial recount result. That seems mistaken. Given that the Court was willing to permit the certification of a partial tally— Secretary Harris’ certification included valid votes found by manual recounts undertaken in Broward and Volusia Counties but ignored any valid votes that a manual recount might have found in Florida’s other counties—a partial count, particularly a partial count that picks up some but not all of the votes missed by the vote-counting machinery, does not deny equal protection. The voters whose ballots are not counted in the recount are not made worse off by the fact that a partial recount picked up valid votes cast by other voters. A constitutional objection would arise only if the partial count is skewed to benefit particu- 368 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 E. Fundamental Fairness and the Recount Order Apart from the inclusion of the Palm Beach and Miami-Dade recount results, the Florida recount order did not create a problem of fundamental unfairness. The recount order did not violate the right to vote or the right to an equally weighted vote. To the limited extent that the variations in counting standards created equal protection problems, those variations were justified by the Florida court’s effort to maximize the ability of Florida voters to have their ballots counted while avoiding the constraints on judicial innovation in the presidential election context imposed by the United States Supreme Court in Bush v. Palm Beach County Canvassing Board.177 The hallmarks of electoral unfairness have been the disenfranchisement of voters and intentional discrimination against voters. The leading lower court cases involving the invalidation of election practices on federal constitutional grounds have involved state or local rescission of rules on which voters had reasonably relied in deciding how to vote. In Griffin v. Burns,178 for example, state election officials told Rhode Island voters that they could vote by absentee or shut-in ballots in party primaries. After an election in which nearly ten percent of the total vote recorded was cast by absentee or shut-in ballots, the Rhode Island Supreme Court held that the use of absentee or shut-in ballots in primaries was unauthorized by state law and invalidated those ballots. The First Circuit held that the state court’s action “severely impugned” the integrity of the election and amounted to “patent and fundamental unfairness” since the voters had reasonably relied on well-established practice, and the state court’s order thus unfairly disenfranchised them.179 In Roe v. State of Alabama,180 a case heavily relied on by Bush in his effort to block the Florida recounts, the Eleventh Circuit found that an Alabama court’s postelection order, which departed from past practice and required election officials to count absentee ballots that did not include notarization and the signatures of two qualified wit- lar voters. Miami-Dade had been selected for a recount because of its Democratic majority, and the precincts in Miami-Dade that had already been counted might have been selected for partisan reasons. As a result, the Florida court’s inclusion of the partial Miami-Dade results before the general statewide recount was undertaken was an error. But it should have been constitutionally permissible to include partial recount results when a full recount is impossible due to the federal timetable for the selection of presidential electors, provided that the counties and precincts that were counted were not selected on a partisan basis. 177. 531 U.S. 70 (2000). 178. 570 F.2d 1065 (1st Cir. 1978). 179. Id. at 1078; see also Partido Nuevo Progresista v. Barreto Perez, 639 F.2d 825, 827 (1st Cir. 1980) (finding that disenfranchisement was crucial to Griffin’s holding). 180. 43 F.3d 574 (11th Cir. 1995). 2001] EQUAL PROTECTION 369 nesses required by law, “implicate[d] fundamental fairness.”181 The Eleventh Circuit rejected the argument that the state court’s acceptance of otherwise invalid absentee ballots was constitutional because it enfranchised those who had cast the contested absentee ballots. Instead the federal court emphasized the disenfranchising effect of the Alabama court’s order: [T]he change in the rules after the election would have the effect of disenfranchising those who would have voted but for the inconvenience imposed by the notarization/witness requirement. . . . We believe that, had the candidates and citizens of Alabama known that something less than the signature of two witnesses or a notary attesting to the signature of absentee voters would suffice, campaign strategies would have taken this into account and supporters [of the candidates disadvantaged by the rule change] who did not vote would have voted absentee.182 Unlike the Rhode Island court’s order in Griffin and the Alabama court’s order in Roe, the Florida court’s order in Gore v. Harris was enfranchising. It would have provided for the review of tens of thousands of uncounted undervote ballots and potentially would have led to the counting of whatever valid votes were found among those ballots. The Gore order would not have unfairly burdened other voters or potential voters. Both the availability of the manual recount and the use of the “intent of the voter” standard were consistent with preexisting law. Even if the Florida order could somehow be characterized as a change in the law, it is hard to see how such a change disenfranchised anyone. Unlike the situation in Roe, it is simply impossible to believe that some Floridians assumed they would be unable to fully detach their chad and did not bother to vote but that they would have voted if they had known that a manual count, using a liberal standard, would have been required for all undervotes, and thus they were disenfranchised by the Florida order. The Florida case is much closer to Partido Nuevo Progresista v. Barreto Perez183 than it is to Roe. In Partido Nuevo Progresista, the First Circuit reversed the district court and rejected federal intervention in a Puerto Rico election dispute where the election administrator counted as votes ballots with marks above the list of a party’s candidates rather than in the places on the ballot designated for indicating preferences. The court emphasized that the administrator’s order was enfranchising and that “no party or person is likely to have 181. Id. at 581. 182. Id. at 581-82. 183. 639 F.2d 825 (1st Cir. 1980). 370 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 acted to their detriment by relying upon the invalidity of ballots with marks outside the ballots’ drawn rectangles.”184 The Florida court’s order did not intentionally discriminate among voters. As already noted, the Florida court did not itself require the use of a more inclusionary standard for determining legal votes in some counties and a less inclusionary one in others. Any differences in counting standards would have been the result of local action, not state mandate. If anything, the Florida court’s order would have ameliorated the intercounty differences in effective voting rates attributable to the differences in the quality of voting machinery. Voters in counties using punch card ballots had an appreciably lower chance of casting effective votes than voters in counties using optical scan machinery.185 The recount order would have reduced the disparity in effective voting rates between punch card and optical scan counties, although, to be sure, it could have led to the creation of smaller disparities within the punch card counties if the counties in fact used different counting standards. Even then, by directing that the recount be undertaken under the auspices of a circuit court judge who had the authority to hear challenges to counting team decisions and resolve disputed cases, the Florida court’s order provided an opportunity for reducing intercounty differences. Oddly, the Bush v. Gore per curiam cabined the application of equal protection in the election administration context to “the special instance of a statewide recount under the authority of a single state judicial officer.”186 Yet, by placing the recount under the authority of a single state judicial officer, the Florida Supreme Court actually created the possibility of reconciling intercounty differences and harmonizing standards. Certainly, the order created a greater possibility for statewide consistency in ballot evaluating standards than was the case with the protest phase recounts, when no state officer reconciled intercounty variations in counting standards. Even if intercounty variations in the determination of whether a ballot contained a valid vote had resulted, and even if those variations could be attributed to the state supreme court’s failure to spell out ballot assessment standards, the Florida court’s action did not threaten the integrity of the election under the circumstances of the Florida presidential recount dispute. Rather, the Florida court’s fail184. Id. at 828; see also Roe v. Alabama, 43 F.3d at 581-82 (citing Partido Nuevo Progresista with approval). 185. 36 DAYS, supra note 13, at 191. In the counties that used optical scan ballots, just 0.30 percent of ballots had no selection for President whereas in the counties that used Votomatic punch card ballots, 1.53 percent of the votes had no selection for President. Indeed, in every single one of those counties, the undervote fraction was greater than one percent, and in five counties (including Palm Beach county) the undervote fraction was greater than two percent. Id. 186. Bush v. Gore, 531 U.S. 98, 109 (2000). 2001] EQUAL PROTECTION 371 ure to spell out more precise rules for the evaluation of undervote ballots may have been required by the constitutional and statutory rules governing presidential elections. Just four days before Gore v. Harris was decided, the United States Supreme Court had sharply reminded the Florida court of the Florida Legislature’s exclusive constitutional prerogative to determine the rules for selecting presidential electors. The Supreme Court had also hinted broadly that the Florida Legislature would probably have wanted to secure the safe harbor provided by 3 U.S.C. § 5 by avoiding the creation of any rules for the presidential election that could be characterized as postelection day new law. The Florida Legislature had never adopted a standard for evaluating ballots more specific than “the intent of the voter.” Similarly, Florida case law concerning the determination of legal votes prior to November 2000 had been phrased exclusively in terms of the “intent of the voter.” Given Bush v. Palm Beach County Canvassing Board, if the Florida Supreme Court had adopted a more specific standard, such action could have been challenged as a violation of Article II, Section 1 and a threat to the federal statutory safe harbor. In effect, Bush v. Palm Beach County Canvassing Board chastised the Florida Supreme Court for doing what Bush v. Gore subsequently condemned the Florida court for not doing—making new law with respect to the resolution of a dispute in the selection of presidential electors. Under these circumstances, the Florida court’s order did not threaten the integrity of the presidential election. After Bush v. Palm Beach County Canvassing Board, judicial deference to the legislature and avoidance of making new law for the resolution of election disputes are part of the very definition of electoral fairness and integrity in the presidential election setting. If it had spelled out a standard more specific than the “intent of the voter” for the evaluation of undervote ballots, the Florida Supreme Court would surely have been creating new law. The Florida court’s order may have permitted intercounty and intracounty inconsistencies in the evaluation of undervote ballots. But the court’s failure to adopt a standard that would have precluded those inconsistencies was apparently required by the federal policy of avoiding the state judicial creation of new law in the context of a presidential election. In effect, the Florida Supreme Court had two choices. It could have declined to order a recount, thus excluding whatever valid votes might have been found in the undervote ballots and leaving unremedied the effects of the different election machinery in creating different undervote rates in different Florida counties. Or it could have done what it did—order a statewide recount of the undervote, recognizing that the lack of a precise statewide standard for evaluating undervote ballots might lead to certain ballots being counted in some 372 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 counties while comparable ballots were ignored in other counties. The third option implicated by Bush v. Gore—the inclusion of a ballot assessment standard in the recount order—was effectively precluded by Bush v. Palm Beach County Canvassing Board.187 Both of the options available to the Florida court were flawed. Each would have entailed some disparities in the treatment of some Florida voters. But the second option—the manual recount—at least had the benefits of increasing the number of Floridians whose votes were counted and of reducing the disparities in voting rates attributable to differences in the quality of county voting machinery. Moreover, whatever disparity of treatment might have occurred due to the uncorrected differences in vote assessment standards, it would have been the result not of an intent to discriminate among counties or among partisan interests but of a constitutional constraint on the Florida court’s authority to take the action necessary to avoid that disparity. Each option posed issues of fairness but, given the circumstances, neither option would have caused a fundamental unfairness or threatened the integrity of the election. Certainly it is hard to see how the option the Florida court did choose—the manual recount without specific standards—was more unfair than no recount, and therefore the failure to review tens of thousands of undervote ballots at all. The Florida court’s recount order may have been debatable as a matter of Florida election law—an issue beyond the scope of this Article—but it did not create the kind of fundamental unfairness that ought to be necessary to support a finding that it violated the Equal Protection Clause. Indeed, in seeking to expand the number of voters counted while reducing the differences in effective voting rights among counties, the Florida court’s order was far more consistent with the United States Supreme Court’s use of the Equal Protection Clause in its modern voting rights cases than was the Supreme Court majority in Bush v. Gore. V. CONCLUSION: FEDERALISM AND EQUAL PROTECTION IN BUSH V. GORE Many observers were struck by the obvious fact that the five Justices who led the Court’s unprecedented intervention into a state’s vote counting process, and who injected federal equal protection concerns into an area hitherto seen as a matter largely committed to the 187. A fourth option, of course, was to grant Gore’s request and order a manual recount limited to the uncounted Miami-Dade undervote ballots. For the reason given in the text— that Miami-Dade was selected for the protest phase recount because of its Democratic majority and thus a recount limited to Miami-Dade would have been unfairly proDemocratic—this option would have created a fundamental unfairness. 2001] EQUAL PROTECTION 373 states,188 were also the five Justices who have mounted the Court’s recent aggressive defense of states rights against federal power. These are the Justices who read the “anti-commandeering” doctrine into the Tenth Amendment,189 reinvigorated the Eleventh Amendment,190 and imposed new limits on Congress’s power to act under the Commerce Clause191 and under Section 5 of the Fourteenth Amendment.192 The tension between Bush v. Gore and federalism is not simply a matter of the Court’s application of equal protection principles to state election administration. Federalism has always included federal judicial protection of federal constitutional rights. As I have suggested earlier, equal protection principles have been applied to voting in state and local elections, and it is no great stretch of the Constitution’s vindication of the right to vote and the right to an equally weighted vote to subject state voting mechanisms that operate to disenfranchise voters or to discriminate among voters to equal protection principles. The evidence from Florida concerning the disparate voter error rates resulting from different types of voting machines and different ballot designs suggests that equal protection could play a legitimate and useful role in curbing the structural inequalities in voting that currently plague our system. The real tension between Bush v. Gore and federalism is that the gravamen of the particular equal protection violation at the heart of Bush v. Gore is the value at the heart of federalism itself— decentralized decisionmaking and the resulting variations in government action. Bush v. Gore’s greatest concern was with the poten188. The lower federal courts have repeatedly refused to get involved in disputes involving “garden variety” irregularities of state and local election administration, even when such irregularities have resulted in the rejection of some valid votes or undermined the validity of an election. See, e.g., Gold v. Feinberg, 101 F.3d 796, 801 (2d Cir. 1996); Curry v. Baker, 802 F.2d 1302, 1316 (11th Cir. 1986); Bodine v. Elkhart County Election Bd., 788 F.2d 1270 (7th Cir. 1986); Partido Nuevo Progresista, 639 F.2d 825; Gamza v. Aguirre, 619 F.2d 449, 454 (5th Cir. 1980). Rare exceptions are Roe v. Alabama, 43 F.3d 574 (11th Cir. 1995) (finding that a change in law concerning absentee ballots unfairly surprised those who had relied on restrictive law by not voting absentee); Duncan v. Poythress, 657 F.2d 691 (5th Cir. Unit B 1981) (failing to hold an election altogether was held to be a denial of voting rights); and Griffin v. Burns, 570 F.2d 1065 (1st Cir.1978) (holding that a postelection ruling by a state court that absentee ballots should not have been allowed and thus that nearly ten percent of the ballots cast should not be counted; such a surprise massive disfranchisement after the election was unconstitutional). 189. See, e.g., Printz v. United States, 521 U.S. 898 (1997). 190. See, e.g., Board of Tr. of Univ. of Ala. v. Garrett, 531 U.S. 356 (2001); Kimel v. Fla. Bd. of Regents, 528 U.S. 62 (2000); Alden v. Maine, 527 U.S. 706 (2000); Coll. Savs. Bank v. Fla. Prepaid Postsecondary Educ. Expense Bd., 527 U.S. 666 (2000); Fla. Prepaid Postsecondary Educ. Expense Bd. v. Coll. Savs. Bank, 527 U.S. 627 (2000); Seminole Tribe of Fla. v. Florida, 517 U.S. 44 (1996). 191. See, e.g., United States v. Morrison, 529 U.S. 598 (2000); United States v. Lopez, 514 U.S. 549 (1995). 192. See, e.g., Morrison, 529 U.S. 598; City of Boerne v. Flores, 521 U.S. 507 (1997). 374 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 tial for varying local standards when determining whether an undervote ballot contained a valid vote. These variations occurred because state institutions—the legislature, the Secretary of State as chief administrative officer of the elections system, and the state supreme court—had not adopted rules guiding the actions of county canvassing boards and their counting teams. Although there is no evidence that Florida intentionally chose to devolve this difficult question to the county canvassing boards, that would not have been an unreasonable thing to do. Given the absence of one right answer for balancing the competing factors of voter inclusion, objectivity, and ease of administration, the state could have chosen to let the counties decide this, with different counties balancing these factors differently. Those that valued inclusion could have adopted a more liberal rule, while those that believed that the voters should be required to make a greater effort to confirm that their chads had detached or those more concerned about the objectivity of local election administrators could have adopted a more restrictive rule. A longstanding principle of federalism has been that state-local relationships and the nature and scope of a state’s delegation of power to its local units is, as a matter of federalism, largely for the states.193 To be sure, state delegations or, more commonly, state modifications of traditionally delegated powers that target particular groups or burden fundamental rights, are subject to federal constitutional review.194 But even when decentralization imposes some constraints on the ability to vindicate fundamental rights or protect important interests, the Court has generally treated the state-local relationship as primarily a state matter while expressing support for the state’s decision to favor local autonomy over other concerns.195 So long as any of the standards open to the local canvassing boards— hanging chad, pierced chad, or indented chad—is constitutional, then, as a matter of federalism, the states should be free to leave the matter to the local units. Bush v. Gore does not expressly preclude such decentralization. By limiting its opinion to “the special instance of a statewide recount under the authority of a single state judicial officer,” the Court does not even reach recounts that might be con193. Hunter v. City of Pittsburgh, 207 U.S. 161 (1907). 194. See, e.g., Romer v. Evans, 517 U.S. 620 (1996); Bd. of Educ. v. Kiryas Joel Vill. Sch. Dist., 512 U.S. 687 (1994); Washington v. Seattle Sch. Dist. No. 1, 458 U.S. 457 (1983); Gomillion v. Lightfoot, 364 U.S. 339 (1960). 195. See Milliken v. Bradley, 418 U.S. 717 (1974) (holding that the protection of a system of decentralized education administration justifies reversal of lower court order requiring inclusion of suburban school districts in a metropolitan area desegregation plan); San Antonio Indep. Sch. Dist. v. Rodriguez, 411 U.S. 1, 49-50 (1973) (finding that decentralization of education administration justifies the interdistrict inequalities in revenues and expenditures resulting from differences in local wealth and reliance on local wealth in funding schools). 2001] EQUAL PROTECTION 375 ducted under the auspices of a state administrative body.196 But the implication is clearly that the decentralization of the determination of counting standards is unconstitutional. Bush v. Gore does not simply challenge the state’s power to decentralize some aspects of election law decisionmaking. It raises questions about the variations that result from such decentralization. The central principle of federalism is diversity. Federalism necessarily results in differences—differences that grow from variations in needs, circumstances, preferences, and decisionmaking processes. These differences reflect the local political participation that federalism promotes and the local innovation that federalism protects. Yet, where federalism ordinarily celebrates diversity, local participation, and local experimentation, the Bush v. Gore per curiam saw arbitrary and disparate treatment in different counties. Certainly federalism produces “disparate” rules for different places. But the theory of federalism suggests that such differences, although perhaps arbitrary in the sense that there is no justification that links a particular local rule to specific local circumstances, are also an inevitable outgrowth of decentralized decisionmaking. Different groups of people look at the same problem differently and reach different conclusions about that problem. As a result, they adopt different responses to the same problems. There may be no objective explanation for the different preferences and different results, but they exist nonetheless. Federalism means those differences may be translated into different legal rules in different places. The tension between the per curiam’s federalism jurisprudence and its Bush v. Gore opinion is thus not that the Court has invaded an area that has hitherto been a province of the states. Rather, it is the Court’s apparent discomfort with the varying local standards that are the inevitable accompaniment of decentralized decisionmaking—the very decentralized decisionmaking that is the heart and soul of federalism. Decentralization and the resulting variations in local standards and practices are not always desirable. Election administration could very well benefit from state legislative decisions that provide for voting machinery that is of uniform quality statewide, that standardize ballot design, or that specify consistent statewide procedures for resolving questions concerning improperly marked ballots. But where, as in Bush v. Gore, the intrastate variations do not exclude otherwise valid ballots, involve state-ordered geographic discrimination, upset voters’ expectations, or otherwise undermine fundamental fairness, a commitment to the spirit of federalism would appear to counsel in fa- 196. Bush v. Gore, 531 U.S. 98, 109 (2000). 376 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:325 vor of accepting local decisionmaking rather than undertaking federal judicial intervention. Certainly, Bush v. Gore was far from compelled by the Court’s voting rights cases. Decentralized decisionmaking has most typically fallen to a constitutional challenge when some state or local decisionmakers fail to abide by federally required standards, or when decentralization is a guise for discrimination against locally vulnerable groups. In Bush v. Gore, the intercounty and intracounty variations were just that—variations. None of them fell below federal standards concerning the undervote (since no such standards exist). And there was no evidence that the variation in standards would have discriminated against geographic interests or partisan groups. The variations in counting standards were troublesome to the Justices and, ultimately, held unconstitutional, simply because they were variations. Bush v. Gore is an unusual equal protection case. The Florida Supreme Court’s order did not exclude voters; rather, it would have expanded the ability of Florida voters to cast effective votes. Nor did the Florida order discriminate against any class of voters or ratify the discriminatory actions of other institutions. But what is perhaps more striking is the U.S. Supreme Court majority’s failure even to consider the federalism implications of the case. The Court did not appear to recognize that it was federalizing a state-local relationship. Nor did it even attempt to reconcile its usually strong commitment to federalism with its apparent discomfort with decentralization and, especially, its hostility to the variation in local standards that inevitably follows from decentralized governance. BUSH V. GORE AND THE FUTURE OF EQUAL PROTECTION LAW IN ELECTIONS RICHARD L. HASEN* I. INTRODUCTION ..................................................................................................... II. WHY WE SHOULD NOT TAKE BUSH V. GORE’S EQUAL PROTECTION HOLDING SERIOUSLY ........................................................................................................... A. Optimism and the Equal Protection Holding ............................................. B. Reasons for Doubting Bush v. Gore’s Precedential Value .......................... 1. Limiting Language ................................................................................. 2. The Court’s Failure to Engage in Serious Analysis .............................. 3. Inconsistency in Equal Protection Analysis .......................................... III. TAKING BUSH V. GORE’S EQUAL PROTECTION HOLDING SERIOUSLY ................. IV. THE BENEFITS, COSTS, AND LIMITS OF THE NEW EQUAL PROTECTION JURISPRUDENCE OF BUSH V. GORE ..................................................................... A. Benefits.......................................................................................................... B. Costs .............................................................................................................. C. Beyond Third-Level Equality Claims .......................................................... 1. Equality of “Electoral Structures”.......................................................... 2. Equality of Campaign Finance .............................................................. V. CONCLUSION ........................................................................................................ 377 380 380 386 386 387 390 392 398 399 399 402 402 404 406 I. INTRODUCTION Before the recent Florida controversy, co-authors of the only two election law casebooks drew a distinction between the “big picture” issues of election law—such as representation, the nature of political equality, the role of money in politics—and the “nuts-and-bolts” of election law.1 The conventional wisdom was that the former was more important (and no doubt more interesting) to study than the latter.2 * Professor and William M. Rains Fellow, Loyola Law School, Los Angeles. B.A., University of California, Berkeley, 1986; M.A. 1988, J.D. 1991, Ph.D. (Political Science) 1992, University of California, Los Angeles. Thanks to Evan Caminker, Sam Issacharoff, Stephen J. Kay, Clark Kelso, Hal Krent, Dan Lowenstein, Bill Marshall, Richard McAdams, Andrew Sabel, Paul Schwartz, Georgene Vairo, and symposium participants, for useful comments and suggestions and to Caroline Djang for research assistance. The author served as an unpaid consultant to the Gore legal team on questions related to the Miami-Dade recount. 1. The two casebooks are DANIEL H. LOWENSTEIN & RICHARD L. HASEN, ELECTION LAW—CASES AND MATERIALS (2d ed. 2001); SAMUEL ISSACHAROFF ET AL., THE LAW OF DEMOCRACY (2d ed. 2001). 2. See Samuel Issacharoff & Richard H. Pildes, Not By “Election Law” Alone, 32 LOY. L.A. L. REV. 1173, 1173-74 (1999) (objecting to the term “election law” as focusing on “elections and their administrative mechanisms,” narrowing the field “to microscopic regulatory details” and running the risk of “signaling to potential newcomers a tedious focus on the narrow regulatory questions of most interest to political junkies . . . .”); Daniel H. Lowenstein, Election Law as a Subject—A Subjective Account, 32 LOY. L.A. L. REV. 1199, 1202 (1999) (although “[n]uts and bolts questions . . . have increased in number . . . [f]or the most part we do not teach these issues and we do not write about them in law reviews; not 377 378 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:377 The Florida controversy challenged that conventional wisdom. Although resolution of the dispute depended essentially upon two nutsand-bolts questions—how does one determine the intent of the voter from examining a paper ballot and what are the mechanics for contesting a statewide election?—the controversy illustrated in numerous ways that the line between big picture questions and nuts-andbolts questions is fuzzy. Indeed, this nuts-and-bolts dispute raised big picture questions regarding the nature of representation, the meaning of political equality, and the role of money in politics.3 It is no wonder that the new editions of both casebooks include material on the Florida dispute.4 The Supreme Court’s per curiam (unsigned) majority opinion in Bush v. Gore5 eviscerated the distinction between nuts-and-bolts questions and big picture questions by holding that Florida law, at least as construed by the Florida Supreme Court, violated the Equal Protection Clause of the Fourteenth Amendment. The Court held that a state violates equal protection when it fails to have uniform standards for the recounting of votes during a statewide election contest.6 The opinion is potentially far-reaching, translating just about any disparity regarding the means of voting into a justiciable question. Indeed, if Bush v. Gore were already on the books at the time the Palm Beach County “butterfly ballot” controversy arose, we could have asked how that controversy should have been handled based on equal protection grounds.7 Part II of this Article argues that although some have heralded the opinion as the (perhaps unintended) dawn of a new era in the jurisprudence of equal protection in elections, there are good reasons for doubting that the Supreme Court majority intended anyone to take their equal protection holding seriously. Language in the per curiam opinion limits it to the facts of the case, or, at most, to cases where jurisdiction-wide recounts are ordered. Moreover, the Court’s because they are not there but because, for various reasons, we do not find them sufficiently interesting”). 3. Questions of who votes, how votes are counted, and the reasons for and critiques of the Electoral College raised questions of representation and political equality. Toward the end of the controversy, the media focused on how the presidential candidates raised money for recount funds, thereby exploring the relationship of money and politics. On the latter issue, see John M. Broder, Contesting the Vote: Many Donors to Campaigns are Financing Recount Fight, N.Y. TIMES, Dec. 8, 2000, at A33. The Center for Responsive Politics website lists donors to the Gore recount committee and provides a searchable database of donors to the Bush recount committee. See http://www.opensecrets.org/alerts/v5/ alertv5_65b.asp (Gore); http://www.opensecrets.org/2000elect/other/bush/recountdonorsform.asp (Bush). 4. LOWENSTEIN & HASEN, supra note 1, chs. 3-4; ISSACHAROFF ET AL., supra note 1, chs. 4, 12. 5. 531 U.S. 98 (2000). 6. Id. at 109. 7. See infra Part III. 2001] EQUAL PROTECTION’S FUTURE IN ELECTIONS 379 own analysis was superficial. It failed to explain or justify its large extension of precedent, and, most importantly, given the fact that a “fundamental right” was involved, the Court appeared to speak the language of strict scrutiny but apply something much less than strict scrutiny. Finally, the kind of equal protection claim favored by the conservative Justices in the Bush v. Gore majority is a strong departure from the usual equal protection jurisprudence they favor. Time will tell whether the Court backs away from its ambitious new equal protection jurisprudence. To the extent that the Court does back away, it further undermines the already-questioned legitimacy of the opinion. Part III of this Article considers not whether the Court meant what it said, but rather what the consequences would be if the Court indeed meant what it said. The equal protection jurisprudence of Bush v. Gore moves election law to an uncharted third level of political equality. Various amendments to the Constitution and Supreme Court cases decided by the Warren Court established the first level of equality, requiring that if a jurisdiction holds an election, every citizen, adult resident has the right to vote in that election.8 The Warren Court in Reynolds v. Sims9 and its progeny, relying upon the Equal Protection Clause of the Fourteenth Amendment, established the second level of equality—the right to an equally weighted vote.10 In Bush v. Gore, the Court relied upon the Reynolds line of cases to move to a third level of equality—equality in the procedures and mechanisms used for voting. Part III explores a range of election law cases that may be subject to a “third level” political equality claim.11 It concludes that, if the case were taken seriously, Bush v. Gore should have great precedential value in changing a host of voting procedures and mechanisms, particularly when those procedures and mechanisms are challenged prospectively. Part IV of this Article explores the benefits, costs, and implications of expanding equal protection to such third level claims. The benefits of the approach are fairly obvious: a precedent requiring scrupulous equality in the holding of elections will increase resources used to conduct elections, so that at least twentieth century voting technology will be applied as we enter the twenty-first century. It will provide a means for those in poor, urban areas to have just as accurate a voting system as those used in wealthier areas. It also likely will ensure more reliable vote counting. 8. See infra note 81 and accompanying text. 9. 377 U.S. 533 (1964). 10. See infra note 82 and accompanying text. 11. I ignore the relevance that Bush v. Gore may have for equal protection claims outside the election law context. 380 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:377 But expanding political equality to the third level would be a mixed blessing. Putting aside the considerable costs associated with upgrading voting equipment, rewriting state and local election laws involving contested elections, and litigation over both types of changes, three concerns arise with extending equal protection jurisprudence to the nuts-and-bolts of elections. First, third-level claims provide a new reason and a pretext for federal courts to nullify state and local election results, thereby threatening both democracy and the judiciary. Second, third-level claims undermine federalism in a way that first- and second-level equal protection claims do not. Claims of local control over nuts-and-bolts voting mechanisms resonate more genuinely than claims of localities to deny the franchise to certain groups of individuals or to count votes unevenly. Third, thirdlevel claims create a disincentive for jurisdictions to experiment with new methods of voting, such as internet voting. Finally, it is worth thinking about the doctrinal implications of extending equal protection jurisprudence to the third level. It is unclear whether extension of equal protection to the third level differs meaningfully from arguments calling for greater political equality in terms of electoral structures (such as Justice Marshall’s argument in his dissent in Mobile v. Bolden12) and financing election campaigns (such as the arguments of Jamin Raskin and John Bonifaz13). The main, albeit unintended, precedent of Bush v. Gore may be to ease the way for future Supreme Court majorities to pursue their own visions of political equality without much thought about whether that vision is supported by existing case law. II. WHY WE SHOULD NOT TAKE BUSH V. GORE’S EQUAL PROTECTION HOLDING SERIOUSLY A. Optimism and the Equal Protection Holding Two days after the Supreme Court issued its opinion in Bush v. Gore, Professor Sam Issacharoff wrote in a New York Times op-ed of the “surprising expansion of voting rights”14 wrought by the opinion: [T]he Supreme Court may have given us an advancement in voting rights doctrine. It has asserted a new constitutional requirement: to avoid disparate and unfair treatment of voters. And this obligation obviously cannot be limited to the recount process alone. . . . 12. 446 U.S. 55, 116-17 (1980) (Marshall, J., dissenting). 13. Jamin Raskin & John Bonifaz, Equal Protection and the Wealth Primary, 11 YALE L. & POL’Y REV. 273 (1993) [hereinafter Raskin & Bonifaz, Wealth Primary]; Jamin Raskin & John Bonifaz, The Constitutional Imperative and Practical Superiority of Democratically Financed Elections, 94 COLUM. L. REV. 1160 (1994) [hereinafter Raskin & Bonifaz, Constitutional Imperative]. 14. Samuel Issacharoff, The Court’s Legacy for Voting Rights, N.Y. TIMES, Dec. 14, 2000, at A39. 2001] EQUAL PROTECTION’S FUTURE IN ELECTIONS 381 The court’s new standard may create a more robust constitutional examination of voting practices.15 Echoing Issacharoff’s conclusion that the legacy of the case “could be a substantial jolt of justice into the voting arena,”16 Village Voice columnist Nat Hentoff concluded that the Justices planned their precedent to be far-reaching: “The justices knew that with the way opened to election reforms, a lot of cases will be heading to the courts throughout the nation until all votes are counted according to uniform standards.”17 As the Yiddish expression goes, “From your mouth to God’s ears.” Whether the sentiments above represent a prediction by Issacharoff and Hentoff about the future of equal protection jurisprudence in elections or merely wishful thinking, I am far less sanguine that the case will have much precedential effect. To explain why, I begin by noting precisely what conduct the Bush v. Gore majority opinion held violated the Equal Protection Clause.18 The Florida election controversy reached the United States Supreme Court for the second time in Bush v. Gore, an appeal of the Florida Supreme Court’s second opinion.19 In the Florida Supreme Court opinion, a four to three majority reversed the trial court. The trial court held that Democratic presidential candidate Al Gore failed to meet Florida’s statutory requirement to contest the Florida vote and therefore rejected Gore’s demand for manual recounts of “undervotes” in selected Florida counties with large Democratic majorities.20 “Undervotes” were ballots that vote-counting machines recorded as containing no vote for President. Gore asserted that a recount of these votes would show enough legally valid votes cast in his favor, but not counted by the machine, to make up the extremely small difference in votes between Gore and Bush. 15. Id. 16. Id. 17. Nat Hentoff, A Jolt of Justice for All Voters: Supreme Court Redeemed, VILLAGE VOICE, Jan. 10-16, 2001. For additional optimistic assessments, see Steve France, Equal Protection Claims Likely to Prod States to Address Voting System Problems, 69 U.S.L.W. 2483 (Feb. 20, 2001). 18. This analysis assumes the reader is familiar with the facts of the case. As this incident fades into memory, that may no longer be true. Readers looking for more extended factual background on the case may consult LOWENSTEIN & HASEN, supra note 1, ch. 3; 37 Days: A Special Report: An American Diary; The Battle Unfolds Day by Day, L.A. TIMES, Dec. 17, 2000, at V3. 19. The election controversy first came before the Court in Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000), rev’g Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d 1220 (Fla. 2000). 20. Gore v. Harris, 772 So. 2d 1243, 1247 (Fla. 2000), rev’d sub nom. Bush v. Gore, 531 U.S. 98 (2000). 382 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:377 The Florida Supreme Court held that the trial court applied the wrong legal standards in judging the merits of Gore’s claim.21 However, rather than remand the case for the trial court to apply the correct legal standard to the facts, the Florida Supreme Court ordered that certain recounts conducted after the deadline it had set in an earlier case should be included in the totals and that recounts of undervotes should go forward.22 And rather than allow Gore to pick the counties for the recounts, the Florida court held that all Florida counties—and not just the counties singled out by Gore—had to conduct manual recounts of the undervotes.23 The court failed to respond to Chief Justice Wells’ observation in dissent that it was unfair to count only undervotes and not “overvotes”—that is, ballots that the vote-counting machines recorded as containing more than one valid vote for President.24 The court further held that in examining the undervotes to determine if the ballots indeed contained a valid vote for a presidential candidate, the counters should judge the ballots using a “clear intent of the voter” standard.25 The court failed to be more specific, perhaps out of fear that a more specific standard would open up the decision to charges that it violated Article II of the United States Constitution.26 In any case, the court ordered that the trial judge manage the statewide recount,27 which needed to be completed in short order. 21. Id. at 1252. 22. Id. at 1261-62. 23. Id. 24. Id. at 1264 n.26 (Wells, C.J., dissenting). It is not clear that all overvotes recounted by hand would necessarily be classified as invalid votes. For example, a voter who wrote Al Gore’s name in the write-in portion of a ballot and also punched out the chad for Al Gore clearly intended to vote for Al Gore, but the counting machine would record that vote as an overvote. 25. Id. at 1262. 26. Article II of the Constitution vests in each state’s legislature the power to prescribe the state’s rules for choosing presidential electors. The Article II issues are beyond the scope of this Article. The main thrust of the Article II argument in Chief Justice Rehnquist’s concurrence is that the Florida Supreme Court’s novel interpretation of legislatively enacted statutes regulating election contests in the Bush-Gore dispute constituted a change in the law in violation of Article II. On these issues, see James Gardner, The Regulatory Role of State Constitutional Structural Constraints in Presidential Elections, 29 FLA. ST. U. L. REV. 625 (2001); Robert Schapiro, Conceptions and Misconceptions of State Constitutional Law in Bush v. Gore, 29 FLA. ST. U. L. REV. 661 (2001); see also Richard H. Pildes, Judging “New Law” in Election Disputes, 29 FLA. ST. U. L. REV. 691 (2001). Suffice it to say that it was clear from the United States Supreme Court’s first opinion in the case, Bush v. Palm Beach County Canvassing Board, 531 U.S. 70, 76 (2000), that even if the Florida Supreme Court thought it was legitimately filling gaps and reconciling conflicting statutes in Florida’s election law, it ran the risk of its opinion being characterized as a “change of law” in violation of Article II. But cf. Richard A. Posner, Florida 2000: A Legal and Statistical Analysis of the Election Deadlock and the Ensuing Litigation, 2000 SUP. CT. REV. 1, 37 (“That was a gap in the statute that a court applying normal principles of statutory interpretation might fill.”). 27. Gore v. Harris, 772 So. 2d at 1262. 2001] EQUAL PROTECTION’S FUTURE IN ELECTIONS 383 The Florida court remanded the case to the original trial judge, who recused himself. Another trial judge ordered the manual recounts to begin of the Miami-Dade ballots (that had been shipped to Tallahassee for the election contest) and in counties across Florida. Just as the counts began on Saturday, December 9, the United States Supreme Court, by a five to four vote, stayed the Florida Supreme Court’s order, thereby suspending the recount.28 Justice Scalia, in his opinion concurring to the granting of a stay, explained that a majority of the Court believed that Bush had a substantial likelihood of success on the merits and stood to face irreparable harm29—the Court’s standard for issuing a stay.30 Late in the evening of Tuesday, December 12, the Supreme Court issued its opinion on the merits. Five Justices (Chief Justice Rehnquist and Justices Kennedy, O’Connor, Scalia, and Thomas) joined in a per curiam opinion reversing the Florida court on equal protection grounds.31 The Chief Justice, joined by Justices Scalia and Thomas, issued a concurring opinion presenting as alternative grounds for reversal that the Florida Supreme Court’s order violated Article II of the Constitution.32 Four Justices dissented (Justices Breyer, Ginsburg, Souter, and Stevens),33 although Justices Breyer and Souter expressed some support for the equal protection argument but not the remedy.34 28. Bush v. Gore, 531 U.S. 1046 (2000). 29. Id. at 1047 (Scalia, J., concurring). I could write much beyond the scope of this Article about Justice Scalia’s concurrence, particularly his view of what constituted “irreparable harm” to Bush and why he failed to balance the equities of harm to Gore from granting the stay. I note here only that I am aware of no empirical support available at the time the Court issued the stay for Justice Scalia’s statement that “it is generally agreed that each manual recount produces a degradation of the ballots, which renders a subsequent recount inaccurate.” Id. 30. Rostker v. Goldberg, 448 U.S. 1306, 1308 (1980) (Brennan, J., in chambers). The four requirements are: (1) There is a reasonable probability that four Justices will vote to grant certiorari or note probable jurisdiction; (2) There is “a fair prospect that a majority of the Court will conclude that the decision below was erroneous”; (3) Irreparable harm is likely to result from the denial of a stay; and (4) In balancing the equities, taking into account the harm to both parties as well as the interests of the public at large, the stay should be granted. Id. 31. Bush v. Gore, 531 U.S. 98, 100 (2000). 32. Id. at 111 (Rehnquist, C.J., concurring). 33. Id. at 123 (Stevens, J., dissenting); id. at 129 (Souter, J., dissenting); id. at 135 (Ginsburg, J., dissenting); id. at 143 (Breyer, J., dissenting). 34. I can conceive of no legitimate state interest served by these differing treatments of the expressions of voters’ fundamental rights. . . . I would . . . remand the case to the courts of Florida with instructions to establish uniform standards for evaluating the several types of ballots that have prompted differing treatments, to be applied within and among counties when passing on such identical ballots in any further recounting (or successive recounting) that the courts might order. Id. at 134-35 (Souter, J., dissenting). 384 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:377 The per curiam opinion’s analysis began by setting forth the applicable law. It noted that individual citizens have no federal constitutional right to vote for electors for the President of the United States.35 However, “[w]hen the state legislature vests the right to vote for President in its people,” as Florida had, “the right to vote as the legislature has prescribed is fundamental; and one source of its fundamental nature lies in the equal weight accorded to each vote and the equal dignity owed to each voter.”36 The Court continued: The right to vote is protected in more than the initial allocation of the franchise. Equal protection applies as well to the manner of its exercise. Having once granted the right to vote on equal terms, the State may not, by later arbitrary and disparate treatment, value one person’s vote over that of another. See, e.g., Harper v. Virginia Bd. of Elections, 383 U.S. 663, 665 (1966). . . . It must be remembered that “the right of suffrage can be denied by a debasement or dilution of the weight of a citizen’s vote just as effectively as by wholly prohibiting the free exercise of the franchise.” Reynolds v. Sims, 377 U.S. 533, 555 (1964).37 After noting that “[t]he question before us . . . is whether the recount procedures the Florida Supreme Court has adopted are consistent with its obligation to avoid arbitrary and disparate treatment of the members of its electorate,”38 the Court answered the question in the negative. It held that the recount mechanism adopted by the Florida Supreme Court did “not satisfy the minimum requirement for nonarbitrary treatment of voters necessary to secure the fundamental right”39 under the Equal Protection Clause for four related reasons: 1. Although the Florida court had instructed that those individuals conducting the manual recounts judge ballots by discerning the “intent of the voter,” it failed to formulate uniform rules to determine such intent, such as whether to count as a valid vote a ballot with a chad hanging by two corners. The standards for whether to count a I agree that, in these very special circumstances, basic principles of fairness should have counseled the adoption of a uniform standard to address the problem. . . . Nonetheless, there is no justification for the majority’s remedy, which is simply to reverse the lower court and halt the recount entirely. An appropriate remedy would be, instead, to remand this case with instructions that, even at this late date, would permit the Florida Supreme Court to require recounting all undercounted votes in Florida . . . and to do so in accordance with a single uniform substandard. Id. at 146 (Breyer, J., dissenting). 35. Id. at 104. 36. Id. 37. Id. at 104-05. 38. Id. at 105. 39. Id. 2001] EQUAL PROTECTION’S FUTURE IN ELECTIONS 385 ballot differed “not only from county to county but indeed within a single county from one recount team to another.”40 2. The recounts already undertaken included a manual recount of all votes in selected counties, including both undervotes and overvotes, but the new recounts ordered by the Florida court included only undervotes. As a result, the citizen whose ballot was not read by a machine because he failed to vote for a candidate in a way readable by a machine may still have his vote counted in a manual recount; on the other hand, the citizen who marks two candidates in a way discernable by the machine will not have the same opportunity to have his vote count, even if a manual examination of the ballot would reveal the requisite indicia of intent.41 3. The Florida Supreme Court had ordered that the current vote totals include results of a partial recount from Miami-Dade County. From this fact, the Supreme Court concluded that “[t]he Florida Supreme Court’s decision thus gives no assurance that the recounts included in a final certification must be complete.”42 4. The Florida Supreme Court did not specify who would count the ballots, forcing county boards to include team members without experience in recounting ballots. Nor were observers permitted to object during the recount.43 After reaching its holding, the U.S. Supreme Court declined to remand the case to the Florida Supreme Court to order procedures satisfying these concerns, as two dissenting Justices urged.44 Putting aside the Article II problem,45 a remand order would have been entirely manageable.46 Nonetheless, the Court held that the Florida Supreme Court had recognized the Florida Legislature’s intention to participate fully in the federal electoral process. Under a federal 40. Id. at 106. The Court noted that the vote totals already approved by the Florida Supreme Court included recount totals from counties using various methods of counting: “Broward County used a more forgiving standard than Palm Beach County, and uncovered almost three times as many new votes, a result markedly disproportionate to the difference in populations between the counties.” Id. at 107. The Court did not respond to Justice Stevens’ point in dissent that concerns about the new recounts “are alleviated—if not eliminated—by the fact that a single impartial magistrate will ultimately adjudicate all objections arising from the recount process.” Id. at 126 (Stevens, J., dissenting). 41. Id. at 108. The Court continued: “Furthermore, the citizen who marks two candidates, only one of which is discernable by the machine, will have his vote counted even though it should have been read as an invalid ballot.” Id. 42. Id. 43. Id. at 109. 44. See supra note 34. 45. See supra note 26. 46. The Court disagreed. Bush v. Gore, 531 U.S. at 110. The Electoral College did not vote until December 18, and Congress did not count the electoral votes until January 6, 2001. 386 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:377 statute,47 states that designate their electors by a certain date, in this election by December 12, cannot have their choice challenged in Congress when Congress later counts the electoral votes. That date [of December 12] is upon us, and there is no recount procedure in place under the State Supreme Court’s order that comports with minimal constitutional standards. Because it is evident that any recount seeking to meet the December 12 date will be unconstitutional for the reasons we have discussed, we reverse the judgment of the Supreme Court of Florida . . . .48 B. Reasons for Doubting Bush v. Gore’s Precedential Value At first glance, Issacharoff’s and Hentoff’s optimism is entirely understandable. As Justice Stevens noted in dissent, never before had the Supreme Court “called into question the substantive standard by which a State determines that a vote has been legally cast.”49 The Court for the first time expressed its willingness to get its hands into to the nitty-gritty details of vote counting—one would have been hard-pressed before this case to imagine the Supreme Court delving into the law of hanging chads. If chads could be questioned, then why not a more “robust” exploration of not only the mechanics of elections but state and local laws governing election contests as well? And if courts are to look at the minutiae of election contests, perhaps courts would also consider structural issues, such as the financing of elections, which might create conditions of inequality among voters. Nonetheless, for at least three reasons I doubt this optimistic assessment. 1. Limiting Language First, the Court’s language explicitly limiting its holding to the facts of this case is extraordinary. After stating the four ways in which the Florida procedures violated the Equal Protection Clause, the Court wrote: “Our consideration is limited to the present circumstances, for the problem of equal protection in election processes generally presents many complexities.”50 By this statement, the Court appeared to dismiss any precedential value this case may have for future election law cases. This is a strong deviation from the Court’s usual practice in election cases. Take campaign finance for example. When the Court considered the constitutionality of the 1974 amendments to the Federal 47. 48. 49. 50. 3 U.S.C. § 5 (1994). Bush v. Gore, 531 U.S. at 110. Id. at 125 (Stevens, J., dissenting). Id. at 109. 2001] EQUAL PROTECTION’S FUTURE IN ELECTIONS 387 Election Campaign Act in Buckley v. Valeo,51 the Court, also in a per curiam opinion decided under rushed circumstances, resolved excruciatingly difficult and complex issues related to the First Amendment, corruption, political equality, and democracy. Yet the Court did not limit the holding in Buckley to the particular facts of the case. Far from it; more than twenty-five years after the opinion was issued, the Court continues to look to Buckley as providing the proper starting point for evaluating the constitutionality of various campaign finance laws.52 Nor was the Court merely silent on the issue of Bush v. Gore’s precedential value. It expressly denied the case had any precedential value, something the Court could have suggested more subtly in distinguishing Bush v. Gore’s facts in future cases to come before it. The Court further noted that “[t]he question before the Court is not whether local entities, in the exercise of their expertise, may develop different systems for implementing elections.”53 When future courts consider litigation challenging the electoral practices of local entities, no doubt the lawyers representing these entities will point out that Bush v. Gore is expressly limited to those situations where “a court orders a statewide remedy” and then fails to give “at least some assurance that the rudimentary requirements of equal treatment and fundamental fairness are satisfied.”54 Following Bush v. Gore, it is hard to imagine many cases falling into that category. 2. The Court’s Failure to Engage in Serious Analysis Second, we should not take Bush v. Gore’s holding seriously because the Court itself did not take its holding seriously.55 The per curiam opinion, no doubt, amounted to a great extension of precedent, yet the Court never explained why it was extending precedent in this case. As authority for its holding, the Court relied principally upon two cases, Reynolds v. Sims56 and Harper v. Virginia Board of Elec- 51. 424 U.S. 1 (1976). 52. In one of the most recent of these cases, Nixon v. Shrink Missouri Government PAC, 528 U.S. 377, 397-98 (2000), the Court went out of its way to show how its opinion was entirely consistent with Buckley. 53. Bush v. Gore, 531 U.S. at 109. 54. Id. 55. Judging by the dissents’ focus on Chief Justice Rehnquist’s concurring opinion on Article II grounds rather than on equal protection, one wonders if the equal protection ground was an afterthought conjured up by Justices O’Connor and Kennedy, who may have been uncomfortable with Chief Justice Rehnquist’s sharp rebuke of the Florida Supreme Court’s reasoning and integrity. See Linda Greenhouse, Bush v. Gore: A Special Report; Election Case a Test and Trauma for Justices, N.Y. TIMES, Feb. 20, 2001, at A1 (“[A]lthough intended as a majority opinion, the chief justice’s opinion failed to get the support of Justices Kennedy and O’Connor.”). 56. 377 U.S. 533 (1964). 388 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:377 tions.57 Reynolds held that it is an equal protection violation to elect members of a state or local legislative body from unequally populated districts.58 According to Reynolds, “[d]iluting the weight of votes because of place of residence impairs basic constitutional rights under the Fourteenth Amendment just as much as invidious discriminations based upon factors such as race, or economic status.”59 The Reynolds Court took forty pages in the U.S. Reports to justify this deviation from past precedent. In Harper, the Court relied upon Reynolds in striking down a poll tax on equal protection grounds: “wealth or fee paying has . . . no relation to voting qualifications; the right to vote is too precious, too fundamental to be so burdened or conditioned.”60 Neither case involved the mechanics of elections which had, heretofore, been seen to be a matter for local officials. Indeed, the Court in recent years has expressed great deference to local officials who wish to structure their elections in the way they see fit. In Munro v. Socialist Workers Party,61 for example, the Court held that the state of Washington’s interest in preventing ballot “confusion” by voters justified its rules that kept most third-party candidates off the general election ballot. The Court held that the state need not even provide any empirical evidence that its rules were necessary to prevent such confusion: To require States to prove actual voter confusion, ballot overcrowding, or the presence of frivolous candidacies as a predicate to the imposition of reasonable ballot access restrictions would invariably lead to endless court battles over the sufficiency of the “evidence” marshaled by a State to prove the predicate. Such a requirement would necessitate that a State’s political system sustain some level of damage before the legislature could take corrective action.62 That is not to say that the Court was wrong in Bush v. Gore in extending equal protection to the mechanics of elections. However, even under the admittedly great time pressure of the case, the Court could have gone a long way toward showing that it took the exercise seriously by including a sentence or two justifying, or at least acknowledging, that the holding greatly expanded past precedent. Perhaps the best evidence that the Court did not take the analysis seriously was its resolution of the case. The Court recognized that voting is a “fundamental right,” and that “the State may not, by later 57. 58. 59. 60. 61. 62. 383 U.S. 663 (1966). Reynolds, 377 U.S. at 583-84. Id. at 566 (citations omitted). Harper, 383 U.S. at 670. 479 U.S. 189 (1986). Id. at 195. 2001] EQUAL PROTECTION’S FUTURE IN ELECTIONS 389 arbitrary and disparate treatment, value one person’s vote over that of another.”63 It is hornbook law that laws infringing on fundamental rights, including voting, must be judged under the standard of strict scrutiny—that is, that the state must have a compelling interest in treating voters differently and that the means must be narrowly tailored to meet that interest. The Court did nothing to suggest that anything less than strict scrutiny, such as an easier to meet “rational basis test,” should apply to analyze burdens on the fundamental right of voting in this context.64 Nonetheless, the Court held that the Florida Legislature’s interest (which the Supreme Court said was recognized by the Florida Supreme Court) in taking advantage of the “safe harbor” provisions of federal law for counting the state’s electoral votes trumped the rights of all Florida voters to have valid votes counted.65 It is not selfevident that such a state interest was compelling and trumped the right, recognized in Reynolds but ignored by the Court in Bush v. Gore, to have every vote count: It has been repeatedly recognized that all qualified voters have a constitutionally protected right to vote and to have their votes counted. In Mosley the Court stated that it is “as equally unquestionable that the right to have one’s vote counted is as open to protection . . . as the right to put a ballot in a box.” The right to vote can neither be denied outright, nor destroyed by alteration of ballots, nor diluted by ballot-box stuffing. As the Court stated in Classic, “Obviously included within the right to choose, secured by the Constitution, is the right of qualified voters within a state to cast their ballots and have them counted . . . .”66 Suppose evidence existed that Florida officials had failed to count the votes of African-American voters because of racial animus, and the Florida Supreme Court ordered a recount of votes that would require time beyond December 12, 2000. It is not clear that the “safe harbor” provision should have trumped the right to have every vote count. Now, perhaps one could argue, even under those circumstances, that Florida’s interest in meeting the deadline was indeed compelling and that there were no other means to achieve that goal. 63. Bush v. Gore, 531 U.S. 98, 104-05 (2000). 64. It might well be that the Court would have held that the Florida recount procedures flunked even a rational basis test given their “arbitrary” nature. But the Court did not say or even suggest that it was relaxing the strict scrutiny it had applied in the past to these voting cases. Indeed, Harper and Reynolds, the only cases relied upon by the majority, are among the important cases establishing that strict scrutiny applies to burdens on voting. 65. Bush v. Gore, 531 U.S. at 110. 66. Reynolds v. Sims, 377 U.S. 533, 554-55 (1964) (emphases added and citations omitted). It will not do to argue that the votes were counted in the machine count. Florida law, as the law of many other states, allowed for manual recounts of votes precisely because machines made errors and sometimes failed to count valid votes. 390 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:377 But the Supreme Court never even bothered to undertake the analysis in Bush v. Gore, suggesting that the fundamental right to vote was not so fundamental after all.67 As Judge Posner remarked, “[t]here was an air of non sequitur to ruling that the Florida supreme court had violated the Constitution by failing to prescribe uniform criteria for a recount, yet terminating the recount rather than permitting it to go forward under proper criteria.”68 3. Inconsistency in Equal Protection Analysis The final reason not to take the Supreme Court’s equal protection holding seriously is that it constitutes a strong break from the conservative majority’s usual approach to equal protection and, therefore, it will not likely be extended or embraced by them in future cases. The argument here is not the “crude” one that “[t]he five Justices are ‘conservative,’ and ‘conservative’ judges don’t ‘like’ the Equal Protection Clause.”69 These Justices have shown that they like the Equal Protection Clause just fine, when it is used to pursue claims more consistent with their ideology. There was no such thing as a claim of an “unconstitutional racial gerrymander” before these same five Justices decided Shaw v. Reno70 in 1993, a holding grounded in equal protection.71 Shaw and its progeny have been used to limit the extent to which race may be taken into account in redistricting to benefit minority-preferred candidates for elective office. It is not so much that these Justices do not “like” equal protection as that we would not have expected them to use the Equal Protection Clause to create new federal oversight of the minutiae of state and local elections. Besides the federalism costs which make the majority’s holding surprising,72 no Rehnquist Court opinion had ever relied upon Reynolds or Harper to expand oversight of the electoral process or to expand the franchise.73 One would have expected these Justices to agree with Judge Posner’s observations about the case: 67. Thus, at least under Florida law, those votes were not counted. The Supreme Court’s opinion kept these votes uncounted. Note that this hypothetical situation also appears to run afoul of the Fifteenth Amendment’s prohibition on abridgement of the right to vote on account of race. 68. Posner, supra note 26, at 48. 69. Id. at 56. 70. 509 U.S. 630 (1993). 71. The Court, with different personnel, rejected a similar (if not identical) claim in 1977 in United Jewish Org. of Williamsburgh, Inc. v. Carey, 430 U.S. 144 (1977). See Shaw, 509 U.S. at 658 (White, J., dissenting) (“The facts of the case here mirror those in” United Jewish Organization). 72. I discuss these costs below in Part IV.B. There are significant federalism costs with the Shaw line of cases as well. See Daniel Hays Lowenstein, You Don’t Have to Be Liberal to Hate the Racial Gerrymandering Cases, 50 STAN. L. REV. 779 (1998). 73. When the butterfly ballot case arose, Erwin Chemerinsky and I each independently suggested that a revote could be demanded under the authority of Reynolds. I made 2001] EQUAL PROTECTION’S FUTURE IN ELECTIONS 391 Such differences [in how votes are counted] had not previously been thought to deny equal protection of the laws and if they are now to do so this portends an ambitious program of federal judicial intervention in the electoral process, a program the Supreme Court seems, given the haste with which it acted, to have undertaken without much forethought about the program’s scope and administrability. The last thing we need is more election litigation.74 Judge Posner defended the Supreme Court’s decision not on equal protection grounds, which he dismissed in three paragraphs of his lengthy article, but instead on grounds of “rough justice,” if not “legal justice.”75 “I cannot see the case for precipitating a political and constitutional crisis merely in order to fuss with a statistical tie that, given the inherent subjectivity involved in hand counting spoiled ballots, can never be untied.”76 Similarly, Professor Charles Fried, who represented the Florida Legislature in an amicus curiae brief before the Supreme Court, wrote that the Court’s analysis on why it failed to remand the case to the Florida Supreme Court to implement its novel equal protection holding was “the least convincing portion of the Court’s opinion.”77 In sum, the limiting language in the opinion, the lack of seriousness with which the Court undertook its own analysis, and the inconsistency of the opinion with other jurisprudence by this majority of Justices all point in the direction of assuming that Bush v. Gore is not good precedent for an expansive reading of equal protection law in elections. Embarrassment provides the only hope that the case will have precedential value. Conservative Justices decided a case in which their decision effectively chose a President who was far more likely than the losing candidate to choose additional conservative Justices my comments on an election law internet discussion group that I manage with Dan Lowenstein and to numerous reporters. Chemerinksy made the argument in Erwin Chemerinsky, Palm Beach County Must Vote Again, L.A. TIMES, Nov. 12, 2000, at M5. The general response to such claims was skepticism given their novelty. Georgene M. Vairo, Bush v. Gore, NAT’L L.J., Feb. 12, 2001, at A16 (“Before the Supreme Court decided Bush v. Gore . . . it is likely that only a liberal, results-oriented law professor would have thought Mr. Gore would have a prayer in federal court to obtain injunctive relief on . . . [equal protection] claims.”). 74. Posner, supra note 26, at 41. Posner then argues that a due process claim would have been more defensible. “Yet even this would not be an inconsequential doctrinal step— the creation of a federal duty to use uniform precise criteria in a recount.” Id. at 42. 75. Id. at 60. For a critique of Posner’s argument in favor of the result in Bush v. Gore, see Richard L. Hasen, A “Tincture of Justice”: Judge Posner’s Failed Rehabilitation of Bush v. Gore, 80 TEX. L. REV. 137 (2001). 76. Posner, supra note 26, at 46. 77. Charles Fried, A Badly Flawed Election: An Exchange, N.Y. REV. OF BOOKS, Feb. 22, 2001, available at http://www.nybooks.com/articles/14004. 392 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:377 to fill Supreme Court vacancies.78 An opinion whose holding is limited to the facts of the case will lead to further claims that the Court decided the case using the not-so-venerable principle of constitutional interpretation: “Bush wins.” To blunt that criticism, the Court may tolerate giving the opinion some precedential value. I turn now to that possibility. III. TAKING BUSH V. GORE’S EQUAL PROTECTION HOLDING SERIOUSLY As Professor Jack Balkin of Yale Law School observed shortly after the Court decided Bush v. Gore, a rule applied only to one case “isn’t consistent with rule-of-law principles.”79 To be consistent with such principles, like cases are to be treated alike. This Part examines which cases are “like cases” compared to Bush v. Gore. This is not a futile exercise even if I am correct in Part II that the Supreme Court ultimately will limit Bush v. Gore to its facts. Lower courts will first apply Bush v. Gore as precedent to cases coming before it, and the Supreme Court might decline to review some of those cases. So there is at least a window of time in which the case may serve as valid precedent. To make a determination of which cases are “like cases,” I begin by restating the holding of the case as briefly as I can. The Court held that “[h]aving once granted the right to vote on equal terms, the State may not, by later arbitrary and disparate treatment, value one person’s vote over that of another.”80 Florida violated this rule by: (1) failing to formulate uniform rules for judging the “intent of the voter” in a manual recount; (2) ordering only a selective recount of “undervotes” rather than a recount of all votes; (3) leaving open the possibility of certifying vote totals from incomplete recounts; and (4) failing to specify who would count the ballots or the procedures for objection. This holding moves equal protection analysis in election law cases to a third level of equality. Various amendments to the Constitution and Supreme Court cases decided by the Warren Court established the first level of equality, requiring that if a jurisdiction holds an election, every citizen, adult resident has the right to vote in that 78. See Bruce Ackerman, The Court Packs Itself, AM. PROSPECT, Feb. 12, 2001. Jon Elster refers to the “civilizing force of hypocrisy,” which in this case could lead the Court to grant more precedential value to the case than first planned. Jon Elster, Introduction to DELIBERATIVE DEMOCRACY 12 (Jon Elster ed., 1998). 79. Marcia Coyle, Gauging ‘Bush v. Gore’ Fallout: Will Equal Protection Language Open a Can of Electoral Worms?, NAT’L L.J., Dec. 25, 2000, at A4 (quoting Professor Balkin). 80. Bush v. Gore, 531 U.S. 98, 104-05 (2000). 2001] EQUAL PROTECTION’S FUTURE IN ELECTIONS 393 election.81 The Warren Court in Reynolds and its progeny, relying upon the Equal Protection Clause of the Fourteenth Amendment, established the second level of equality—the right to an equally weighted vote.82 In Bush v. Gore, the Court relied upon the Reynolds line of cases to move to a third level of equality—equality in the procedures and mechanisms used for voting. The Bush v. Gore Court did not explain which kinds of procedures and mechanisms used for voting constitute “arbitrary and disparate treatment” that “value one person’s vote over another” in violation of the Equal Protection Clause. Accordingly, the most we can do is look at the four ways in which the Court held that Florida did so and compare those ways to factual allegations in new cases.83 Thus, consider which, if any, of the following five hypothetical allegations should be cognizable as an equal protection violation: 1. In the state of Pacifica, voters in some counties vote using punch card voting systems in which they must vote by punching out a chad with a stylus. Voters in other counties vote using optical scanning systems in which they must vote by filling in a bubble with a pencil. The rate at which punch card votes are rejected by vote tabulating equipment is almost 4% compared to an approximately 1.5% rate for rejection of ballots read by optical scanning equipment.84 Optical scanning equipment is more expensive, and perhaps for that reason it has been adopted in counties with higher per capita incomes.85 81. See generally LOWENSTEIN & HASEN, supra note 1, ch. 3; ISSACHAROFF ET AL., supra note 1, at 304 (referring to Harper as a “first generation” voting rights claim). 82. LOWENSTEIN & HASEN, supra note 1, ch. 4; ISSACHAROFF ET AL., supra note 1, at 303-05 (discussing Harper, Reynolds, and Bush v. Gore). The latter source’s authors call cases dealing with statutory design of democratic institutions “second generation” cases, a different meaning than my “second level” term. Id. at 304. 83. The opacity of the Court’s equal protection holding may be the best thing about the opinion because it gives lower courts a chance to experiment with the new equal protection holding. I explain this point more fully in Richard L. Hasen, The Benefits of Judicially Unmanageable Standards in Election Cases Under the Equal Protection Clause, 80 N.C. L. REV. (forthcoming 2002). 84. The percentage of nonvotes in [Florida’s 2000 presidential] election in counties using a punch-card system was 3.92%; in contrast, the rate of error under the more modern optical scan systems was only 1.43%. Put in other terms, for every 10,000 votes cast, punch-card systems result in 250 more nonvotes than optical-scan systems. Bush v. Gore, 531 U.S. at 126 n.4 (Stevens, J., dissenting) (citation omitted). In the wake of the Florida controversy, a number of empirical studies have examined rates of undervotes across different voting mechanisms. One of the most thorough studies has been conducted by the Caltech-MIT Voting Technology Project. The study concluded that “[p]unch cards . . . lose at least 50 percent more votes than optically scanned paper ballots.” CALTECH-MIT VOTING TECHNOLOGY PROJECT, VOTING: WHAT IS, WHAT COULD BE 21 (July 2001). 85. At least five lawsuits have been filed in light of Bush v. Gore alleging that such differences constitute a denial of equal protection. See Compl. for Injunctive and Declara- 394 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:377 2. Same facts as in the first example, but the systems differ across states voting in a presidential election. 3. Two candidates in a local election in the state of Pacifica compete for the job of county dogcatcher. Smith defeats Jones by ten votes in an election using a punch card voting system. Jones demands a recount, which is conducted by hand using the statemandated “intent of the voter” standard. The state statute provides no further guidance on how to judge intent. Under the recount, Jones wins by three votes. Smith, and voters supporting Smith, sue to have the results overturned. 4. State voting officials “purge” from their voter rolls the names of voters who have not voted in the last two elections. Officials claim that they are doing so to prevent fraud—a number of names removed are of people who the state said, incorrectly, were convicted felons who have lost the right to vote—but plaintiffs claim the purpose is to remove as many African-American voters as possible. Alternatively, plaintiffs claim the purge law has a disparate impact on AfricanAmerican voters.86 5. In an effort to make it easier for elderly voters to see the ballot and vote effectively, county election officials design a ballot using a “butterfly ballot” design. In this design, the place for voters to record votes is along the center spine of the ballot. Voters allege after voting that they were confused by the ballot design, leading many of them to vote for a third-party candidate for an elective office rather than for their preferred candidate. Statistics unambiguously show that there is virtually no chance that this third-party candidate simply received proportionally more votes in this county than in other counties in the state.87 Voters in other counties in the state did not use the butterfly ballot. tory Relief at 19, Common Cause v. Jones, No. 01-03470-SVW (C.D. Cal. filed Apr. 17, 2001), available at http://election2000.stanford.edu/Complaint.aclu.sc.pdf; Class Action Compl. at 13-15, NAACP v. Harris, No. 01-CIV-120-GOLD (S.D. Fla. filed Jan. 10, 2001), available at http://election2000.stanford.edu/11001harris.pdf; Compl. at 1-2, 9-10, Black v. McGuffage, No. 01C-0000208 (N.D. Ill. filed Jan. 11, 2001), available at http://election2000.stanford.edu/Illinois.aclu.pdf; Class Action Compl. for Declaratory and Injunctive Relief at 40, Wirth v. Election Sys. Software, Inc., No. 01-MR-6 (Ill. Cir. Ct. filed Jan. 9, 2001), available at http://election2000.stanford.edu/votomaticil.pdf; Compl. at 1, 6, Andrews v. Cox, No. 01-CZ-32490 (Ga. Fulton County Ct. filed Jan. 5, 2001), available at http://election2000.stanford.edu/aclu.cox.pdf. For a summary of these cases through September 2001, see B.J. Palermo, Bush-Gore Lives On, NAT’L L.J., Sept. 17, 2001, at A1. 86. For an example of such a lawsuit, see Class Action Compl. at 18-22, NAACP v. Harris, No. 01-CIV-120-GOLD (S.D. Fla. filed Jan. 10, 2001), available at http://election2000.stanford.edu/11001harris.pdf. 87. On the statistics related to overvoting, voting for Reform Party candidate Pat Buchanan, and the butterfly ballot used in Palm Beach County, Florida, during the disputed election, see Henry E. Brady, Report on Voting and Ballot Form in Palm Beach County (Nov. 16, 2000), at http://elections.fas.harvard.edu/statement/hbrady/hbrady.pdf; see also 2001] EQUAL PROTECTION’S FUTURE IN ELECTIONS 395 Each of these five examples present fact patterns with both similarities to and differences from the facts of Bush v. Gore. The question is which similarities and differences should matter legally. In other words, which cases are sufficiently “like” Bush v. Gore so that Bush v. Gore should be precedent? In the first hypothetical, there is little question that the use of different voting systems with different error rates treats voters differently and makes it less likely that voters in punch card districts will cast votes that count. Voters in counties using optical scanning equipment have a much better chance of having their votes counted than those in counties using a punch card ballot system. The disparate treatment is all the more disturbing to the extent that it correlates with wealth,88 looking functionally like the poll tax the Court struck down in Harper. Under strict scrutiny, this disparate treatment in the counting of votes appears just as “dilutive” of the right to vote and just as “arbitrary” as the different methods of recounting votes struck down in Bush v. Gore. There is no compelling interest for the different treatment; a decision about resource allocation by localities should not be able to trump a “fundamental right.” Furthermore, it appears irrelevant that the choice of voting machine technology was not the product of intentional discrimination or animus against any voters or groups of voters. In Harper, the Court held that a poll tax is unconstitutional even absent evidence that its intent was to discriminate against voters on the basis of race or wealth. In Bush v. Gore, the Court did not base its holding on intentional discrimination by Florida officials (or the Florida Supreme Court). In sum, if Bush v. Gore indeed has precedential value, it clearly should apply to prevent the use of these different voting systems in the same election.89 Different voting systems function in the same discriminatory manner as different means to count votes in a manual recount. The result of this case might be different if a court applied only a rational basis standard to the different procedures. The decision of which voting systems to use appears to be a resource allocation deciDon Van Natta, Counting the Vote: The Ballot; Gore Lawyers Focus on Ballot in Palm Beach County, N.Y. TIMES, Nov. 16, 2000, at A29. 88. In fact, as an empirical matter, it appears (counterintuitively) that wealthier areas are somewhat more likely to use punch card ballots than poorer areas. See E-mail from Professor Stephen Ansolabehere, Professor of Political Science, Massachusetts Institute of Technology, to author (Oct. 10, 2001) (on file with the Florida State University Law Review) (finding this statistical relationship but noting that it may not be statistically significant when the model controls for other variables). 89. Under this reasoning, one may rightly question whether McDonald v. Board of Election Commissioners, 394 U.S. 802 (1969), remains good law. In McDonald, the Court held that a state need not provide for absentee voting at all, and if it does so the state need not provide it for all voters. 396 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:377 sion that a court could deem a legitimate one. Nonetheless, Bush v. Gore appears to mandate strict scrutiny, not application of rational basis review.90 The second hypothetical is more complex. The first hypothetical establishes that the Equal Protection Clause affords a right to jurisdiction-wide uniformity in the methods for conducting elections. In a presidential election, the jurisdiction is the entire nation. The need for uniformity itself is echoed in the Constitution, which requires a uniform day for choosing presidential electors.91 On the other hand, each state picks its own electors for the Electoral College, so equality in the weighting of votes across states is affirmatively rejected in the Constitution. Moreover, the Equal Protection Clause by its own terms provides that “no state shall” deny equal protection of the laws; in differences across states, perhaps the Clause is not even implicated.92 Thus, a textual constitutional argument might allow treating the second hypothetical differently. The third hypothetical appears the easiest to resolve under Bush v. Gore. In its equal protection analysis, the Court spent most of its time explaining its view that the Florida Supreme Court’s failure to further define the “intent of the voter” standard violated equal protection.93 Justice Stevens, in dissent, pointed out that numerous states used such a standard or its equivalent in setting forth the standards for manual recounts.94 The majority did nothing to suggest that Florida law on this point was unique in some way. It is difficult to see how any of these standards survive Bush v. Gore. But, as explained below, a court’s finding that the standard violates the Constitution does not require that the election results be overturned.95 The fourth hypothetical is easy to resolve if plaintiffs can prove intentional discrimination. In that case, plaintiffs do not need Bush v. Gore to make the equal protection claim. Preexisting case law established that purposeful race discrimination in voting is unconstitutional.96 Where Bush v. Gore might be helpful is in getting around Bolden’s holding that disparate racial impact of an electoral structure (like the failure to use districting to elect members of a city commission) does not violate the Equal Protection Clause. Since it is often difficult to meet Bolden’s discriminatory intent requirement, to the extent that plaintiffs can recast their case as a Bush v. Gore claim—one involving “arbitrary and disparate treatment” that 90. 91. 92. 93. 94. 95. 96. See supra note 64 and accompanying text. U.S. CONST. art. II, § 1, cl. 4. Id. amend. XIV, § 1. Bush v. Gore, 531 U.S. 98, 105-06 (2000). Id. at 125 n.2 (Stevens, J., dissenting). See infra note 103 and accompanying text. Mobile v. Bolden, 446 U.S. 55, 65-70 (1980). 2001] EQUAL PROTECTION’S FUTURE IN ELECTIONS 397 “value[s] one person’s vote over that of another” in violation of the Equal Protection Clause—perhaps their claim will fare better.97 The fifth and final hypothetical presents the most difficult issue, but one that should still be considered a violation of equal protection. On the one hand, the case fits comfortably into the holding of Bush v. Gore and the other hypothetical cases: voters are being treated differently depending upon the county in which they live. Imagine if voters in one county could walk right up to the polls, but voters in another county had to walk up a steep hill to get to the polls. The confusing ballot is like the steep hill, and it should not matter that election officials picked the hill because they thought it would be a good place to vote without distractions. On the other hand, the plaintiff’s voting complaint may stem less from state action (as in the manual recount case where state officials count the votes) than from the capabilities of different voters. In Lassiter v. Northampton County Board of Elections,98 the Court held that fairly applied literacy tests are constitutional.99 This holding is of questionable value following cases like Harper100 and Kramer v. Union Free School District No. 15.101 But if Lassiter remains good law, it stands for the proposition that the state can condition the franchise on voters’ ability to follow instructions—thereby ensuring that only educated voters vote.102 I find this argument unpersuasive, and not only because I reject Lassiter as misconstruing the nature of voting as an exercise in efficient decisionmaking rather than an allocation of political power among co-equal citizens. In the butterfly ballot hypothetical, the state did not design the ballot in one county to “test” elderly citizens’ ability to vote. The ballot more likely tested their ability to see, and no one will claim blindness as a valid reason to deny the vote. Moreover, accepting the legitimacy of such a test, why conduct the test only in one county? Finally, the result of a literacy test as in Lassiter is to prevent or hinder illiterate voters from voting. The result of the butterfly ballot apparently is to cause voters to vote for candidates they do not prefer. Surely the state cannot have a legitimate, much 97. On the relationship of Bush v. Gore to Mobile v. Bolden, see infra Part IV.C.1. 98. 360 U.S. 45 (1959). 99. Id. at 53-54. Such tests are now banned by the Voting Rights Act. 42 U.S.C. § 1973aa (1994). 100. 383 U.S. 663 (1966). 101. 395 U.S. 621 (1969). On why Lassiter may not survive Kramer, see ISSACHAROFF ET AL., supra note 1, at 58. 102. See Lassiter, 360 U.S. at 51-52 (“The ability to read and write likewise has some relation to standards designed to promote intelligent use of the ballot. . . . [I]n our society where newspapers, periodicals, books, and other printed matter canvass and debate campaign issues, a State might conclude that only those who are literate should exercise the franchise.”). 398 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:377 less compelling, interest in that. On balance, this looks like a case to which the precedent of Bush v. Gore should apply. With that conclusion, a word here is in order about remedies. There may be a difference between a challenge to a voting procedure or election mechanism before an election takes place and a postelection challenge seeking to throw out the results of a vote or recount or to demand a revote. Even if each of these five hypothetical lawsuits presents violations of the Equal Protection Clause, the appropriate remedy may not be to void an election or the results of a recount or to require a revote.103 Courts may be uncomfortable with remedies that overturn elections. In the third hypothetical, for example, a court could rule that the recount violates equal protection but that Smith’s claim is barred by laches: she should have sought an injunction preventing the manual recounting of the votes under existing state law. In the actual butterfly ballot case, the trial court ruled that a revote in Palm Beach County alone would violate the Constitution’s requirement of a uniform election day for presidential electors.104 Thus, if Bush v. Gore has any precedential value at all, it may have such value primarily when used prospectively to change election practices. IV. THE BENEFITS, COSTS, AND LIMITS OF THE NEW EQUAL PROTECTION JURISPRUDENCE OF BUSH V. GORE In Part II, I set forth my suspicions that Bush v. Gore ultimately will have little precedential value. In Part III, I explored the precedential value the case likely would have if the Supreme Court took its holding seriously. In this final Part, I consider the benefits of Bush v. Gore’s ostensible extension of equal protection jurisprudence in elections to the third level of equality, the costs of the extension, and the implications of the extension for other, broader equal protection claims in elections. 103. On the variety of potential remedies, see ISSACHAROFF ET AL., supra note 1, ch. 12 (listing as potential remedies for defective elections: ordering a new election, enjoining an upcoming election, adjusting the vote totals, permanently enjoining a particular election practice, damages, and criminal prosecution). 104. The trial court held that a revote could not be ordered in a presidential election because it would violate, among other things, the Constitution’s provision of a uniform day for the choosing of presidential electors. See Order on Plaintiff’s Compl. for Declaratory, Injunctive, and Other Relief Arising from Plaintiffs’ Claims of Massive Voter Confusion Resulting from the Use of a “Butterfly” Type Ballot during the Election Held on Nov. 7, 2000, Fladell v. Elections Canvassing Comm’n of Fla., No. CL 00-10965 AB (Nov. 20, 2000), available at http://election2000.stanford.edu/fladell1120.pdf. The Florida Supreme Court did not reach the issue of remedy, finding that the ballot was in substantial compliance with Florida law. Fladell v. Palm Beach County Canvassing Bd., 772 So. 2d 1240, 1242 (Fla. 2000). 2001] EQUAL PROTECTION’S FUTURE IN ELECTIONS 399 A. Benefits The benefits of a precedent requiring scrupulous equality in the procedures and mechanics of elections are fairly obvious: such a precedent will increase resources used to conduct elections, so that at least twentieth century voting technology will be applied as we enter the twenty-first century. It will provide a means for those in poor, urban areas to have just as accurate a voting system as those used in wealthier areas. It will also likely ensure more reliable vote counting. Before the Florida debacle, state and local governments had little incentive to invest in better voting technology or to reconsider the fairness of their laws regulating the contesting of elections. In an era of tight government budgets, an argument to upgrade from punch card technology to optical scanner equipment had to compete with arguments to pay teachers more, to devote more money to crime prevention, or to return money to taxpayers. Bush v. Gore provides legal cover—if not a legal mandate—for expending resources to upgrade voting; prudent municipal attorneys would well advise their clients that failure to invest in better election processes will invite litigation. No one wants to be “the next Florida.” By increasing the salience of these issues, Bush v. Gore may have the salutary effect of causing governments to pay attention to these issues and devote resources toward solving voting problems, even if the case ultimately holds little or no precedential value. B. Costs Expanding political equality to the third level, as Bush v. Gore may have done, is a mixed blessing. Obviously, the costs associated with upgrading voting equipment, rewriting state and local election laws involving contested elections, and litigation over both types of changes will be considerable. One estimate to upgrade voting equipment ranged as high as $9 billion nationally.105 These are real costs, and obviously in a time of limited budgets such spending takes money away from teacher raises, better police protection, or tax reduction. But we can chalk up the $9 billion to the cost of having a democracy that takes seriously the mandate to ensure that all votes are counted and counted fairly. I focus here on three other concerns arising from extending equal protection jurisprudence to the nutsand-bolts of elections—concerns that go to whether extension to the third level of political equality necessarily furthers democratic values. 105. Alan C. Miller & Nick Anderson, America Waits: Voting Reforms Join Race for Funding, L.A. TIMES, Dec. 13, 2000, at A1; see also Jim Drinkard, Updating Voting Machines Could Take Nation a Decade, USA TODAY, Feb. 14, 2001, at 1 (noting that Congress “may pour up to $2.5 billion into upgrades” of voting equipment). 400 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:377 First, third-level claims provide more reasons, and in some cases a pretext, for courts to nullify election results. The courts’ further entry into the political thicket thus threatens both democracy and the legitimacy of courts, whose integrity may be questioned even when a court justifiably nullifies an election on equal protection grounds. In Bell v. Southwell,106 the Fifth Circuit courageously voided the results of a local election in which African-American voters were intimidated from voting in voting booths segregated by race and gender. The court voided the results even while recognizing the power to do so as “[d]rastic, if not staggering.”107 As correct as Bell was, court intervention should be used sparingly. Bush v. Gore is a dangerous precedent to the extent that it eases the way for federal court intervention in state and local elections over nuts-and-bolts disputes better left to local authorities. Second, third-level claims undermine federalism in a way that first- and second-level equal protection claims do not. Claims of local control over nuts-and-bolts voting mechanisms resonate more genuinely than claims of localities to deny the franchise to certain groups of individuals or to count votes unevenly. The Court rightly observed long ago that the right to vote is fundamental because it is “preservative of all [other] rights.”108 Politicians are less likely to be responsive to a group of citizens who cannot vote. Moreover, legislatures are constructed to respond to demands of a group of legislators in proportion to the group’s power in the legislature, rather than in proportion to the number of people the group of legislators represents; that is the essence of the vote dilution claim in Reynolds v. Sims.109 Thus, first- and second-level political equality claims allow courts to solve political market failures. A similar political market failure does not exist with respect to most nuts-and-bolts election issues.110 Consider again the issue of punch card systems versus optical scanners. Although it is true that punch card voters will be marginally more likely not to have their votes counted compared to those using optically scanned ballots, that difference will not neatly translate into a loss of political strength. Legislators elected from districts in which the punch card ballots are used will represent the same number of voters as those legislators from other districts, and politicians cannot ignore the wishes of those whose votes do not count because nobody knows who these people are. Thus, it is more likely that a nuts-and-bolts problem like the op106. 107. 108. 109. 110. A31. 376 F.2d 659 (5th Cir. 1967). Id. at 662. Yick Wo v. Hopkins, 118 U.S. 356, 370 (1886). 377 U.S. 533 (1964). See Pamela S. Karlan, The Court Casts Its Vote, N.Y. TIMES, Dec. 11, 2000, at 2001] EQUAL PROTECTION’S FUTURE IN ELECTIONS 401 tical scanning problem may be solved politically, if the voters and legislators in districts using the system make it a priority on the local or state level. The political process will not always work, however, to the extent the poor lack political power in the legislature generally—they may lack the power to get voting changes enacted as well. Not only is the case for court intervention weaker for third-level claims than for first- and second-level claims, but the state and local interests in local variations are also on stronger moral grounds. At least under political theories currently accepted by the Supreme Court, any state interest in deviating from roughly equipopulous districts is illegitimate, as is any state interest in denying the franchise to some group of citizen adult residents. On the other hand, the state may have a good reason unrelated to voting for at least some variations in the nuts-and-bolts of elections. Bush v. Gore is tantamount to a holding that the purchase of ambulances by a relatively poor county is less important than a move from punch cards to optical scanners. That may be a valid trade-off to make, but note that it is being made on the federal level for all jurisdictions by unelected federal judges. The third and final cost of accepting third-level equality claims is the disincentive the claims create for jurisdictions to experiment with new methods of voting. Oregon has adopted vote-by-mail, and jurisdictions are considering internet voting.111 How do these new methods get adopted in one jurisdiction alone, at least in presidential elections, following Bush v. Gore?112 California, for example, has wisely chosen to explore a move to internet voting slowly through a number of discrete steps with evaluations conducted after each step.113 As part of that experimentation, “touch-screen voting,” much like voting with an ATM screen, was used in Riverside County, California, as a pilot project in the 2000 general election.114 It seems far from frivolous to argue that, depending upon the error rates of such systems or other factors, either the voters of Riverside County or, alternatively, voters outside Riverside County have suffered discrimination under Bush v. Gore by the countywide experiment in an election for state and national office. Now perhaps the state has an important, indeed compelling, interest in conducting such tests. (Or perhaps not; could these tests be done in nonbinding elections or elections featuring only 111. See generally Richard L. Hasen, Introduction to Symposium, Internet Voting and Democracy, 34 LOY. L.A. L. REV. 979 (2001). 112. See supra Part III (discussing hypothetical 2). 113. CALIFORNIA INTERNET VOTING TASK FORCE, A REPORT ON THE FEASIBILITY OF INTERNET VOTING (January 2000), available at http://www.ss.ca.gov/executive/ivote/final_ report.htm. 114. Katharine Q. Seelye, California County Touches Future of Voting, N.Y. TIMES, Feb. 12, 2001, at A1. 402 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:377 candidates for local office?) But if the state knows it will face litigation for such experimentation, it will be wary of engaging in it. Thus, Bush v. Gore could have the unintended effect of freezing our voting mechanics at the current level of technology. That means that all voters may suffer as more accurate voting technology emerges. C. Beyond Third-Level Equality Claims Finally, it is worth thinking about the doctrinal implications of extending equal protection jurisprudence to the third level. It is unclear whether extension of equal protection jurisprudence to the third level differs meaningfully from arguments calling for greater political equality in terms of electoral structures and financing election campaigns. In other words, the Court in Bush v. Gore set the precedent of moving to a more intrusive and comprehensive view of political equality in terms of the nuts-and-bolts of elections without much discussion or defense of the move. Bush v. Gore can therefore serve to justify an analogous move by a future, more liberal Supreme Court toward a more intrusive and comprehensive view of political equality in other areas. My claim is not that such moves would flow from the holding of Bush v. Gore itself. The case’s holding is no doubt distinguishable from the equal protection claims discussed below. Rather, the applicable precedent here is the means by which Bush v. Gore adopted a new level of political equality. 1. Equality of “Electoral Structures” Consider first political equality in the means of aggregating votes. In Mobile v. Bolden,115 African-American residents of the city of Mobile, Alabama, brought a class action lawsuit challenging the constitutionality of the city’s at-large method of electing its three city commissioners under the Equal Protection Clause of the Fourteenth Amendment and under the Fifteenth Amendment.116 The evidence showed that African-American voters made up about one-third of the Mobile electorate, but given the persistence of severe voting along racial lines and the use of at-large voting rather than single-member districts, no African-American-preferred candidate had ever been elected commissioner or was likely to be elected commissioner in the foreseeable future.117 Had voting taken place using single-member districts rather than at-large, African-American voters would have had a better chance to elect a candidate of their choice or at least to exert greater political influence.118 115. 116. 117. 118. 446 U.S. 55 (1980). Id. at 58. Id. at 122 (Marhsall, J., dissenting). Id. at 105 n.3 (Marshall, J., dissenting). 2001] EQUAL PROTECTION’S FUTURE IN ELECTIONS 403 The Court rejected the argument that the at-large method violated the Equal Protection Clause of the Fourteenth Amendment.119 A four-Justice plurality stated that the plaintiffs’ claim failed because the plaintiffs lacked evidence that the electoral system was designed with a racially discriminatory purpose.120 Justice Blackmun concurred in the result on grounds that the relief afforded by the trial court “was not commensurate with the sound exercise of judicial discretion.”121 Justice Stevens concurred essentially on grounds that a contrary ruling would be impossible to administer.122 Three Justices dissented. Justice Marshall, joined by Justice Brennan, relied explicitly on Reynolds123 in arguing that the at-large system constituted a denial of equal protection: Reynolds v. Sims and its progeny focused solely on the discriminatory effects of malapportionment. They recognize that, when population figures for the representational districts of a legislature are not similar, the votes of citizens in larger districts do not carry as much weight in the legislature as do votes cast by citizens in smaller districts. The equal protection problem attacked by the “one person, one vote” principle is, then, one of vote dilution: under Reynolds, each citizen must have an “equally effective voice” in the election of representatives. In the present cases, the alleged vote dilution, though caused by the combined effects of the electoral structure and social and historical factors, rather than by unequal population distribution is analytically the same concept: the unjustified abridgement of a fundamental right. It follows, then, that a showing of discriminatory intent is just as unnecessary under the vote-dilution approach . . . as it is under our reapportionment cases.124 The plurality rejected Justice Marshall’s reliance on Reynolds, seeing Marshall’s position as an endorsement of proportional representation and thus “not the law. The Equal Protection Clause of the Fourteenth Amendment does not require proportional representation as an imperative of political organization.”125 119. Id. at 65-70. The Court also rejected the Fifteenth Amendment claim, but I focus here only on the Fourteenth Amendment claim, which is the claim in Bush v. Gore. 120. Id. 121. Id. at 80 (Blackmun, J., concurring). 122. Id. at 93 (Stevens, J., concurring) (“A contrary view ‘would spawn endless litigation concerning the multi-member district systems now widely employed in this country,’ and would entangle the judiciary in a voracious political thicket.”) (citation omitted). 123. 377 U.S. 533 (1964). 124. Bolden, 446 U.S. at 116-17 (Marshall, J., dissenting) (citations and footnotes omitted). 125. Id. at 75-76. 404 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:377 Regardless of whether Justice Marshall’s position should properly be characterized as an endorsement of proportional representation,126 it seems no more a stretch to extend the equal protection analysis of Reynolds to the means of aggregating votes (what Marshall refers to as “electoral structures”) than to the mechanics of voting. In other words, the principle of promoting political equality has no “natural” stopping point, even if we can draw distinctions among the cases. Congress essentially codified Justice Marshall’s position in Bolden through an amendment to section 2 of the Voting Rights Act in 1982.127 Thus, there has been no need for the Court to revisit the constitutional question. However, if Congress were to repeal the Voting Rights Act or the current Court majority were to hold it unconstitutional,128 the constitutional question could arise again. A future liberal Supreme Court could reverse Bolden, citing no more than Reynolds and Bush v. Gore’s holding that “[h]aving once granted the right to vote on equal terms, the State may not, by later arbitrary and disparate treatment, value one person’s vote over that of another.”129 The argument would be simply that at-large voting, much more so than counting undervotes but not overvotes, values one person’s vote over that of another. 2. Equality of Campaign Finance In Buckley v. Valeo,130 as mentioned above,131 the Court considered the constitutionality of a law limiting the amount individuals could spend supporting or opposing candidates for federal office. Plaintiffs argued that the law violated their rights under the First Amendment to freedom of speech and association, while the government defended the regulation in a number of ways.132 One argument the government raised was that the law was justified by an interest in promoting political equality.133 The Court rejected the argument: It is argued, however, that the ancillary governmental interest in equalizing the relative ability of individuals and groups to influ126. See Thomas M. Boyd & Stephen J. Markman, The 1982 Amendments to the Voting Rights Act: A Legislative History, 40 WASH. & LEE L. REV. 1347, 1356 n.56 (1983) (discussing whether revising section 2 of the Voting Rights Act to conform with Justice Marshall’s dissent would lead to “proportional representation”). 127. 42 U.S.C. § 1973(b) (1994). 128. This is more than an abstract possibility. See LOWENSTEIN & HASEN, supra note 1, at 339-40. 129. Bush v. Gore, 531 U.S. 98, 104-05 (2000). For a brief argument just along these lines, see Lani Guinier, A New Voting Rights Movement, N.Y. TIMES, Dec. 18, 2000, at A29. 130. 424 U.S. 1 (1976). 131. See supra notes 51-52 and accompanying text. 132. Buckley, 424 U.S. at 11. 133. Id. at 25-26. 2001] EQUAL PROTECTION’S FUTURE IN ELECTIONS 405 ence the outcome of elections serves to justify the limitation on express advocacy of the election or defeat of candidates imposed by [the statute’s] expenditure ceiling. But the concept that government may restrict the speech of some elements of our society in order to enhance the relative voice of others is wholly foreign to the First Amendment, which was designed “to secure the widest possible dissemination of information from diverse and antagonistic sources,” and “to assure unfettered interchange of ideas for the bringing about of political and social changes desired by the people.”134 Liberal scholars have sharply attacked the Buckley Court’s rejection of political equality as a compelling interest.135 One criticism has specifically tied the equality interest to equal protection analysis. In a pair of articles predating Bush v. Gore, Professor Jamin Raskin and attorney John Bonifaz argued that Reynolds requires that candidates for election receive equal public financing. They argue that “[i]n market societies where wealth is unevenly distributed yet crucial to the processes of election and governance, the inegalitarian logic of the economy undermines the egalitarian logic of one person, one vote democracy.”136 The authors explicitly argue that a constitutional requirement mandating equality in campaign finances follows from earlier Supreme Court equal protection precedents striking down “grandfather clauses, exclusionary white primaries, state poll taxes, restrictions on the suffrage rights of citizens in the armed services, unnecessarily long residency requirements, excessively high candidate filing fees, and malapportioned legislative districts that dilute the potency of the vote.”137 No doubt, Raskin and Bonifaz can now add Bush v. Gore to their list of precedents creating greater political equality in elections. If “the State may not . . . value one person’s vote over that of another”138 in how it counts votes, it similarly should not sanction the use of private wealth to influence the outcome of an election in a way that values one person’s vote over that of another. As with reversal of Bolden, the conservative Supreme Court in Bush v. Gore has set the precedent for a future liberal Supreme Court to embrace Raskin and Bonifaz’s novel equal protection analysis. 134. Id. at 48-49 (citations omitted). 135. See, e.g., JOHN RAWLS, POLITICAL LIBERALISM 362-63 (Columbia Univ. Press ed., 1993); CASS R. SUNSTEIN, DEMOCRACY AND THE PROBLEM OF FREE SPEECH 93-101 (The Free Press ed., 1993). 136. Raskin & Bonifaz, Constitutional Imperative, supra note 13, at 1162. 137. Raskin & Bonifaz, Wealth Primary, supra note 13, at 273-74 (citations omitted). For criticism of the authors’ analogy to Reynolds, see Bradley A. Smith, Money Talks: Speech, Corruption, Equality, and Campaign Finance, 86 GEO. L.J. 45, 84-88 (1997). 138. Bush v. Gore, 531 U.S. 98, 104-05 (2000). 406 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:377 V. CONCLUSION Whether a future Supreme Court should reverse course from either Bolden or Buckley is a large topic well beyond the scope of this paper. My point is only that the equality principle is difficult to cabin. Whether or not Reynolds, Harper, and Bush v. Gore were correct or incorrect decisions, they inevitably flow from the Justices’ views of how much (and what kinds of) equality the Constitution should mandate, and what is better left to state variation and the political processes. As times and Court personnel change, such views on equality, and therefore the law of equal protection in elections, will likely change as well. To the extent Bush v. Gore paves the way toward constitutional challenges of electoral structures and campaign finance reform, it may be a good development to at least some observers. It certainly would not be a development intended by at least some of the five Justices in the Bush v. Gore majority.139 That would just add to the list of ironies that the Florida controversy has wrought, and, perhaps for those who are disappointed by the Court in Bush v. Gore, create a sense of “rough justice” as well. 139. Then-Justice Rehnquist was in the Mobile plurality in 1980 that criticized Justice Marshall’s dissent. None of the other four members of the Bush v. Gore majority were on the Court at the time. Justice Thomas’ opinion concurring in the judgment in Holder v. Hall, 512 U.S. 874 (1994), joined by Justice Scalia, affirms that he would find Justice Marshall’s arguments to extend Reynolds unacceptable and against the Constitution. ThenJustice Rehnquist agreed with the majority per curiam opinion in Buckley v. Valeo, 424 U.S. 1 (1976), rejecting political equality as a rationale for campaign finance regulation. Justice Thomas, joined by Justice Scalia, has rejected any campaign finance regulation more onerous than disclosure in his dissenting opinion in Nixon v. Shrink Missouri Government PAC, 528 U.S. 377 (2000). Surely he would reject the equality rationale. Justices Scalia and Kennedy (joined by Justice O’Connor) also rejected the equality rationale for campaign finance reform in Austin v. Michigan Chamber of Commerce, 494 U.S. 652, 69293 (1990) (Scalia, J., dissenting); id. at 705 (Kennedy, J., dissenting). NEW WINE IN OLD BOTTLES: A COMMENT ON RICHARD HASEN’S AND RICHARD BRIFFAULT’S ESSAYS ON BUSH V. GORE HEATHER K. GERKEN* I. NEW WINE: BUSH V. GORE’S NOVEL EQUAL PROTECTION CLAIM ...................... II. OLD BOTTLES: THE COURT’S BAD HABITS IN VOTING-RIGHTS CASES ................ A. Judicial Agnosticism .................................................................................... B. Agnosticism’s Effects .................................................................................... 1. The Right is Defined in Abstract Terms................................................ 2. The Absence of Limiting Principles....................................................... 3. The Prevalence of Bright-Line Rules and Mechanical Proxies. ........... 4. The Court’s Failure to Come to Grips With the Normative Stakes of the Questions Before It ........................................................................... 409 413 414 415 415 416 417 421 The conventional wisdom in the wake of Bush v. Gore1 was that the decision represented a significant departure from prior equal protection jurisprudence, and the contributions by Richard Briffault and Rick Hasen to this symposium provide confirmation of that view.2 I want to make two points in response to their fine essays. First, it is a mistake to try to fit Bush v. Gore into existing equal protection frameworks. As I explain in Part I, Bush v. Gore is best understood as a new type of equal protection claim. On one reading, it addresses broad structural concerns rather than conventional individual harms. Second, I argue in Part II that the structural reading is probably unfounded.3 Bush v. Gore could, in theory, represent a sophisticated * Assistant Professor, Harvard Law School. I would like to thank Einer Elhauge, Richard Fallon, Lani Guinier, Ellen Katz, Spencer Overton, David Simon, William Stuntz, and Laurence Tribe for their thoughtful comments and suggestions. Thanks also to my research assistants, Felix Gilman and Sarah Pheasant. All mistakes are, of course, my own. Copyright © Heather K. Gerken. 1. 531 U.S. 98 (2000). 2. See Richard Briffault, Bush v. Gore as an Equal Protection Case, 29 FLA. ST. U. L. REV. 325 (2001); Richard L. Hasen, Bush v. Gore and the Future of Equal Protection Law in Elections, 29 FLA. ST. U. L. REV. 377 (2001). 3. There is, of course, a less generous reading of Bush v. Gore: it may be an exercise in naked partisanship. Plenty of academics have hashed out that question. See, e.g., Frank I. Michelman, Suspicion, or the New Prince, 68 U. CHI. L. REV. 679 (2001); Richard A. Posner, Bush v. Gore: Prolegomenon to an Assessment, 68 U. CHI. L. REV. 719 (2001); David A. Strauss, Bush v. Gore: What Were They Thinking?, 68 U. CHI. L. REV. 737 (2001); Cass R. Sunstein, Order Without Law, 68 U. CHI. L. REV. 757 (2001); John C. Yoo, In Defense of the Court’s Legitimacy, 68 U. CHI. L. REV. 775 (2001); Bruce Ackerman, The Court Packs Itself, AM. PROSPECT, Feb. 12, 2001, at 48; Ronald Dworkin, A Badly Flawed Election, N.Y. REV. OF BOOKS, Jan. 11, 2001, at 53, 53-55; Charles Fried, A Badly Flawed Election: An Exchange, N.Y. REV. OF BOOKS, Feb. 22, 2001, at 8, 8-10; Neal Kumar Katyal, Politics Over Principle, WASH. POST, Dec. 14, 2000, at A35; Jeffrey Rosen, Disgrace, NEW REPUBLIC, Dec. 25, 2000, at 18, 18-21. I have little to add to that discussion. Moreover, even if one thinks Bush v. Gore involves bad faith decisionmaking, it is worth considering the doctrinal package in which the Court chose to wrap its decision, for that choice itself tells us something about the judicial culture. See generally Richard H. Pildes, Democracy and Dis- 407 408 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:407 effort to conceptualize democratic principles in structural terms. But it is far more likely that the Court, in announcing a new type of equal protection claim, is simply reverting to one of its bad habits in voting-rights cases: decisionmaking unmoored from an explicit normative theory. One of the great oddities in the Supreme Court’s votingrights jurisprudence dating back to the Warren Court is that the Justices often disavow the notion that they are importing a particular theory of democracy into the decision.4 Their claim to agnosticism is, of course, implausible. And the Court’s self-conscious preference for avoiding any discussion of its normative premises has led to the type of decisionmaking we see in the Bush v. Gore per curiam: an opinion that articulates the injury in an abstract, formal manner; announces a legal rule with no easily discernible limits; defines equality in mechanical, quantitative terms; and fails to address the hard normative issues embedded in the questions it resolves.5 The Court has, in effect, poured new wine (the novel claim recognized in Bush v. Gore) into the old bottle of past jurisprudential habits.6 order, 68 U. CHI. L. REV. 695, 696 (2001) (placing Bush within the context of “[the] judicial culture, . . . the empirical assumptions, historical interpretations, and normative ideals of democracy that seem to inform and influence the current constitutional law of democracy”). 4. See, e.g., Holder v. Hall, 512 U.S. 874, 893 (1994); Lucas v. Forty-Fourth Gen. Assembly of Colo., 377 U.S. 713 (1964); Baker v. Carr, 369 U.S. 186, 300 (1962) (Frankfurter, J., dissenting). 5. Cass Sunstein has written extensively on the value of “incompletely theorized agreements,” and the benefits derived from the Court’s failure to explain and develop the broad theories underlying decisions rendered on a case-by-case basis. See, e.g., CASS R. SUNSTEIN, ONE CASE AT A TIME: JUDICIAL MINIMALISM ON THE SUPREME COURT (1999); Cass R. Sunstein, The Supreme Court, 1995 Term—Foreword: Leaving Things Undecided, 110 HARV. L. REV. 4 (1996). I share Richard Fallon’s view that “incompletely theorized agreements” represent “a second-best approach.” RICHARD H. FALLON, JR., IMPLEMENTING THE CONSTITUTION 106-07 (2001). Moreover, as I explain in greater detail elsewhere, even setting aside the benefits of transparency discussed by Fallon, supra, I believe that this strategy engenders more costs than benefits in the context of voting rights because of the difficult nature of the issues presented to the Court. In the voting context, the Court’s failure to articulate its normative premises leads it to mistake structural claims about the aggregation of votes for conventional individual rights, a mistake that leads to circular descriptions of the constitutional injury, doctrinal incoherence, and an inappropriate reliance on mechanical rules. See Heather K. Gerken, Lost in the Translation: Baker v. Carr and the Supreme Court’s Inability to Translate the Equality Norm in Voting Cases, 80 N.C. L. REV. (forthcoming 2002) (manuscript at 15-25, on file with author). 6. In this sense, I join Pam Karlan and Rick Pildes in thinking of Bush v. Gore as a continuation of the Court’s prior jurisprudence, albeit for different reasons. See Pamela S. Karlan, Nothing Personal: The Evolution of The Newest Equal Protection from Shaw v. Reno to Bush v. Gore, 79 N.C. L. REV. 1345 (2001) [hereinafter Karlan, Nothing Personal] (arguing that Bush v. Gore resembles the Shaw line of cases in that it demonstrates the Supreme Court’s confidence in its own abilities, its distrust of other institutions for resolving voting controversies, its solicitousness toward those who do not require judicial protection, and its willingness to suspend traditional standing requirements); Pamela S. Karlan, The Newest Equal Protection: Regressive Doctrine on a Changeable Court, in THE VOTE: BUSH V. GORE AND THE SUPREME COURT 77 (Richard Epstein & Cass Sustein eds., 2001) [hereinafter Karlan, The Newest Equal Protection] (same); Pildes, supra note 3, at 696-97 2001] NEW WINE IN OLD BOTTLES 409 I. NEW WINE: BUSH V. GORE’S NOVEL EQUAL PROTECTION CLAIM In trying to explain the new equal protection injury announced in Bush v. Gore, Hasen’s and Briffault’s papers neatly complement one another.7 Both struggle mightily to fit Bush v. Gore into conventional equal protection analysis, and both make interesting observations along the way. But it is the struggle itself that I find most intriguing, because it raises the question of whether we are making a mistake by trying to squeeze Bush v. Gore into the framework the courts have traditionally used to evaluate equal protection harms. Hasen and Briffault, of course, do not seek to answer that question here.8 They both take Bush v. Gore on its own terms. But the fact that neither of these able academics can offer a coherent theory for Bush v. Gore, let alone agree on its basic application, is worth noting. It suggests that we may need to discard the frameworks we have used in the past to understand voting claims. Specifically, the reason that Hasen and Briffault struggle so much here is that they are talking about a different category of harm. Because they analyze the case in conventional terms, they try to figure out why the conduct challenged in Bush v. Gore harms an individual or a group. But even if the Bush v. Gore majority suspected that Florida officials were reading ballots in a manner that undermined Bush’s chances of election, no one had solid proof of this bias. Because there was no evidence of a skew, intentional or otherwise, the right here is very different from traditional equal protection analysis—it is the right to equality in the abstract.9 Hasen’s and Briffault’s (arguing that Bush v. Gore reflects the Justices’ own views about democracy as strong or fragile, chaotic or robustly competitive). 7. The papers also complement each other in that both assess the implications of Bush v. Gore for the Court’s federalism theories. Hasen sees the Bush v. Gore claim as more intrusive on federalism values than prior doctrine because, in his view, the nuts and bolts of redistricting are more intuitively local concerns than what he terms the “claims of localities to deny the franchise to certain groups of individuals or to count votes unevenly.” Hasen, supra note 2, at 380. I disagree with him to some extent, but it may simply be a matter of how we frame the question. I understand those prior doctrines not as depriving the state of the right to disenfranchise or to count votes unfairly, but as depriving the state of the ability to choose how to structure its government and electoral system, which has also long been understood as an issue of local concern. Thus, in my view, the Baker and Shaw lines involve significant intrusion into local concerns. Briffault’s important contribution to this debate provides a further puzzle. He argues that the Court’s opposition to the use of a standard to count votes necessarily deprives local decisionmakers of the discretion one might think federalism would accord them. See Briffault, supra note 2, at 373-76. One leaves both pieces with a strong sense that “federalism” is so vaguely defined in the Court’s jurisprudence that it does not provide a meaningful principle for deciding cases. 8. Few things are more unforgivable than a person commenting on a paper who, in fact, talks about a prior question or fails to address the author on his own terms. All I can offer by way of apology is that academics usually commit this sin when the author has written something intelligent and persuasive, leaving little room for direct criticism. 9. I do not mean to say that Bush and his supporters had no means of asserting standing in this case. If one analogizes Bush v. Gore to Miranda, see infra text accompany- 410 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:407 search for the elements of a Washington v. Davis claim, intent and effect, is therefore futile.10 That is why we need a new framework. The one the two authors employ cannot possibly work here because it is not geared to this type of injury. So where do we look to find this new framework? One possible explanation for the harm in Bush v. Gore, and its departure from traditional equal protection analysis, is that Bush v. Gore is really a structural claim.11 It is a claim about how to order a well-functioning democracy, not a suit about individual rights. In this sense, Bush v. Gore may resemble other claims that do not fit easily into a conventional individual rights paradigm, like the right to an undiluted vote or the one-person, one-vote claims.12 To the extent that such claims implicate broad democratic structures, we should not be surprised that it is difficult to identify a conventional individual harm. If structural concerns are the source of Bush v. Gore, then we need to think of the claim differently. Perhaps we should think of the cause of action as something akin to a due process claim, as Briffault notes in passing and as others have argued here and elsewhere.13 Procedural due process claims may be structural in this sense; they address the way a system of adjudication is supposed to work rather than intentional injury or differences in substantive outcomes. That is why a violation of procedural due process can be established withing notes 18-19, then Bush may raise a claim on behalf of his supporters regarding the risk of hidden discrimination. Similarly, as Larry Tribe helpfully pointed out to me, Bush was the officially certified winner of Florida’s electoral votes at the time of the Supreme Court’s decision. Any decision that might call that position into question—even the order of a recount where Bush was as likely to emerge the winner as Gore—would represent a concrete injury to Bush and his supporters. Imagine, however, that this case had been brought before any candidate had been certified the winner. In this context, the underlying injury asserted by the voters—the right not to be subject to a recount process that involved no skew against any type of voter or candidate—is quite abstract when viewed against the traditional equal protection doctrine. It is in this conception of the harm that I discuss here. 10. 426 U.S. 229, 245-48 (1976). 11. See Karlan, The Newest Equal Protection, supra note 6, at 78 (concluding that Bush v. Gore involves a structural principle designed “to regulate the institutional arrangements within which politics is conducted”); see generally Samuel Issacharoff & Richard H. Pildes, Politics as Markets: Partisan Lockups of the Democratic Process, 50 STAN. L. REV. 643 (1998) (proposing a structural analysis of voting-rights cases). 12. I have briefly explored these questions elsewhere. See Heather K. Gerken, Understanding the Right to an Undiluted Vote, 114 HARV. L. REV. 1665, 1727 (2001) [hereinafter Undiluted Vote]; Gerken, supra note 5 (manuscript at 25-52). One key difference between the harm recognized in Bush v. Gore and the dilution injury is that in dilution cases, it is possible to identify a discrete group of individuals who have suffered a concrete injury. Gerken, Undiluted Vote, supra, at 1724-25. The same cannot be said of the Bush v. Gore injury, which seems to fall equally upon all voters. See infra text accompanying note 30. 13. See, e.g., Briffault, supra note 2, at 343; Richard A. Epstein, “In such Manner as the Legislature Thereof May Direct”: The Outcome in Bush v. Gore Defended, 68 U. CHI. L. REV. 613, 615 (2001); Peter M. Shane, Disappearing Democracy: How Bush v. Gore Undermined the Federal Right to Vote for Presidential Electors, 29 FLA. ST. U. L. REV. 535, 550-52 (2001). 2001] NEW WINE IN OLD BOTTLES 411 out proof that a different result would have obtained had the treatment been nonarbitrary.14 If we still need to identify a source of individual harm, then perhaps we should look to an expressive harm theory.15 It may be that the arbitrary and capricious treatment of ballots the Justices perceived in Florida conveys an improper message about the value of one’s vote. If so, we might still care about the differential treatment of like ballots even if the conduct does not stem from an improper motive or result in a skew. For example, we might think a healthy, well-functioning democracy demands that voters have confidence in the sanctity of the ballots they cast, which in turn requires that state officials accord adequate respect to those ballots by treating similarly situated ballots alike.16 In the words of the per curiam, the recount process violated equal protection because it was “not well calculated to sustain the confidence that all citizens must have in the outcome of elections.”17 Alternatively we could envision Bush v. Gore as a prophylactic rule, like Miranda,18 or as an effects test in discrimination cases. As14. See, e.g., Carey v. Piphus, 435 U.S. 247, 264 (1978). 15. See generally Elizabeth S. Anderson & Richard H. Pildes, Expressive Theories of Law: A General Restatement, 148 U. PA. L. REV. 1503 (2000); Deborah Hellman, The Expressive Dimension of Equal Protection, 85 MINN. L. REV. 1, 13-14, 34-35 (2000); Ellen D. Katz, Race and the Right to Vote After Rice v. Cayetano, 99 MICH. L. REV. 491, 512-17 (2000); Richard H. Pildes & Richard G. Niemi, Expressive Harms, “Bizarre Districts,” and Voting Rights: Evaluating Election-District Appearances After Shaw v. Reno, 92 MICH. L. REV. 483 (1993); Richard H. Pildes, Why Rights Are Not Trumps: Social Meanings, Expressive Harms, and Constitutionalism, 27 J. LEGAL STUD. 725 (1998). In a recent e-mail exchange, Larry Tribe raised yet a third possibility: that the constitutive aspects of voting might serve as the source of the harm. See E-mail from Larry Tribe, Professor of Law, Harvard Law School, to Ellen Katz, Assistant Professor of Law, University of Michigan Law School (May 26, 2001, 14:45:50 EST) (on file with author); see also Katz, supra, at 491; Frank I. Michelman, Conceptions of Democracy in American Constitutional Argument: Voting Rights, 41 FLA. L. REV. 443, 453-57 (1989). 16. I do not mean to suggest here that the ballots in question were similarly situated, just that the Court thought of them as such. After all, they were the products of quite different voting procedures and machines. See Laurence H. Tribe, eroG .v hsuB and Its Disguises: Freeing Bush v. Gore From Its Hall of Mirrors, 115 HARV. L. REV. 170, 177 (2001); see also Karlan, Nothing Personal, supra note 6, at 1364-65. Another bad habit of the Court’s jurisprudence is its failure to pay adequate attention to the factual context of its decisions. See Heather K. Gerken, Morgan Kousser’s Noble Dream, 99 MICH. L. REV. 1298 (2001) (reviewing J. MORGAN KOUSSER, COLORBLIND INSURANCE: MINORITY VOTING RIGHTS AND THE UNDOING OF THE SECOND RECONSTRUCTION (1999)). 17. Bush v. Gore, 531 U.S. 98, 109, (2000). 18. Miranda v. Arizona, 384 U.S. 436 (1966). A concern about the possibility that a broad recount standard would facilitate sub rosa discrimination has been raised by Einer Elhauge in a recent article, see Einer Elhauge, The Lessons of Florida 2000, 111 POL’Y REV. 15, 18-27 (2001-02), and in an amicus brief he and Charles Fried filed in Bush v. Gore, see Brief of the Florida House of Representatives and Florida Senate as Amici Curiae in Support of Neither Party and Seeking Reversal, Bush v. Gore, No. 00-949, 2000 WL 1868121. For an argument that Shaw can also be understood as a prophylactic rule, see Melissa L. Saunders, Reconsidering Shaw: The Miranda of Race-Conscious Districting, 109 YALE L.J. 1603, 1605-07 (2000). 412 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:407 suming that discrimination in the counting process is hard to detect, we might be especially concerned about the possibility of discrimination in the classic sense (an intentional effort to skew the count in favor of a particular candidate) when dealing with recounts governed by a flexible standard. After all, in contrast to rules adopted ex ante (or even variations in machinery that existed ex ante), at the point of a recount we may know what outcome will result from a more or less generous approach to determining the voter’s intent.19 Thus, we could read Bush v. Gore’s demand for uniformity as embedding within the structure of the electoral process a prophylactic protection against invidious discrimination. There is one final possibility to play with. We could read Bush v. Gore as a reflection of the Court’s efforts to vindicate a structural concern of a different sort: the value of “stability and measured change,” to borrow Justice O’Connor’s phrase.20 As Richard Pildes has recently observed, Bush v. Gore may reflect the Justices’ strongly held assumptions about democracy, whether they view the roughand-tumble political process as threatening chaos or healthy democratic competition.21 The Court has often expressed concern about destabilizing threats to the two-party system: blanket primaries,22 fusion candidates,23 party raiding,24 and the like. In Bush v. Gore, the threat to stability may have been polarization—the fear that the disorderly recount process was exacerbating political divisions. In this sense, the danger the Court perceived in Bush v. Gore may have resembled the “exacerbat[ion of] . . . racial bloc voting” the Court feared in Shaw v. Reno,25 the “divisive sore-loser candidacies” it sought to avert in Burdick v. Takushi,26 or the “destabilizing effects of 19. Such an intuition may also help explain our collective concern over whether the Florida Supreme Court had changed the rules of the game midstream, a concern that may have animated Chief Justice Rehnquist’s concurrence to Bush v. Gore as well. In offering this hypothetical justification for Bush v. Gore, I do not mean to make too much of the ex ante/ex post distinction, as the line is too often difficult to draw. Here, for example, at the time of the litigation both the Bush team and the Gore team were guessing as to which standard would most benefit their candidate, and neither guess appears to have been terribly accurate. See, e.g., Dan Keating & John Mintz, From Election Audit, Most Uncertainty; Miami Herald Review Shows Result Hinges on Standard Used in Recount, WASH. POST, Apr. 5, 2001, at A15. I am grateful to Larry Tribe for raising this point. 20. Davis v. Bandemer, 478 U.S. 109, 145 (1986) (O’Connor, J., concurring). I am indebted to Lani Guinier for suggesting this point. See also Lani Guinier & Gerald Torres, THE MINER’S CANARY: ENLISTING RACE, RESISTING POWER, TRANSFORMING DEMOCRACY ch. 6, at 18-19 (forthcoming Feb. 2002) (manuscript on file with author). 21. Pildes, supra note 3, at 704-05. 22. See, e.g., Cal. Democratic Party v. Jones, 530 U.S. 567, 577-78 (2000). 23. See, e.g., Timmons v Twin Cities Area New Party, 520 U.S. 351, 366-67 (1997). 24. See, e.g., Burdick v. Takushi, 504 U.S. 428, 439 (1992). 25. 509 U.S. 630, 648 (1993). 26. 504 U.S. at 439. But see Eu v. S.F. County Democratic Cent. Comm., 489 U.S. 214, 227 (1989) (no state interest in avoiding “intraparty friction”). 2001] NEW WINE IN OLD BOTTLES 413 party-splintering and excessive factionalism” it tried to avoid in Timmons v. Twin Cities Area New Party.27 The problem for the members of the Bush v. Gore majority, however, is that there was no easy doctrinal path for vindicating this concern. In cases like Burdick and Timmons, it has been quite easy for the Court to express its preference for stability-promoting democratic structures through the balancing test the Court uses to evaluate First Amendment claims; the Court could simply place a judicial thumb on the scale in favor of the state’s interest in stability.28 Bush v. Gore, however, presented the Court with the same dilemma it faced in Shaw: it was the actions of the state itself that threatened to factionalize the electorate.29 This may help explain the abstractness of the injury articulated in both Bush v. Gore and Shaw.30 Because the legal claim at stake did not allow for the inclusion of the value of stability in the doctrinal equation, this concern may have seeped into the definition of the injury itself. On this view, the polarizing and chaotic recount process was the source of the harm in Bush. Because such a harm is what Justice Frankfurter would call “a wrong suffered by [the state] as a polity,”31 it is not surprising that the injury seems abstract and amorphous when viewed through the traditional lens of Article III standing. II. OLD BOTTLES: THE COURT’S BAD HABITS IN VOTING-RIGHTS CASES There is a second possible explanation for Bush v. Gore, one that is a more likely candidate for explaining the decision. While it may be tempting to think of the Bush v. Gore claim as a structural harm, the Court has not yet explicitly accepted scholars’ invitation to think of democracy in structural terms, and the opinion itself offers no hint that it is taking a structural view.32 Thus, while Bush v. Gore may 27. 520 U.S. at 367. 28. See, e.g., id.; Burdick, 504 U.S. at 439. 29. In Shaw, the Department of Justice may also have been an instability-inducing culprit in the Court’s eyes, as it often intervened in the redistricting process during the 1990s to promote its policy of “maximization.” This may explain why the Department of Justice was so frequently and harshly rebuked by the Court. See, e.g., Shaw v. Hunt, 517 U.S. 899, 912-13 (1996); Miller v. Johnson, 515 U.S. 900, 907-08, 921-23 (1995). 30. See Karlan, Nothing Personal, supra note 6, at 1349-53, 1357-63 (arguing that Shaw and Bush v. Gore both involve attenuated injuries that are not sufficiently concrete to confer standing); Karlan, The Newest Equal Protection, supra note 6, at 20-21. 31. Colegrove v. Green, 328 U.S. 549, 552 (1946). 32. Compare Bush v. Gore, 531 U.S. 98, 104-05 (2000) (discussing injury in terms of individual right to vote), with Nixon v. Shrink Mo. Gov’t, 528 U.S. 377, 401 (2000) (Breyer, J., concurring) (suggesting that campaign finance doctrine should be recast as a principle “to protect the integrity of the electoral process—the means through which a free society democratically translates political speech into concrete governmental action”), and Laurence H. Tribe, Saenz Sans Prophecy: Does the Privileges or Immunities Revival Portend the Future—Or Reveal the Structure of the Present?, 113 HARV. L. REV. 110 (1999) (discussing 414 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:407 have announced a new type of injury, it is best viewed as exemplifying a common problem in voting cases dating back to the days of the Warren Court: formalist reasoning unmoored from an explicit normative theory. The same tendency emerges in cases involving one person, one vote; vote dilution; and Shaw. This Part explores the weaknesses that result from the Court’s avoidance of normative debate and argues that Bush v. Gore provides a good example of these problems. This Part concludes by speculating about why the Supreme Court is so uncomfortable with making normative pronouncements about democracy. A. Judicial Agnosticism Members of every generation of the Supreme Court’s Justices have claimed that they have no theory about the way democracy should work.33 It is an odd claim for many reasons. First, it seems strange that a group constantly making rules about how the game is played should admit that they have no view on why we play it and who should win. Second, the claim of agnosticism is implausible. Whenever the Court inserts itself into the democratic process, it is making a judgment about how that process should work. One person, one vote presumably embodies a theory about majoritarianism;34 vote dilution doctrine tells us something about the power that should be accorded to electoral minorities;35 the right-to-vote cases incorporate a judgment about representation, participation, and community membership.36 The Supreme Court similarly makes a decision about democratic values when it declines to act. The political question doctrine is a theory about how our constitutional system should work;37 Justice structural approach of federalism cases and relating to questions of individual rights), and Ernest A. Young, Alden v. Maine and the Jurisprudence of Structure, 41 WM. & MARY L. REV. 1601 (2000) (documenting and analyzing the structural approach adopted by the Court in its recent federalism decisions). 33. See, e.g., Holder v. Hall, 512 U.S. 874, 893 (1994) (Thomas, J., concurring) (arguing that courts should not recognize a harm that requires them to “resort to political theory [in order to] determine which electoral systems provide the ‘fairest’ levels of representation or the most ‘effective’ or ‘undiluted’ vote[ ]”); Lucas v. Forty-Fourth Gen. Assembly of Colo., 377 U.S. 713, 748 (1964) (Stewart, J., dissenting) (criticizing the majority opinion because “it imports and forever freezes one theory of political thought into our Constitution”); Baker v. Carr, 369 U.S. 186, 300 (1962) (Frankfurter, J., dissenting) (warning against courts “choos[ing] among . . . competing theories of political philosophy”); see also Einer Elhauge, Are Term Limits Undemocratic?, 64 U. CHI. L. REV. 83 (1997) (critiquing Court for failure to consider democratic theory in term limits decision). 34. See, e.g., Reynolds v. Sims, 377 U.S. 533, 545-56 (1964). 35. See Lani Guinier, [E]racing Democracy: The Voting Rights Cases, 108 HARV. L. REV. 109, 124 (1994). 36. See Michelman, supra note 15, at 443. 37. See Samuel Issacharoff, Political Judgments, 68 U. CHI. L. REV. 637, 650-55 (2001). 2001] NEW WINE IN OLD BOTTLES 415 Thomas’s insistence that dilution claims are noncognizable implicates a particular understanding about the aggregative aspects of voting.38 Even when the Court does not self-consciously eschew normative theory, it often fails to articulate the normative premises of its decisions or offers them cloaked in such vague generalities as to be indecipherable. In Bush v. Gore, for example, the per curiam opinion stated that one source of the right at issue “lies in the equal weight accorded to each vote and the equal dignity owed to each voter,”39 a phrase highly reminiscent of the Warren Court’s suggestion that the mere failure to achieve exact population equality in districts somehow renders each person “that much less a citizen.”40 While the words have a lovely ring to them, it is hard to describe the injuries in question, let alone explain why those injuries rise to constitutional status. The same is true of the early vote-dilution cases handed down by the Burger Court.41 While these decisions are often filled with stirring language about the right to vote, it is difficult to identify their precise doctrinal or normative parameters. B. Agnosticism’s Effects The problem is not just that the Court claims to be agnostic while it continues to worship at various altars of democratic theory. The problem is that the Court’s purported agnosticism allows it to indulge in a number of bad habits in voting cases, all of which are readily identifiable in Bush v. Gore. 1. The Right is Defined in Abstract Terms First, when the Court fails to articulate the normative premises of its decision, it tends to describe the injury in abstract terms.42 Consider, for example, the one person, one vote cases. These cases were originally fashioned as equal protection claims based on the assump- 38. See Guinier, supra note 35, at 122-23. 39. Bush v. Gore, 531 U.S 98, 104 (2000). 40. Reynolds v. Sims, 377 U.S. 533, 567 (1964). 41. Compare Reynolds, 377 U.S. at 533, and Baker v. Carr, 369 U.S. 186 (1962), with White v. Regester, 412 U.S. 755 (1973), and Whitcomb v. Chavis, 403 U.S. 124 (1971). 42. I should emphasize that when I argue that the right is defined in abstract terms, I do not mean to suggest that the doctrine itself lacks concrete, real-world consequences. To the contrary, as Richard Fallon has reminded me, one of the odd things about Bush and many other voting cases is that the Court relies on an abstractly defined right to justify inserting itself into the nitty-gritty details of redistricting. See Gerken, supra note 5 (manuscript at 2). What is missing in these cases—what makes the right seem abstract despite its real-world effects—is a principle for translating the broad equality norms on which the Court relies into a sensible definition of the right it is vindicating. For a detailed analysis of the problem of translation theories in voting cases, see id. 416 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:407 tion that rural voters had different interests than urban residents.43 And the Court’s early articulation of the one person, one vote principle might have developed into a sufficiently robust theory of democratic representation to take that truth into account.44 Instead, the Court simply stopped talking about which voters were affected by the skewed districting system. It similarly abandoned any effort to explain why the disparities in question demanded judicial intervention or to offer a theory of representation that required equally populated districts. Without a normative theory to define the parameters of the constitutional harm, the Court’s description of the injury became circular: population deviations cause an injury because they depart from the principle of one person, one vote. In short, the description of the harm was transformed into its normative justification.45 The same seems true of Bush v. Gore. Without a normative theory for why minor variations in recount efforts matter, it is difficult to say anything meaningful about the injury. For example, if Bush v. Gore had invoked a theory of aggregation, we could define the injury as the arbitrary treatment of votes that affects the election outcome. We would know to look for the skew that Briffault and Hasen search for in the opinion, and we would recognize that randomly distributed errors are constitutionally acceptable. If the Court had instead invoked an expressive harm theory, we would have to decide whether human error matters more or less than machine error for purposes of measuring the expressive injury. That analysis would, in turn, help us decide whether minor variations in a recount process are acceptable when remedying more significant disparities in machine counts. If the Court had adopted a prophylactic rule, we would know that variations in machinery are not constitutionally problematic, but that uniformity would be demanded of any ex post decisionmaking. As the opinion stands, however, we cannot describe the right in sufficiently concrete terms to resolve these questions. 2. The Absence of Limiting Principles This brings me to a second parallel between Bush v. Gore and past voting-rights jurisprudence. In the absence of a theory like one of those identified in Part I, it is very difficult to figure out the limiting principle for an abstractly defined right like the one articulated in 43. See Baker, 369 U.S. at 273 (Frankfurter, J., dissenting) (quoting complaint as challenging discrimination against a “geographical class of persons”). 44. See Reynolds, 377 U.S. at 563 n.40, 565-66 (asserting that the Constitution mandated the “full and effective participation” of citizens within the democratic process, guaranteed “fair and effective representation” to them, and forbade efforts to undermine citizens’ voting power “by any method or means”); Wesberry v. Sanders, 376 U.S. 1, 14 (1964) (describing one person, one vote principle as fostering equal representation). 45. For a more detailed analysis of these problems, see Gerken, supra note 5. 2001] NEW WINE IN OLD BOTTLES 417 Bush v. Gore, as Hasen’s and Briffault’s papers demonstrate.46 Under traditional equal protection analysis, it is hard to understand why we should care about a minute percentage of ballots being treated differently as long as there is not an outcome-affecting skew. But the odd result in Bush v. Gore seems a natural outgrowth of the Court’s vague and abstract approach. Consider again what happened in the one person, one vote cases.47 On one view, population equality represents a means to an important end, a well-functioning majoritarian system. But without an explanation as to why equality in population was necessary to achieve this important democratic value or whether that principle should always trump others, population equality became an end unto itself. Because equal population became the definition of the harm rather than the means to measure or prevent it, no limiting principle existed for discerning when variations from population equality would be acceptable. At least at the federal level, absolute numerical equality became the standard for measuring the constitutional harm in these cases, and it became difficult to argue that the one-person, one-vote rule should be suspended when necessary to achieve the broader democratic aims that (one presumes) the rule was originally designed to serve. One could not, for instance, argue that some population disparities should be tolerated in order to implement other values that promoted the same goal of a well-functioning majoritarian system (for example, keeping communities of interest intact in order to facilitate better representation, or providing a forum for electoral minorities to obtain an adequate voice).48 The Court’s failure to articulate its normative premises in the one person, one vote cases has thus led to the type of formalism we do not often associate with the Warren Court;49 it has become a jurisprudence where the rule is all that matters.50 3. The Prevalence of Bright-Line Rules and Mechanical Proxies The third problem that arises from the Court’s reluctance to articulate its normative premises is a marked trend toward using bright-line rules and mechanical proxies. Such tests can, of course, provide an efficient and sensible means to achieve a broader, norma46. See Briffault, supra note 2; Hasen, supra note 2. 47. See, e.g., Reynolds, 377 U.S. at 533; Wesberry, 376 U.S. at 1; Baker, 369 U.S. at 186. 48. See Lucas v. Forty-Fourth Gen. Assembly of Colo., 377 U.S. 713 (1964). 49. See Morton J. Horwitz, The Warren Court: Rediscovering the Link Between Law and Culture, 55 U. CHI. L. REV. 450, 456 (1988) (praising the Warren Court for avoiding formalist judging). 50. For a more detailed analysis of these problems, see Gerken, supra note 5 (manuscript at 21-25). 418 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:407 tive agenda. The problem in the voting context, however, is that too often courts lose track of the normative foundations of the rules and proxies they have adopted. Again, the test becomes all that matters, and courts rigidly, often unthinkingly, apply it in contexts where it does not fit. As I have discussed elsewhere,51 section 2 of the Voting Rights Act52 provides a good example of this problem. Section 2 prohibits vote dilution by preventing states from taking advantage of racially polarized voting among whites and racial minorities by drawing district lines so as to deprive minority voters of a fair share of political power. The qualitative approach to assessing dilution claims offered by the Supreme Court53 and endorsed by Congress in 198254 all but demanded that courts make some normative judgments. To assess a dilution claim, they had to examine the dynamics of the election process and assess the quality of representation minority group members received. The courts, however, have gradually moved away from this qualitative analysis toward a more rigid, quantitative approach for assessing dilution claims. To begin, in 1986 the Supreme Court adopted the three Gingles preconditions,55 mechanical proxies for assessing whether racial minorities’ potential voting strength has been undermined. Many thought that the Supreme Court’s decision in Johnson 51. Gerken, supra note 16, at 1303-04. 52. 42 U.S.C. § 1973 (1994 & Supp. V 1999). 53. See White v. Regester, 412 U.S. 755 (1973). 54. Congress endorsed a number of qualitative factors for assessing vote dilution in the Senate report accompanying the 1982 amendments. See S. REP. NO. 97-417, at 23 (1982), reprinted in 1982 U.S.C.C.A.N. (96 Stat. 131) 177, 200-01. Termed “the Senate factors,” they were pulled from a variety of dilution cases arising prior to 1982. See, e.g., White, 412 U.S. at 755; Zimmer v. McKeithen, 485 F.2d 1297 (5th Cir. 1973), aff’d sub nom. E. Carroll Parish Sch. Bd. v. Marshall, 424 U.S. 636 (1976) (per curiam). The factors include: the history of voting-related discrimination in the State or political subdivision; the extent to which voting in the elections of the State or political subdivision is racially polarized; the extent to which the State or political subdivision has used voting practices or procedures that tend to enhance the opportunity for discrimination against the minority group, such as unusually large election districts, majority vote requirements, and prohibitions against bullet voting; the exclusion of members of the minority group from candidate slating processes; the extent to which minority group members bear the effects of past discrimination in areas such as education, employment, and health, which hinder their ability to participate effectively in the political process; the use of overt or subtle racial appeals in political campaigns; and the extent to which members of the minority group have been elected to public office in the jurisdiction. Thornburg v. Gingles, 478 U.S. 30, 44-45 (1986). 55. According to Gingles, plaintiffs must prove that (1) the state could have drawn an additional, compact majority-minority district but failed to do so; (2) the minority group is politically “cohesive”—that is, its members vote in a similar fashion; and (3) the white electorate votes as a bloc, thus enabling whites usually to defeat the minority group’s preferred candidates at the polls. 478 U.S. at 50-51. 2001] NEW WINE IN OLD BOTTLES 419 v. DeGrandy56 would revive the role of qualitative, contextual analysis in the dilution context. Although the Court specifically reminded courts to pay attention to the qualitative concerns endorsed by Congress in 1982,57 it offered a new standard for evaluating dilution—the proportionality standard—that was equally amenable to unthinking, mechanical implementation. Proportionality requires that members of a racial group have a chance to exercise electoral control in a number of districts roughly proportional to their share of the population.58 It is an easily applied mathematical standard that requires a court to count the number of districts group members are capable of controlling on election day.59 Courts have quickly seized upon proportionality as the preeminent measure of vote dilution60 and often apply it in a rigid way. In Barnett v. City of Chicago,61 for example, the Seventh Circuit assumed that section 2 demanded precise mathematical equality for all racial groups and hinted that proportionality may even serve as a ceiling upon the number of majority-minority districts a locality can create.62 As the Gingles test and DeGrandy’s proportionality test take on a momentum of their own, as these numerical proxies become equated with democratic equality rather than measures of it, we can discern the potential costs to this approach. To begin, in adjudicating section 2 cases, courts may end up neglecting other important aspects of our democracy. For example, if all that matters is proportionality, courts will think that dilution claims deal solely with what happens on election day.63 They will thus ignore other important aspects of the de56. 512 U.S. 997 (1994). 57. Id. at 1011-12. 58. It is worth emphasizing that the proportionality standard does not suggest that only African Americans can represent African Americans and only Latinos can represent Latinos, nor does it guarantee a win to minority voters whether or not they show up to vote. To the contrary, in DeGrandy, the Court explicitly rejected the view that courts should look to election outcomes, specifically, how many minority legislators are elected in a district, in order to assess whether the districting scheme is “fair.” Id. at 1014 n.11. The Court was unwilling to assume that racial minorities can be represented only by other racial minorities, and it similarly refused to grant racial minorities a guaranteed right to electoral success. Id. Thus, DeGrandy held that fairness should be measured against the potential voting power of a minority group, that is, whether group members are capable of electing a candidate of choice (whatever the race of that candidate) if they choose to do so. In the words of the Court, “the ultimate right of § 2 is equality of opportunity, not a guarantee of electoral success for minority-preferred candidates of whatever race.” Id. 59. Id. at 1013-14 (announcing the proportionality standard). 60. See, e.g., African Am. Voting Rights Legal Def. Fund, Inc. v. Villa, 54 F.3d 1345, 1356 (8th Cir. 1995); Rural W. Tenn. African-Am. Affairs Council v. McWherter, 877 F. Supp. 1096, 1100-03 (W.D. Tenn. 1995), aff’d sub nom. Rural W. Tenn. African-Am. Affairs Council v. Sundquist, 516 U.S. 801 (1995). 61. 141 F.3d 699 (7th Cir. 1998). 62. Id. at 703-05. 63. See Kathryn Abrams, Relationships of Representation in Voting Rights Act Jurisprudence, 71 TEX. L. REV. 1409, 1415 (1993); see also Kathryn Abrams, “Raising Politics Up”: Minority Political Participation and Section 2 of the Voting Rights Act, 63 N.Y.U. L. 420 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:407 mocratic process (for example, the amount of political power minority groups enjoy once they have elected a candidate of choice to the legislature or the quality of representation they receive).64 Courts will also be far less open to other types of voting concerns, for example, influence-district claims65 or other models of representation.66 After all, the proportionality standard, standing alone, offers no means for assessing whether the interests of racial minorities are better served by majority-minority districts that concentrate their voting power or districting plans that spread their influence across many districts but do not grant them political independence in any single district. That judgment necessarily requires a normative theory for identifying what constitutes the interest of a racial minority and a theory of representation.67 Further, courts may apply rough proxies and mechanical tests even when the normative theory behind those rules would counsel a different result.68 The question is whether Bush v. Gore will continue this trend.69 Initially, for the reasons Hasen provides in his analysis of the costs of extending Bush v. Gore, I had thought that the Court would never pursue this line of reasoning because it would involve significant intrusion upon state decisionmakers carrying out traditional state duties. Now I am not so sure.70 It may be that judicial intrusion into REV. 449, 455-56 (1988); Pamela S. Karlan, Maps and Misreadings: The Role of Geographic Compactness in Racial Vote Dilution Litigation, 24 HARV. C.R.-C.L. L. REV. 173, 186-87, 198 (1989); Judith Reed, Sense and Nonsense: Standing in the Racial Districting Cases as a Window on the Supreme Court’s View of the Right to Vote, 4 MICH. J. RACE & L. 389, 441, 450 (1999). Pamela Karlan has argued that the Supreme Court has adopted an unduly narrow definition of equality in the Shaw case as well, focusing on the “claims of individual voters to the exclusion of claims about race-conscious districting’s contributions to the achievement of effective political equality for minority communities.” Karlan, Nothing Personal, supra note 6, at 1364. 64. See, e.g., Presley v. Etowah County Comm’n, 502 U.S. 491, 510 (1992) (rejecting third-generation governance claim regarding the allocation of political power within local governing body). 65. See, e.g., Voinovich v. Quilter, 507 U.S. 146, 158 (1993) (rejecting influence-district claim). 66. See, e.g., Cousin v. Sundquist, 145 F.3d 818, 822 (6th Cir. 1998) (declining to adopt a cumulative voting scheme as remedy for vote dilution). 67. Gerken, supra note 16, at 1312-13. 68. The same problems seem to plague the Court’s jurisprudence under section 5, where the retrogression standard has been applied in a similarly rigid fashion. See, e.g., Reno v. Bossier Parish Sch. Bd., 528 U.S. 320 (2000); Reno v. Bossier Parish Sch. Bd., 520 U.S. 471 (1997). 69. To be fair, I suppose that if Bush v. Gore is indeed intended as a prophylactic rule guarding against intentional discrimination, there is a stronger reason for the bright-line rule to be given such prominence. A prophylactic rule remains a means to an end (deterring discrimination), but it cannot function effectively as such if judges constantly return to its normative underpinnings and engage in a case-by-case contextual analysis to root out discrimination; they must apply the rule in something akin to the unthinking fashion I have criticized above. 70. My confidence has been undermined in part because of the questions both pieces raised about whether the Court is serious about federalism issues in this context. If Shaw 2001] NEW WINE IN OLD BOTTLES 421 state affairs is less likely to deter the Court than the absence of easily applied mechanical standards. After all, the latter allow the Court to think it is avoiding, often incorrectly, assessments about individual substantive preferences or broad structural judgments about representation and the democratic process. Thus, it may be that the Court was willing to restructure virtually every state redistricting scheme on equal protection grounds precisely because the one person, one vote cases were eventually reduced to what Justice Stewart termed “sixth-grade arithmetic.”71 We may find in the future that courts are more willing to strike down state redistricting plans if section 2 liability becomes nothing more than a mechanical application of Gingles and a rigid adherence to proportionality. If this hypothesis proves accurate, then the Bush v. Gore injury is just the kind of claim that the Court is likely to develop. The injury is easy to quantify, and it does not appear (at least superficially) to require endorsement of a particular theory of democracy or representation. Absolute equality among counting mechanisms will become an end unto itself, and we will never have a theory explaining why this type of democratic equality matters, save a few Kennedyesque phrases about equal treatment. That is also why I think that Hasen’s cautious hope that Bush v. Gore might lead to development of new principles in the context of campaign finance and vote aggregation rules is misplaced.72 For the Court to venture down that road, it would have to adopt a thicker normative theory about representation and the democratic process, and this Court seems to lack the judicial imagination to do so. 4. The Court’s Failure to Come to Grips With the Normative Stakes of the Questions Before It This leads me to my fourth and final concern about the Court’s agnosticism regarding democratic theory: the Court should not avoid these inquiries. The reasons for this trend are easy to explain and sympathize with—these questions are hard. For example, qualitative assessments of dilution claims would require courts to make judgments about the interests of minority voters, the quality of representation they receive, and the allocation of political power among groups. The one-person, one-vote cases demand an assessment of what constitutes effective representation and how to balance the dictates of majoritarianism against the interests of distinct minorities. It is understandable that judges prefer to latch on to a mechanical and the one-person, one-vote cases provide any indication of this Court’s proclivities, the level of intrusion is unlikely to bother the Court. 71. Lucas v. Forty-Fourth Gen. Assembly of Colo., 377 U.S. 713, 750 (1964). 72. Hasen, supra note 2, 402-05. 422 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:407 proxy or a thinly defined injury and thereby avoid, or more accurately, think they are avoiding, these difficult judgments.73 We know, however, that the Court is necessarily making such judgments and it should therefore face them squarely in doing so. Consider, for example, Hunt v. Cromartie,74 the Court’s latest iteration of Shaw. In the early days of the Shaw doctrine, Shaw was all norms and no empirics.75 The doctrine centered upon a normative assessment of the role race should play in politics, which led to a spirited debate among the Justices that filled hundreds of pages in the U.S. Reports. It is striking how easily the Court was able to set aside those normative debates in Cromartie. Both the majority and the dissent are formal and mechanical. Each takes an agreed-upon set of facts, applies the same standard, and reaches different conclusions.76 And both opinions are bereft of any sense of the broader issues at stake. They debate whether registration rates accurately reflect political affiliation, whether a given precinct contained more Democrats than African Americans, and whether particular legislative statements signaled a racial motive.77 But the two opinions do not wrestle with the normative questions embedded in the controversy. This is the jurisprudence of the technocrat: mechanical, seemingly neutral adjudication that conceals, but does not eliminate, the normative theories that necessarily undergird the decision. The same may prove true of Bush v. Gore. The Court seems to be vindicating a principle of equality, but it is equality of an extraordinarily narrow sort. In mandating the equal treatment of ballots that appear identical, the Court simply ignores the deep, systemic inequalities that plague our democratic process. For example, as Spencer Overton discusses in this symposium, the poor and racial minorities in Florida may have been less likely to have their ballots counted under any recount standard due to disparities in electoral resources;78 thus, as Larry Tribe has noted, straining ballots “through a uniformly metered sieve will [still] yield an unequal stream of results.”79 The Court’s failure to wrestle with these questions—what does equality mean, and how far should we go to attain 73. For a more in-depth analysis of the reasons behind the Court’s agnosticism, see Gerken, supra note 5 (manuscript at 25-52). 74. 121 S. Ct. 1452 (2001). 75. See Gerken, supra note 16, at 1317-18. 76. Cromartie, 121 S. Ct. at 1466, 1475. 77. Id. at 1459-65, 1474-75. 78. Spencer Overton, A Place at the Table: Bush v. Gore Through the Lens of Race, 29 FLA. ST. U. L. REV. 469, 469-71 (2001). Recently released studies by various media consortia reveal that this issue is more complex than previously suspected. See, e.g., Ford Fessenden & John Broder, Examining the Vote, N.Y. TIMES, Nov. 12, 2001, at A1. 79. Tribe, supra note 16, at 256. 2001] NEW WINE IN OLD BOTTLES 423 it when the twin problems of race and poverty permeate our democratic structures?—gives an unwarranted patina of legitimacy to the election system. Nor does the Court address the normative questions necessarily implicated by the more limited injury it recognizes. For example, if the injury at stake is an expressive harm, how do we decide whether such a harm exists? Should the Court consult its own intuitions? Public opinion polls?80 Similarly, how would the Court respond to Richard Briffault’s argument that the source of the injury in Bush v. Gore, the grant of discretion to local decisionmakers, is federalism itself?81 How should we weigh the concrete costs of judicial intrusion Hasen has documented against the equal protection values the Court seeks to vindicate?82 What do we make of the contested and contestable assumptions about the locus of democratic decisionmaking Spencer Overton has identified in the Court’s preference for a rule over a standard?83 I do not mean to suggest that every court must engage in normative hand-wringing each time it renders an opinion or that we should abandon bright-line rules. Nonetheless, we should expect, or at least hope for, some recognition of the normative stakes in the opinions judges render and a concomitant willingness to reexamine those normative premises where appropriate. That is especially true where, as here, the Court is announcing a newly minted voting claim in a highly charged political environment. If the Court decides to recognize a right that does not fit easily within conventional equal protection doctrine, it should provide a normative map so we can tell where we are heading within this political thicket. The unfortunate fact is that explicit normative engagement has not always been evident in voting-rights decisions dating back to the Warren Court, and there is little hope that it will be forthcoming should Bush v. Gore spawn progeny of its own. 80. If so, then the problem of butterfly ballots and overseas voters should have concerned the Court more than the claim Bush raised. As my British research assistant observed, people seemed far more upset about “elderly Holocaust survivors voting for Buchanan” or “our boys in uniform, fighting for our rights on foreign soil, being denied the right to vote” than the differences in the standards used to count ballots. 81. Briffault, supra note 2, at 375. 82. Hasen, supra note 2, at 399-402. 83. See Spencer Overton, Rules, Standards, and Recounts: Form and the Law of Democracy, 36 HARV. C.R.-C.L. L. REV. (forthcoming 2001) (manuscript at 16-18, on file with author). COUNTS, RECOUNTS, AND ELECTION CONTESTS: LESSONS FROM THE FLORIDA PRESIDENTIAL ELECTION STEVE BICKERSTAFF* I. GENERAL PRINCIPLES GOVERNING POSTELECTION DISPUTES ........................... A. Canvassing Ballots and Certifying the Outcome of an Election................. B. Recounting Votes........................................................................................... C. The Contesting of an Election ...................................................................... II. A DISPUTED ELECTION OUTCOME ....................................................................... III. IN PARTIAL DEFENSE OF THE FLORIDA SUPREME COURT OPINIONS .................. A. Palm Beach County Canvassing Board v. Katherine Harris (Round One) ............................................................................................................... B. Albert Gore v. Katherine Harris (Round Two) ........................................... IV. THE PRECEDENTIAL SIGNIFICANCE OF BUSH V. GORE ....................................... A. Requirements for Equal Protection.............................................................. B. Discerning Voter Intent ................................................................................ C. The Single Impartial Arbiter ....................................................................... V. GORE’S LOSING LEGAL STRATEGY ....................................................................... VI. CONCLUSION ........................................................................................................ 427 428 429 431 434 440 440 446 456 456 458 461 463 467 The thirty-six days following the November 7, 2000, presidential election was an amazing period in the legal and political history of the United States. Lawyers, judges, and election officials in Florida found themselves participating sometimes unwillingly in a national media spectacle. Some observers saw this postelection turmoil in Florida as a crisis for this country. Others viewed it as an international embarrassment. The outcome brought elation for many and cynicism from others who explained the result as a triumph of power and politics over justice and the will of the people. For columnists and law faculty, the events in Florida and the ruling of the United States Supreme Court on December 12 in Bush v. Gore1 provided an opportunity for decrying the “frail underside of elections”2 and speculating about the “legacy” of the Supreme Court opinion for voting rights nationwide.3 In an op-ed piece in The New York Times, Columbia Law School professor Sam Issacharoff posited that: * Visiting Professor of Law, University of Texas at Austin. The author is a partner at Bickerstaff, Heath, Smiley, Pollan, Kever & McDaniel LLP, in Austin, Texas. He has practiced election law for more than 20 years and has represented clients in statewide recounts in Texas. 1. 531 U.S. 98 (2000) (per curiam). 2. SAMUEL ISSACHAROFF ET AL., WHEN ELECTIONS GO BAD: THE LAW OF DEMOCRACY AND THE PRESIDENTIAL ELECTION OF 2000, at i (rev. ed. 2001). 3. See Samuel Issacharoff, The Court’s Legacy for Voting Rights, N.Y. TIMES, Dec. 14, 2000, at A39. 425 426 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 [T]he Supreme Court may have given us an advancement in voting rights doctrine. It has asserted a new constitutional requirement: To avoid disparate and unfair treatment of voters. And this obligation obviously cannot be limited to the recount process alone. . . . The court’s new standard may create a more robust constitutional examination of voting practices.4 Elsewhere in this issue, Professors Richard Briffault and Richard L. Hasen offer their informed opinions regarding the effect of Bush v. Gore on election law and state election systems.5 In general, the recent wave of articles tends to analyze the events in Florida without regard to how an election recount and contest might proceed under more normal circumstances.6 As a result, these articles underestimate the integrity and vitality of the existing state election processes nationwide and overestimate the need for corrective rules or laws by Congress or state legislatures. This Article suggests that the events in Florida and the resulting decisions of the Florida Supreme Court and United States Supreme Court are best understood against the backdrop of the state law processes and jurisprudence that usually govern the outcome of an ordinary election dispute. Measured against this backdrop, the election recount and contest process in Florida functioned in reasonably good order under the circumstances and, had adequate time been available, probably would have produced a reasoned and credible outcome. The problems in Florida that plagued the postelection process and ultimately led to the unsettling decision in Bush v. Gore are largely attributable to external factors, not to flaws within the election process itself. The circumstances that confronted the postelection process in Florida were novel only in the magnitude of the external factors affecting the operation of the recount and election contest process. The 4. Id. 5. See Richard Briffault, Bush v. Gore as an Equal Protection Case, 29 FLA. ST. U. L. REV. 325 (2001); Richard L. Hasen, Bush v. Gore and the Future of Equal Protection Law in Elections, 29 FLA. ST. U. L. REV. 377 (2001). 6. See, e.g., Richard A. Epstein, “In such Manner as the Legislature Thereof May Direct”: The Outcome in Bush v. Gore Defended, 68 U. CHI. L. REV. 613 (2001); Samuel Issacharoff, Political Judgments, 68 U. CHI. L. REV. 637 (2001); Michael W. McConnell, Twoand-a-Half Cheers for Bush v. Gore, 68 U. CHI. L. REV. 657 (2001); Frank I. Michelman, Suspicion, or the New Prince, 68 U. CHI. L. REV. 679 (2001); Richard H. Pildes, Democracy and Disorder, 68 U. CHI. L. REV. 695 (2001); Richard A. Posner, Bush v Gore: Prolegomenon to an Assessment, 68 U. CHI. L. REV. 719 (2001); David A. Strauss, Bush v. Gore: What Were They Thinking?, 68 U. CHI. L. REV. 737 (2001); Cass R. Sunstein, Order Without Law, 68 U. CHI. L. REV. 757 (2001); John C. Yoo, In Defense of the Court’s Legitimacy, 68 U. CHI. L. REV. 775 (2001); Nelson Lund, The Unbearable Rightness of Bush v. Gore (Apr. 26, 2001) (unpublished manuscript, from “Votes and Voices” symposium at Benjamin N. Cardozo School of Law), available at http://papers.ssrn.com; Robert H. Bork, Sanctimony Serving Policies: The Florida Fiasco, THE NEW CRITERION, Mar. 2001, at 4; Robert F. Nagel, From U.S. v. Nixon to Bush v. Gore, THE WKLY. STANDARD, Dec. 25, 2000, at 20. 2001] COUNTS, RECOUNTS, CONTESTS 427 legal issues and proceedings took on a surreal appearance primarily because of the importance of the election, the intensity of the media’s scrutiny, the magnitude of the forces employed on both sides of the election conflict, and the brevity of time available for the participants and the courts to address complex legal and factual issues. These same factors often play a significant role in state recounts or election contests. The differences encountered in Florida were a matter of degree, not a matter of kind. Examining the events in Florida in the context of general election law jurisprudence also provides important insights into what went wrong for the Gore legal strategy. The Gore team simply lost the legal battle. Gore’s postelection legal strategy was flawed from the beginning, largely because it failed adequately to appreciate the general principles and dynamics that historically have governed the conduct of recounts and election contests nationwide and that almost certainly would eventually control the outcome in Florida. There was essentially no chance from the beginning of the postelection dispute that any candidate could win an election contest in the time available. Gore’s sole chance for victory laid in the administrative recount process. That opportunity was squandered. Part I of this Article provides an overview of the general principles found in election law nationwide governing the canvass, recount, and contest aspects of elections. Part II describes the postelection events in Florida. Part III discusses the major opinions of the Florida Supreme Court and the opinion of the United States Supreme Court in Bush v. Gore. To minimize unnecessary repetition of the descriptions available elsewhere in the papers of this symposium, I focus primarily on the Gore v. Harris7 opinions of the Leon County Circuit Court and Florida Supreme Court. Part IV discusses the probable impact of Bush v. Gore and the events in Florida on election law generally. Finally, I consider the events in Florida in terms of candidate Gore’s losing legal strategy. I. GENERAL PRINCIPLES GOVERNING POSTELECTION DISPUTES An election is a process, not an event. The objective of this process is to determine the will of the electorate as expressed by qualified voters casting ballots in accordance with applicable state law. Detailed state statutes and administrative regulations govern the process in every state.8 The specifics of these statutes and regulations 7. Gore v. Harris, No. 00-2808, 2000 WL 1770257 (Fla. 2d Cir. Ct. Dec. 4, 2000), rev’d, 772 So. 2d 1243 (Fla. 2000), rev’d sub nom. Bush v. Gore, 531 U.S. 98 (2000). 8. State statutes often are compiled in a separate election code, such as the Texas Election Code. Administrative rules are promulgated and published by the state’s chief elections officer. Florida’s election law statutes appear generally in Florida Statutes chs. 428 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 vary significantly by state. It is possible, however, to make generalizations about the procedures followed by the states in the counting of state and federal ballots and about the election law principles applied by the state courts.9 This Article examines these general procedures and principles as they exist for the canvassing and recounting of ballots and for contesting an outcome in an election. A. Canvassing Ballots and Certifying the Outcome of an Election The Secretary of State serves as the chief elections officer in most states. This official generally is responsible for implementing and enforcing the state’s election laws through regulations, official opinions, and unofficial guidance for local government officials. These local government officials usually bear responsibility for conducting the election, including the obligation as governed by state law to purchase, maintain, and operate voting devices at which votes are cast, as well as any electronic or electromechanical systems used for counting ballots. The primary means of in-person voting include paper ballots read manually, levered machines, prescored punch cards, direct recording electronic (DRE) machines, and paper ballots read by optical scan equipment.10 Absentee ballots are usually paper ballots read manually or with optical scan equipment. Each of these means of voting and counting votes has vulnerabilities.11 Election returns from a general election usually are tabulated by the local election officials.12 For punch card ballots and paper ballots, the counting or tabulation of votes can occur at the precinct or at a central counting location as prescribed by state law. Levered machines are inspected and the vote tabulated at the precinct. This is also true of the DRE machines. Results obtained at the precinct level are communicated to the central election office and added by county election officials to the results from other precincts.13 Within a prescribed period after the day of the election, these results are officially canvassed by designated local government or canvassing officials.14 The canvass usually consists only of officially opening the returns 101-102. The Florida Legislature amended the election code in the 2001 session. See Florida Election Reform Act of 2001, 2001 Fla. Laws ch. 01-40, at 117-73. 9. See generally NAT’L ASS’N OF STATE ELECTION DIRS., ELECTION ADMIN. SURVEY (2001) [hereinafter 2001 ELECTION ADMIN. SURVEY] (compiled by the association following the 2000 election and in response to events in Florida), and the individual state responses on which the survey compilation is based. 10. See ROY G. SALTMAN, U.S. DEP’T OF COMMERCE, ACCURACY, INTEGRITY AND SECURITY IN COMPUTERIZED VOTE-TALLYING (Aug. 1988) (Nat’l Bureau of Standards, Special Pub. No. 500-158). 11. See id. 12. See, e.g., FLA. STAT. § 102.112 (2000) (amended 2001). 13. See, e.g., id. § 101.5614 (amended 2001); TEX. ELEC. CODE § 67.017 (2000). 14. See, e.g., FLA. STAT. § 102.112 (2000) (amended 2001); TEX. ELEC. CODE § 67.003 (2000). 2001] COUNTS, RECOUNTS, CONTESTS 429 from the precincts and recording and tabulating the votes for each candidate.15 In many jurisdictions, this recording and tabulation of votes already has been accomplished by election officials, and the canvass is pro forma. The county governing body or canvassing board certifies the winner of local general elections.16 An official certificate of election is issued.17 The county governing body or canvassing board usually is responsible also for totaling the returns from within the county in multicounty and statewide elections and for forwarding the results for these elections to the state officer, board, or legislature responsible by law for officially canvassing the votes and certifying the outcome for these elections.18 For political party primary elections, the county and state executive committees for the party usually perform the same canvass and certification duties performed respectively by the local and state officials for the general election.19 State laws strictly control the secrecy of the ballots and the access of persons to paper or punch card ballots that could be altered to change or add votes for a candidate. Meticulous records must be kept throughout the process of the names of voters, number of ballots used, number of spoiled ballots, and number of voted ballots to better prevent fraud, as well as to allow a review of the vote counting process, if necessary, after the election. Ballots and levered or DRE voting machines remain in custody for a period as prescribed by state law following the election to ensure the ability of election officials and courts to verify vote totals as necessary to resolve any disputes arising from the election.20 B. Recounting Votes A process for recounting votes exists in virtually every state.21 The process generally is administrative in nature, with the recount conducted through the government authority responsible for conducting the election or through a special government committee or commission created for the purpose of the recount.22 Sometimes, the recount 15. See, e.g., FLA. STAT. § 102.061 (2000) (amended 2001); TEX. ELEC. CODE § 67.004 (2000). 16. See, e.g., TEX. ELEC. CODE § 67.016 (2000). 17. See, e.g., FLA. STAT. § 102.155 (2000) (amended 2001); TEX. ELEC. CODE § 67.016 (2000). 18. See, e.g., FLA. STAT. § 102.141 (2000) (amended 2001); TEX. ELEC. CODE §§ 67.007, 67.008 (2000). 19. See, e.g., TEX. ELEC. CODE §§ 172.119, 172.120 (2000). 20. See, e.g., id. 21. See 2001 ELECTION ADMIN. SURVEY, supra note 9; see, e.g., FLA. STAT. § 102.166(4)-(10) (2000) (amended 2001); TEX. ELEC. CODE §§ 211-15 (2000). 22. See, e.g., State ex rel. Crawford v. Del. Cir. Ct., 655 N.E.2d 499 (Ind. 1995) (noting that under applicable Indiana law, the judge appoints a recount commission). 430 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 must occur before ballots are canvassed or election results are certified, with the recounted totals being substituted for the initial tally.23 In some states, however, a recount may continue after the outcome of the election has been certified.24 Many states provide for a recount of ballots automatically if the difference of votes between candidates meets some statutory threshold.25 This recount or “retabulation” of votes generally consists of simply reprocessing the same ballots back through the same manual process or counting machines used in the original tally of votes. A manual recount generally occurs only in defined circumstances and only if specifically requested by a candidate or voter.26 A recount, including a manual recount, is intended to identify and to correct any inaccuracies in the initial tally caused by human error or equipment malfunction.27 The government officials are concerned with accuracy in the vote totals, not the outcome of the election. It is the candidates and their partisan representatives who are concerned with using the recount to win an election.28 Most recounts occur at the local level even for statewide or multijurisdictional elections. In many states, recounts can proceed on a county-by-county29 or even a precinct-by-precinct basis in multicounty and statewide elections because the alleged human error or equipment malfunction in question probably did not occur throughout the election jurisdiction.30 To require a manual recount in all counties or precincts because of errors in only one or a few counties or precincts within the election jurisdiction would effectively make manual recounts impossible because of cost. Essential to fairness under such a decentralized recount scheme for multicounty elections is the entitlement of candidates, or their partisan representatives, to witness the government officials as they recount the ballots and thereby are able to observe any irregularities in the process or any illegal votes included or legal votes excluded 23. See 2001 ELECTION ADMIN. SURVEY, supra note 9; see, e.g., FLA. STAT. § 102.166(4)(b) (2000) (amended 2001); TEX. ELEC. CODE § 213.057 (2000). 24. See 2001 ELECTION ADMIN. SURVEY, supra note 9; see, e.g., OKLA. STAT. Tit. 26, § 8-116 (2000). 25. See 2001 ELECTION ADMIN. SURVEY, supra note 9; see, e.g., FLA. STAT. § 102.166(4)(d) (2000) (amended 2001); NEB. REV. STAT. § 32-1119 (2001). 26. See 2001 ELECTION ADMIN. SURVEY, supra note 9; see, e.g., FLA. STAT. § 102.166 (2000) (amended 2001); TEX. ELEC. CODE § 214.042 (2000). 27. Siegel v. LePore, 120 F. Supp. 2d 1041, 1050 (S.D. Fla. 2000) (describing the purpose of a manual recount as “detecting and correcting clerical or electronic tabulating errors”), aff’d, 234 F.3d 1163 (11th Cir. 2000) (en banc), cert. denied, 531 U.S. 1005 (2000). 28. See TIMOTHY DOWNS ET AL., THE RECOUNT PRIMER 3 (1994). 29. The reference to “county” in this context is intended as generic for the local government generally charged with carrying out elections. Instead of a county, this local government may be a parish, township, or other local entity. 30. See, e.g., TEX. ELEC. CODE § 212.001(5) (2000). 2001] COUNTS, RECOUNTS, CONTESTS 431 from the vote total.31 If disputes arise that cannot be resolved at the administrative stage, an election contest is available to challenge an uncorrected irregularity or vote tabulation before an impartial arbiter. Specific aspects of the recount process in Florida during 2000 that proved important in the battle for presidential votes included: (1) a provision for an automatic recount statewide;32 (2) authority for a candidate subsequently to file a protest in a county that the returns of the election in the county are erroneous and to request a manual recount;33 (3) authority for a local canvassing board, in response to such a protest, to order a limited manual recount of not less than three precincts;34 and (4) authority for the canvassing board to manually recount all ballots in the county if the limited manual recount “indicates an error in the vote tabulation which could affect the outcome of the election.”35 C. The Contesting of an Election Election contests were unknown at common law.36 They constituted special statutory authority for courts to intervene in the otherwise legislative or political election arena.37 As a result, election contest statutes as a rule are strictly construed. An election contest generally is initiated by a losing candidate or candidates after the outcome of an election has been certified and a winner has been declared. State courts have established presumptions in favor of the accuracy of a certified election outcome.38 Therefore, any contestant has the significant burden of showing that the certified results should be overturned. Moreover, state courts often have established threshold requirements for election contests by requiring specificity in the pleading of alleged election irregularities 31. See FLA. STAT. § 102.166 (2000) (amended 2001); see also Miller v. County Comm’n, 539 S.E.2d 770, 776 (W. Va. 2000) (indicating that a recount gives candidates the opportunity to (1) observe the manner in which the recount is conducted, (2) notify the canvassing board of their intentions regarding requesting a recount in precincts not requested by the candidate originally requesting the recount, and (3) identify ballots that may be challenged as irregular or illegal in an election contest). “Inherent in the recount procedures is the concept of fairness to all interested candidates in an election.” Id. at 776. 32. FLA. STAT. § 102.166(4)(d) (2000) (amended 2001). 33. Id. § 102.166(4)(a) (amended 2001). 34. Id. § 102.166(4)(d) (amended 2001). 35. Id. § 102.166(5) (amended 2001). 36. See Harless v. Lockwood, 332 P.2d 887, 888 (Ariz. 1958); McPherson v. Flynn, 397 So. 2d 665, 668 (Fla. 1981); State ex rel. Lydick v. Brown, 516 P.2d 239, 240 (Okla. 1973); see also Henderson v. Maley, 806 P.2d 626, 634 (Okla. 1991). 37. These state concerns are similar to the concerns underlying the basic political question doctrine of justiciability as explained in Baker v. Carr, 369 U.S. 186 (1962). See McIntyre v. Fallahay, 766 F.2d 1078, 1087-88 (7th Cir. 1985) (Swygert, J., dissenting). 38. See, e.g., In re Contest of the Election for the Offices of Governor and Lieutenant Governor Held at the Gen. Election on Nov. 2, 1982, 444 N.E.2d 170, 172 (Ill. 1983). 432 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 and a showing of probable success by the contestant even before the ballot boxes are opened and the contested ballots are subjected to review.39 The contestant’s failure to make this showing can be a basis for dismissal of the election proceeding.40 In some states the courts will uphold the declared results unless there is clear and convincing evidence that the results are inaccurate and do not reflect the will of the people. Although state election contest statutes vary, most provide that the grounds for contesting an election are established when: (1) an election official has engaged in fraud or other illegal conduct; (2) illegal votes have been counted; (3) legal votes have been rejected or excluded; and (4) an election official has been bribed.41 Misconduct of an election official generally is an insufficient basis for a contest of the election result unless the illegal acts are substantial or intentional and affect the outcome of the election.42 Similarly, the inclusion or exclusion of ballots is a basis for an election contest only if the contestant pleads and can show that the ballots in question were counted or excluded improperly and that these ballots are sufficient to alter the outcome of the election. Such a showing may exist as a threshold requirement for any election contest. Remedies generally available for a court in an election contest are (1) to declare a winner of the election or (2) to order a new election. Some state courts are prohibited from changing the outcome of an election and are limited to declaring the election void and ordering a 39. See id. at 175-77. 40. See, e.g., Gooch v. Hendrix, 851 P.2d 1321, 1327 (Cal. 1993) (“[A] primary principle of law as applied to election contests [is] that it is the duty of the court to validate the election if possible. That is to say, the election must be held valid unless plainly illegal.”); Swift v. Registrars of Voters, 183 N.E. 727, 728-29 (Mass. 1932); Christenson v. Allen, 119 N.W.2d 35, 40 (Minn. 1963) (refusing to allow an election contest as a “fishing expedition”); Jackson v. Maley, 806 P.2d 610, 615 (Okla. 1991) (“[C]ourts indulge every presumption in favor of the validity of an election and, where possible, that validity will be sustained. Mere probabilities will not suffice to carry this initial burden.”) (citation omitted); In re Opening of Ballot Boxes, Montour County, 718 A.2d 774, 777 (Pa. 1998) (invalidating a recount because of a lack of verification of signatures on recount petition); Madigan Appeal, 253 A.2d 271, 275 (Pa. 1969) (indicating that a showing that a partial recount of precincts discloses possible errors in other precincts is insufficient to obtain a recount of the other precincts in a state senate election). 41. See, e.g., FLA. STAT. § 102.168(3) (2000) (amended 2001); TEX. ELEC. CODE § 221.003 (2000). 42. See, e.g., Fladell v. Palm Beach County Canvassing Bd., 772 So. 2d 1240 (Fla. 2000) (concluding as a matter of law that the Palm Beach (butterfly) ballot did not constitute substantial noncompliance with statutory requirements); Jacobs v. Seminole County Canvassing Bd., CV No. 00-2816 (Fla. 2d DCA Dec. 8, 2000) (rejecting a challenge to absentee balloting in Seminole and Martin Counties on the basis that plaintiffs’ evidence failed to support a finding of fraud, gross negligence, or intentional wrongdoing by election officials). 2001] COUNTS, RECOUNTS, CONTESTS 433 new election.43 In other states, the calling of a new election is available as a remedy only if the court is unable to determine who has won the prior election.44 At least two states have allowed votes to be added or subtracted from a candidate’s total based on an allocation formula.45 Although local general elections usually are subject to challenge through judicial election contests, a certified result in a statewide or multijurisdictional election often is subject to challenge only before designated legislative or executive officers. For example, any contest of a general election to the United States House of Representatives or Senate ultimately is resolvable only by those legislative bodies.46 State recounts and possibly even state judicial election contests can proceed as a means of policing state election laws, so long as they do not interfere with the exclusive power of the respective houses of Congress to ultimately determine the election dispute.47 Similarly, in most states, state legislatures have exclusive authority to determine the outcome of any contest of the official results of a general election for membership in the state legislative body.48 In addition, many states have designated the state legislature or state officials other than the judiciary to decide contests of statewide elections.49 When entrusted with power to determine an outcome in an election contest, these state50 and federal51 legislative bodies have fol43. See, e.g., Becker v. Pfeifer, 588 N.W.2d 913, 918 (S.D. 1999). Under South Dakota law, the court can only uphold the election as the free and fair expression of the will of the voters or declare the election void. Id. At least one state has allowed a new election applicable only to one precinct or to certain specific voters previously denied the opportunity to cast a qualified ballot. See State ex rel. Olson v. Bakken, 329 N.W.2d 575, 579-82 (N.D. 1983). 44. See, e.g., Green v. Reyes, 836 S.W.2d 203, 207 (Tex. App. 1992) (indicating that in Texas “[w]hen the court, with some degree of certainty, can determine the outcome of the election based upon the evidence presented by the parties, [state law] requires it do so”). If the alleged irregularity makes it impossible to determine the will of the voters, a court may call a new election. See, e.g., Marks v. Stinson, 19 F.3d 873, 879-89 (3d Cir. 1994). 45. See Canales v. City of Alviso, 474 P.2d 417, 421-22 (Cal. 1970) (subtracting illegal votes from the vote totals in a municipal consolidating election based on circumstantial evidence that illegal voters most likely voted for the proposition); In re The Purported Election of Bill Durkin, 700 N.E.2d 1089, 1095 (Ill. App. Ct. 1998) (subtracting illegal votes from candidates by precinct according to the proportion of votes received by the candidate in the precinct). 46. See U.S. CONST. art. I, § 5; see also Roudebush v. Hartke, 405 U.S. 15, 19 (1972). 47. See Roudebush, 405 U.S. at 26 (allowing a statewide recount of votes for U.S. Senate so long as such state action does not interfere with the ability of the U.S. Senate to make a final determination in the dispute; the resulting recount reduced Hartke’s winning margin by only fifteen votes); Thorsness v. Daschle, 279 N.W.2d 166, 168-70 (S.D. 1979) (holding that the United States Constitution does not bar enforcement by the state courts of state procedures designed to insure the legal outcome of its elections). 48. See, e.g., TEX. ELEC. CODE § 221.002 (2000). 49. See, e.g., id. 50. See, e.g., Robert A. Junnell et al., Consideration of Illegal Votes in Legislative Election Contests, 28 TEX. TECH L. REV. 1095 (1997). 434 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 lowed procedures akin to a judicial proceeding with an opportunity before a legislative committee for written submissions, evidence, and oral argument from the opposing candidates. In most states this power entrusted to legislative and executive officers is exclusive and is not subject to challenge in the state courts. Primary elections pose a special problem for resolving election disputes.52 Outcomes must be final in time for inclusion of the winning candidate on the general election ballot. Moreover, sometimes a primary election will not produce a winner and a runoff election is necessary. These severe time constraints have made contests difficult to prosecute. In most states the power to determine disputed primary elections is given to the local and state executive committees of the affected political party. In some states this power is exclusive. In other states the state courts retain the power to entertain a challenge to party primary elections. In each election scenario, a single judicial or administrative arbiter resolves disputes concerning voting irregularities or the inclusion or exclusion of votes from throughout the election district. Potential differences among local canvassing boards are resolved by this arbiter de novo in a contested proceeding by reference to the applicable standard set by state law. This requirement for a single, ostensibly unbiased arbiter is an essential principle of fairness for resolving election contests. II. A DISPUTED ELECTION OUTCOME 53 The polls closed in most of Florida at 7 p.m. Eastern time on November 7, 2000.54 Shortly thereafter, several national news organizations predicted that Gore would win Florida’s twenty-five electoral votes based on voter exit polls, turnout of voters at selected election precincts, and historical voting patterns.55 As the evening progressed, however, candidate Bush remained substantially ahead in the actual 51. See Roudebush, 405 U.S. at 27 (Douglas, J., dissenting in part) (explaining that the United States Senate has established a custom of resolving disputes by allowing the apparent loser to allege the suspected voting irregularities and the votes affected; if the claim is not frivolous, the proceeding may involve the subpoena of ballots and the calling of witnesses to testify). 52. See Curry v. Baker, 802 F.2d 1302, 1304 (11th Cir. 1986) (considering a challenge to the authority of the Alabama Democratic Party Executive Committee to resolve dispute between candidates in the party primary). 53. This Article provides only a very brief and incomplete guide to some of the major events as background for its legal analysis. Many sources provide a more complete retelling of the events in Florida. See, e.g., THE NEW YORK TIMES, 36 DAYS: THE COMPLETE CHRONICLE OF THE 2000 PRESIDENTIAL ELECTION CRISIS (2001); THE WASHINGTON POST, DEADLOCK: THE INSIDE STORY OF AMERICA’S CLOSEST ELECTION (2001) [hereinafter DEADLOCK]. 54. Polls remained open in some counties in the western panhandle of Florida until 8 p.m. Eastern time. DEADLOCK, supra note 53, at 35-36. 55. Id. at 35. 2001] COUNTS, RECOUNTS, CONTESTS 435 vote tally.56 News organizations began to reverse their earlier decision to call Gore the victor in Florida.57 By 2 a.m., it appeared that Bush had won Florida and the Presidency.58 At approximately 2:30 a.m., Gore called Bush to express his intention to concede.59 Nevertheless, as interested officials of both campaigns and the general public watched in amazement, Bush’s lead in Florida shrank from 50,000 votes to fewer than 6,000.60 Gore called Bush to indicate that “circumstances had changed” and that he was no longer prepared to concede.61 By sunrise on November 8, Bush’s lead in the unofficial vote tally in Florida had dwindled to 1,784, with the state’s electoral votes now clearly essential to both candidates for victory in the Electoral College.62 Lawyers from both campaigns were dispatched to Florida like troops in opposing armies, maneuvering for what was to become a gargantuan, chaotic legal and political battle over final certification of the winner of the state’s electoral votes.63 The Presidency of the United States was the reward for the victor. With approximately 6 million ballots cast in Florida, the miniscule difference between the two candidates automatically triggered a statewide recount.64 Most election officials conducted this automatic recount using the same counting equipment and procedures they used on election night, with ballots in counties using optical scanners and punch card electronic counting systems being rerun through the same machines.65 Nevertheless, the vote changed and by one unofficial count the margin between Gore and Bush had narrowed to 327 votes after the recount tabulations from Florida’s sixty-seven counties were totaled on November 14.66 Some absentee ballots from overseas voters remained to be counted.67 Nevertheless, it was clear that the exclusion of some votes that had been counted or the inclusion of even a handful of votes from ballots that had not been counted could change the outcome of the election statewide. Over the next several weeks, different categories of disputed ballots would be identified. In hindsight, at least each the following categories of votes, or potential votes, existed on November 8 and 56. Id. at 43. 57. Id. at 40. 58. Id. at 43-44. 59. Id. at vii. 60. Id. at 46-47. 61. Id. at 49. 62. Id. at 70. 63. Id. at 60-62, 65, 73. 64. See FLA. STAT. § 102.141(4) (2000) (amended 2001). 65. DEADLOCK, supra note 53, at 77. 66. According to the Associated Press tally on November 11, Gore gained 2,520 votes in the recount, while Bush gained 1,063 votes. This reduced the difference between the two candidates from 1,784 to 327. Id. at viii. 67. Id. at 99. 436 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 were subject to possible dispute in a manual recount or through legal proceedings in state or federal court: 1. An estimated 110,000 ballots on which no vote was counted because the electronic counting equipment recorded votes on the ballot for more than one presidential candidate (overvotes);68 2. An estimated 43,000 to 70,000 ballots on which no vote for any presidential candidate was counted because the electronic counting equipment recorded no vote (undervotes);69 3. Absentee votes in at least two counties (Seminole and Martin) where local election officials allowed Republican Party officials to correct absentee ballot applications after the applications had been received in the election official’s office;70 4. Differences between the initial count and a second tabulation in Nassau County that produced 218 fewer votes, with a net gain of 51 votes for Bush;71 5. Overseas absentee ballots that were counted even though the envelopes containing the ballots failed to have the date postmarked as required by state law;72 6. Differences in the condition and type of voting equipment used by Florida counties that are alleged to have caused significant disparities in the percentage of overvotes and undervotes among counties, with the highest percentage of uncounted ballots (primarily overvotes) occurring in areas of the state with large AfricanAmerican or Caribbean-American neighborhoods;73 68. More than 113,000 ballots included a vote for more than one presidential candidate, with approximately 104,000 including a vote for either Gore (75,000) or Bush (29,000). Ford Fessenden & John M. Broder, Study of Disputed Ballots Finds Justices Did Not Cast the Deciding Vote, N.Y. TIMES, Nov. 12, 2001, at A1; Dan Keating & Dan Balz, Election 2000: Closer Than Close, WASH. POST, Nov. 11, 2000, available at http://stacks.msnbc.com/news/656172.asp#BODY. These overvotes included: (1) the result of the confusing butterfly ballot in Palm Beach County, which may have cost Gore approximately 8,000 votes; and (2) the two-page presidential ballot used in Duval County in which 20% of the presidential votes in predominately African-American precincts were thrown out as overvotes, thereby possibly costing Gore approximately 7,000 votes. Id. Other ballots treated as overvotes resulted from erasures or from voters apparently trying to be extra-clear in their choice by both voting for a candidate and writing in the name of the same candidate. Id. These latter categories of overvotes should have been counted under Florida’s voter intent standard, but some counties, such as Lake and Escambia, did not check ballots rejected by the voting machines. The net gain by Gore from counting such ballots in these two counties alone would have been 329 votes. Id. 69. DEADLOCK, supra note 53, at xiv. 70. Id. at 159. 71. Id. at 158. 72. See David Barstow & Dan van Natta, Jr., How Bush Took Florida: Mining the Overseas Absentee Vote, N.Y. TIMES, July 15, 2001, at A1 (describing how Republican lawyers, in coordination with Florida Secretary of State Katherine Harris, ostensibly were successful in adding a net of 739 votes for Bush from absentee ballots received, often without postmarks, after November 7). 73. DEADLOCK, supra note 53, at 116-17; see Ford Fessenden, Ballots Cast by Blacks and Older Voters Were Tossed in Far Greater Numbers, N.Y. TIMES, Nov. 12, 2001, at A17. Alleged racial discrimination during the election has been the focus of an investigation by 2001] COUNTS, RECOUNTS, CONTESTS 437 7. Ballots uncast because of the alleged intimidation of voters, primarily in African-American or Caribbean-American neighborhoods;74 and 8. Registered voters who remained ineligible on election day because they had not corrected an erroneous finding by state election officials that they had out-of-state felony convictions.75 Numerous state or federal lawsuits were filed challenging these different categories of potential votes or alleged voting irregularities. I will focus, however, on the legal actions that largely deal only with the issues surrounding ballots that were recorded by the counting equipment as having no vote for any candidate for President (undervote) or as having a vote for more than one candidate for President (overvote). Gore attorneys timely petitioned under state law for manual recounts of undervotes in only four of Florida’s sixty-seven counties. These requests were styled “protests” and were filed with the respective county canvassing boards as required by state law.76 These counties included Miami-Dade, Broward, Palm Beach, and Volusia. It soon became clear, however, that the manual recounts in three of these counties77 would not be completed by the statutory deadline of November 14 for reporting returns to the Secretary of State for canvassing and certification of a winner of the election. Several of the United States Commission on Civil Rights. The Commission’s Report indicates that African-American voters in Florida were more likely than white voters to have their ballots discarded. See generally UNITED STATES COMM’N ON CIVIL RIGHTS, VOTING IRREGULARITIES IN FLA. DURING THE 2000 PRESIDENTIAL ELECTION (2001) [hereinafter VOTING IRREGULARITIES IN FLA.], available at www.usccr.gov/. The disparity was greatest in Duval County where over 20% of the ballots in predominately African-American precincts were disqualified because they contained votes for more than one presidential candidate. Id. ch. I, at 20-32; see also id. app. VII, at 10 (report by Allan J. Lichtman, The Racial Impact of the Rejection of Ballots in the 2000 Presidential Election in the State of Florida). The dissenting members of the Commission urge that this apparent disparity in the “spoilage rate” for ballots in predominately African-American precincts was not based on race but on factors such as a lower literacy rate and higher number of first time voters among the African-American voters. Id. app. IX, at 3, 15-16 (report entitled The Florida Election Report: Dissenting Statement by Commissioner Abigail Thernstrom and Commissioner Russell G. Redenbaugh). The dissenting statement further suggests that if any “blame” is to be assigned, it should more appropriately go to the local election officials who designed the ballots and purchased the voting machines rather than to Governor Bush or Secretary of State Harris. Id. app. IX, at 20-21. A “sample ballot” in Duval County urged voters to “vote all pages.” Id. ch. 8, at 7. It is likely that this instruction caused many of the first time voters in the African-American precincts of Duval County to cast a vote for Gore on page one of the ballot and another presidential candidate on page two. 74. DEADLOCK, supra note 53, at 158-59; see also VOTING IRREGULARITIES IN FLA., supra note 73, chs. 2, 7. 75. DEADLOCK, supra note 53, at 158; see also VOTING IRREGULARITIES IN FLA., supra note 73, ch. 1, at 33-35; id. ch. 5. 76. See FLA. STAT. § 102.166(2) (2000) (amended 2001). 77. Volusia County completed its manual recount in time to meet the November 14 deadline. The recount added a net of 98 votes for Gore. By the evening of November 14, Bush’s lead was down to 300 votes, with the overseas absentee ballots yet to be added. 438 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 these counties sought authority to later file amended returns that would include the results of the manual recounts. On November 15, Secretary of State Katherine Harris rejected the counties’ reasons for submitting amended returns after the statutory deadline and announced that she would certify the outcome in the presidential election based only on the returns that had earlier been submitted to her by the counties on or before November 14 and the returns of overseas absentee ballots.78 The Florida Democratic Party and Al Gore filed actions in Leon County seeking to compel the Secretary of State to accept the amended returns.79 This action was consolidated with an earlier action brought on behalf of Volusia County (in which Palm Beach County had intervened).80 The cases were certified by the District Court of Appeal to the Florida Supreme Court.81 The Florida Supreme Court, by order on Friday, November 17, enjoined the Secretary of State from certifying the election results until further order of the court.82 After hearing oral argument, the court on November 21 rejected any “hyper-technical reliance upon statutory provisions” and ordered the Florida Secretary of State to accept amended returns received by 5 p.m. on Sunday, November 26.83 George Bush appealed this decision. On December 5, a unanimous United States Supreme Court agreed to vacate the order of the Florida Supreme Court and to remand the case for the Florida Supreme Court to clarify questions that could determine whether the case presented any federal questions within the jurisdiction of the nation’s highest Court.84 In the meantime, Bush’s lead had increased to 930 counted votes after inclusion of the absentee overseas ballots on November 18.85 The deadline of November 26 set by the Florida Supreme Court for amended returns to be received by the Florida Secretary of State passed with only two of the four counties (Broward and Volusia) having completed manual recounts.86 On the evening of November 26, George W. Bush was certified the winner of the Florida election. The 78. Even Professor Epstein acknowledges that the Secretary of State “made the wrong choice in cutting off the recount so precipitously” but concludes that the decision was nevertheless within the limits of her statutory power. Epstein, supra note 6, at 626. 79. See Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d 1220, 1227, 1240 (Fla. 2000), vacated sub nom. Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000) (per curiam). 80. Id. at 1227. 81. Id. 82. Id. 83. Id. at 1227, 1240. 84. Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000). 85. DEADLOCK, supra note 53, at x. 86. Palm Beach County completed its recount on November 26, but after the 5 p.m. deadline. Secretary of State Harris refused to accept either the partial recount or the latefiled complete recounted returns. 2001] COUNTS, RECOUNTS, CONTESTS 439 certified result showed that he had won by 537 votes out of a total of approximately 6 million.87 Gore’s attorneys were now left with only the option of pursuing an election contest to challenge this certified result. An election contest was promptly filed in Leon County Circuit Court88 and was tried on December 3 and 4.89 Circuit Judge N. Sanders Sauls denied Gore’s claims in the election contest.90 The stage was now set for the final battles before the Florida Supreme Court and the United States Supreme Court. Gore’s attorneys quickly appealed Judge Sauls’ ruling, and on December 8 the majority opinion of a sharply divided Florida Supreme Court reversed Judge Sauls’ decision.91 The court ordered that the amended returns from Palm Beach County and the partially recounted returns from Miami-Dade County be added to the candidates’ totals.92 The addition of these amended returns reduced Bush’s lead to less than 193 votes.93 The majority opinion further directed the circuit court on remand to “tabulate by hand the approximate 9000 Miami-Dade ballots, which the counting machine registered as non-votes, but which have never been manually reviewed,” and “to enter such orders as are necessary to add any legal votes to the total statewide certifications.”94 Gore supporters understandably were optimistic that a continuation of the recount would quickly produce additional votes sufficient to overcome Bush’s miniscule lead. Local election officials and political party representatives mobilized throughout Florida. Attorneys for both candidates gathered before Leon County Circuit Court Judge Terry P. Lewis on the evening of December 8 to argue over how best to tabulate the ballots in Miami-Dade County and to count any uncounted legal votes elsewhere in the state. This count began on the morning of December 9 but ended soon thereafter when the United States Supreme Court stayed the order of the Florida Supreme Court.95 On December 12, a divided United States Supreme Court reversed the decision of the Florida Supreme Court and effectively brought the legal battle to an end, 87. The final tallies were 2,912,790 votes for Bush and 2,912,253 for Gore. Barstow & van Natta, supra note 72. 88. See Complaint to Contest Election, Gore v. Harris (Fla. 2d Cir. Ct. Nov. 27, 2000) (No. 00-2808), available at http://199.44.225.4/courtDockets/pdf/election_cases/CV-002808a.pdf. 89. See Gore v. Harris, No. 00-2808, 2000 WL 1770257 (Fla. 2d Cir. Ct. Dec. 4, 2000). 90. Id. at *1. 91. Gore v. Harris, 772 So. 2d 1243 (Fla. 2000), rev’d sub nom. Bush v. Gore, 531 U.S. 98 (2000). 92. Id. at 1262. 93. Whether this difference was 193 or 154 votes depended on the outcome of a separate dispute over the number of audited votes received by each candidate as a result of the recount in Palm Beach County. See id. at 1248 n.6. 94. Id. at 1262 (emphasis added). 95. Bush v. Gore, 531 U.S. 1046 (2000). 440 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 making George W. Bush the certified winner of Florida’s twenty-five electoral votes and the winner of the Presidency.96 III. IN PARTIAL DEFENSE OF THE FLORIDA SUPREME COURT OPINIONS A. Palm Beach County Canvassing Board v. Katherine Harris (Round One) The Florida Supreme Court’s unanimous decision on November 21 to extend the statutory deadline to allow amended election returns to be included in the final certified vote total was based on Florida law.97 The court found that the applicable state statutes conflicted insofar as they set a timeframe for conducting a manual recount that was unworkable for the state’s most populous counties under the timeframe set for submitting county returns to the Secretary of State.98 The court determined that state law99 authorized local canvassing boards under certain circumstances to conduct a manual recount.100 “[L]ogic dictates that the period of time required to complete a full manual recount may be substantial, particularly in a populous county, and may require several days.”101 Thus, the court reasoned, construing state law to require the results of these recounts to be filed with the Secretary of State within as little as one day after a timely request may be filed would create a conflict and effectively make manual recounts impossible, particularly in the largest counties. The court found that such a conflict was avoided because state 96. See Bush v. Gore, 531 U.S. 98 (2000). 97. See 772 So. 2d 1220 (Fla. 2000), vacated sub nom. Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000). 98. See FLA. STAT., §§ 102.111, 102.112(1) (2000) (amended 2001) (providing that returns are to be filed with the Secretary of State by 5 p.m. on the seventh day following the election). 99. See id. § 102.166(4) (2000) (amended 2001) (authorizing a written request for a manual recount to be filed within seventy-two hours after midnight of the day the election was held). 100. The first issue resolved by the Court was whether local canvassing boards had authority under state law to conduct a manual recount countywide in circumstances where a discrepancy of votes found in a sample manual recount of selected precincts exists for some reason other than incorrect election parameters in the vote tabulation software. See Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d at 1229; see also FLA. STAT. § 102.166(5) (2000) (amended 2001). Secretary of State Harris had indicated to Florida counties that a manual recount was allowed only if made necessary by fraud or substantial noncompliance with the state’s election procedures. Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d at 1238-39 (quoting a November 21, 2000, letter from Katherine Harris to the Palm Beach County Canvassing Board). The court concluded that an “error in vote tabulation” allowing a county canvassing board to conduct a countywide recount also included circumstances in which the discrepancy between the original machine return and the sample manual recount is due to the manner in which the ballot has been marked or punched. Id. at 1228. 101. Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d at 1232. 2001] COUNTS, RECOUNTS, CONTESTS 441 law was ambiguous concerning the Secretary of State’s authority to reject late-filed returns. The applicable provisions setting the deadline for submitting returns had both mandatory and permissive language. Applying traditional rules of statutory construction102 and guided by the principle that the will of the people is the paramount consideration,103 the court concluded that the Secretary of State’s authority to ignore amended county returns after the statutory deadline of November 14 was limited to circumstances that would compromise the integrity of the electoral process.104 The court then proceeded to define those circumstances for the present situation and to set a new deadline of November 27.105 Although controversial when issued, the supreme court’s opinion is consistent with Florida law and with the treatment of deadlines in other states where similar state statutory deadlines or requirements often are considered “directory” instead of “mandatory” when the statute itself does not clearly indicate a contrary legislative intent.106 As the Florida Supreme Court recognized, disallowing a county’s more accurate amended election returns because of the possibly dilatory actions of a local canvassing board is a drastic penalty that “misses the constitutional mark.”107 Professor Richard Epstein faults the “sorry performance” of the Florida Supreme Court as justifying what he also acknowledges is a less than perfect performance by the United States Supreme Court.108 He concludes that “there is ample reason to believe . . . that the Florida Supreme Court adopted, under the guise of interpretation, a scheme . . . that deviates markedly from that which the Florida legislature had set out in its statutes.”109 Professor Epstein is particularly critical of the Florida Supreme Court’s attempt to impose its will on the state’s chief election officer, Secretary of State Katherine Harris.110 According to Professor Epstein, the events leading to the decision in Bush v. Gore could have been avoided if the Florida Supreme 102. Id. at 1234. The rules of statutory construction cited by the court include that (1) a specific statute controls over a general statute, (2) a more recently enacted statute controls an older statute, (3) a statutory provision should never be interpreted so as to render it meaningless, and (4) related statutory provisions must be read as a cohesive whole. Id. 103. Id. at 1236. 104. Id. at 1239. 105. Id. at 1240. 106. See, e.g., Wilks v. Mouton, 722 P.2d 187 (Cal. 1986). Even mandatory provisions of state law should be liberally construed to avoid thwarting the fair expression of the peoples’ will. Id at 190; see also Timm v. Schoenwald, 400 N.W.2d 260, 263 (N.D. 1987). 107. Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d at 1240. 108. Epstein, supra note 6, at 635. 109. Id. at 634. 110. Id. at 634. 442 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 Court had properly followed the law as interpreted by Harris.111 Specifically, Professor Epstein urges that Harris was properly exercising her statutory authority by concluding that: (1) the county canvassing boards could not conduct manual recounts where a discrepancy between the original machine return and a sample manual recount is due to the manner in which a ballot has been marked or punched, and (2) Harris could refuse to accept amended returns after November 14 unless persuaded that the recounted numbers were corrections of mathematical errors or the result of an election official’s substantial noncompliance with state law.112 Professor Epstein argues that manual recounts of ballots countywide in Florida were inappropriate because “[t]he reason we have machine counts is to guard against the risk of human error and bias.”113 He offers no authority for this position. It is correct that levered voting machines were first utilized at least partly in response to the fraud that accompanied the use of paper ballots in the nineteenth century.114 Also, hand counting of large numbers of paper ballots is generally considered to be less accurate than counting with machinereadable ballots.115 Nevertheless, all electronic counting systems are vulnerable and susceptible to significant error.116 Usually this margin of error is not relevant for determining the outcome of an election because the recorded difference in votes among the candidates makes the will of the voters clear, even if the recorded number of votes for each candidate is somewhat imprecise. 111. Id. Professor Epstein urges that the Florida Supreme Court erred because it was bound to defer to the legal interpretation given by Secretary of State Harris to the Florida statutes. Court deference to a government agency on issues of statutory interpretation is understandably less absolute than on issues of fact or matters of policy. Harris relied on essentially legal interpretations of state statutes as a basis for refusing to accept amended county tallies. This tactic left her decisions susceptible to court challenge on the basis that her interpretations were contrary to the law. The Florida Supreme Court agreed with this challenge. See Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d at 1228. 112. Epstein, supra note 6, at 622. Professor Epstein’s description of the Secretary of State’s position is somewhat more thoughtfully worded than the actual position taken by Katherine Harris for rejecting amended returns. In her November 15 letter to the canvassing board in Palm Beach County, Harris essentially adopted the state’s case law applicable to a court’s review of certified results in an election contest and used it to create a burden of proof for county canvassing boards. She concluded that she was justified in rejecting amended returns unless the county canvassing boards alleged fraud or substantial noncompliance with state law and could show more than a mere “possibility” that the amended returns could affect the outcome of the election. Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d at 1226 n.5 (quoting the November 15 letter from Katherine Harris to the Palm Beach County Canvassing Board). Although, as discussed elsewhere, such a burden of proof exists for a contestant in an election contest and would affect Gore’s success before Judge Sauls, the Florida Supreme Court was correct in finding that such reasons were inapposite for rejecting amended returns from government canvassing boards. 113. Epstein, supra note 6, at 625. 114. SALTMAN, supra note 10. 115. Id. at 26. 116. Id. at 25-51. 2001] COUNTS, RECOUNTS, CONTESTS 443 Virtually all states provide by law for some means of manually recounting ballots in circumstances in which the candidate vote totals from the electronic counting machines are near equal.117 An attempt on election night to manually count and to record votes in multiple races from thousands of ballots could result in substantial human error. A manual recount of machine-counted ballots, however, generally is limited to only one race and takes place in a strictly regulated process designed to ensure that the interests of all candidates are protected and that the most accurate count possible is achieved. The Florida Supreme Court was almost certainly correct that, as in most other states, manual recounts are an available means by which a local canvassing board can attempt to identify and to correct errors in the machine counting of ballots. Professor Epstein also argues that if manual recounts of ballots countywide were possible under Florida law, “the sole function of the hand recount is to examine ballots to see whether they meet the standards for a ballot that is machine-readable.”118 This position greatly oversimplifies a complex legal issue. If a ballot is machinereadable, the votes on the ballot will have been recorded by the machine. Any variation in a manual count tabulation from that of a machine count necessarily results from the human decisionmaker concluding that a ballot contains a legal vote even when the electronic counting machine failed to record the vote. In each of the four counties in Florida in which a manual recount was conducted, the members of the recount committees found at least some ballots that they agreed had clearly ascertainable but unrecorded votes. This is likely to be true in any manual recount of punch card or optical scan paper ballots. Therefore, the controlling issue in Florida and other states during a manual recount is not whether such ballots are machinereadable. Instead, the issue is whether the ballot contains a legal vote under state law.119 If the vote is legal, it should be counted. Any failure to include such ballots is subject to challenge in an election contest. The issue of what constitutes a legal vote differs according to the laws of the various states. In virtually all states, a legally cast bal117. See 2001 ELECTION ADMIN. SURVEY, supra note 9; see, e.g., FLA. STAT. § 102.166 (2000) (amended 2001); TEX. ELEC. CODE § 214.042 (2000). 118. Epstein, supra note 6, at 623. 119. Professor Epstein correctly points out that it is the Florida Secretary of State who by law is responsible for at least initially providing guidance on what constitutes a valid vote under Florida law. Id. at 624. In many states, the state’s chief elections officer has promulgated rules to provide guidance for discerning a legal vote during a manual recount. See, e.g., TEX. ELEC. CODE § 214.042 (2000). Florida in 2000 had no such guidelines. Instead of furnishing such guidelines after the 2000 election, Secretary of State Harris took official positions that at the time appeared designed to prevent a manual recount altogether. 444 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 lot120 is to be counted according to the “intent of the voter” even if a vote has not been recorded for the ballot by the voting machine.121 If a ballot has been legally cast, courts have been reluctant to leave it uncounted when the ballot reflects the voter’s intent to cast a vote for a particular candidate and that specific vote has later become crucial to the election or defeat of that candidate.122 The Supreme Court of Florida cited the decision of the Illinois Supreme Court in Pullen v. Mulligan for the explanation that “[t]o invalidate a ballot which clearly reflects the voter’s intent, simply because a machine cannot read it, would subordinate substance to form and promote the means at the expense of the end.”123 The Florida Court’s decision to permit the manual counting of ballots by local canvassing boards to discern voter intent when such votes may affect the outcome of an election is consistent with Florida law and with the approach adopted by state courts nationwide. In his concurring opinion in Bush v. Gore, Chief Justice Rehnquist urged that ballots that cannot be read by a counting machine because of the manner in which the ballots have been marked or punched are a result of “voter error” and are “improperly marked ballots” that the Florida Secretary of State could appropriately refuse to include in the state’s certification of results.124 In regard to punch card ballots, the asserted voter error generally is seen as a failure by the voter to fully or effectively comply with the instruction125 to check her ballot after voting and to clear any “chips” hanging from the back 120. State courts will disallow votes that are cast illegally even though the ballot is machine-readable. This issue occurs in many contexts, such as a voter’s nonresidency, ineligibility to vote in the primary of another party, or failure to comply with requirements for voting absentee. Even in the case of ballots cast in violation of a state’s absentee voting laws, however, some courts have permitted the votes to be counted on the basis that the error was not attributable to the voter and that to disallow the vote would unfairly disenfranchise the voter. 121. See, e.g., Delahunt v. Johnston, 671 N.E.2d 1241, 1243 (Mass. 1996) (indicating that “if the intent of the voter can be determined with reasonable certainty from an inspection of the ballot . . . [then] effect must be given to that intent”) (citing Pullen v. Mulligan, 561 N.E.2d 585, 611 (Ill. 1990); McIntyre v. Wick, 558 N.W.2d 347, 359 (S.D. 1996) (indicating that when marks on a ballot are sufficiently plain to gather therefrom a part of the voter’s intent, there is a duty to count the ballot); see also Partido Nuevo Progresista v. Perez, 639 F.2d 825 (1st Cir. 1980) (permitting courts of Puerto Rico to count ballots containing marks outside the designated spaces on the ballot); Duffy v. Mortenson, 497 N.W.2d 437 (S.D. 1993) (finding chad sufficiently dislodged on one ballot after subjecting disputed ballots to examination by “stereoscope”). 122. See, e.g., Duffy, 497 N.W.2d at 439. 123, Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d 1220, 1238 ( Fla. 2000), vacated sub nom. Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000) (citing Pullen, 561 N.E.2d at 611). 124. Bush v. Gore, 531 U.S. 98, 118-119 (2000) (Rehnquist, C.J., concurring). 125. This instruction does not amount to a legal requirement sufficient to disqualify a ballot. 2001] COUNTS, RECOUNTS, CONTESTS 445 of the card.126 Chief Justice Rehnquist cautioned that “[n]o reasonable person would call it ‘an error in the vote tabulation’ or a ‘rejection of legal votes’ when electronic or electromechanical equipment performs precisely in the manner designed, and fails to count those ballots that are not marked in the manner that these voting instructions explicitly . . . specify.”127 The real world operates somewhat differently than seemingly envisioned by Chief Justice Rehnquist. Voters, in the rush of casting their vote, routinely fail to check their ballots afterward and fail to completely knock out any chips or chad that may remain attached to a punch card ballot. Does the disqualifying “voter error” occur at this moment, when a voter fails to adequately comply with the instructions, or later when the machine fails to record the vote? This distinction is important because loosened chad remaining on the ballots after the vote is cast routinely are cleared purposely or knocked inadvertently from punch card ballots throughout the counting process by election workers during handling of the punch cards. The counting machines themselves dislodge significant quantities of loosened chad during the machine count. In fact, experts concede that it is generally not possible to duplicate a machine count obtained on prescored punch cards because chad continues to become detached on each occasion that the punch cards are machine-counted.128 Apparently, no state tries to disqualify voters because they fail to clear chad from their ballot. Nor do states attempt to discriminate between voters based on whether the chad on a punch card ballot is removed by the voter, or is later knocked off during the counting process, or remains by chance attached after the machine count. It is very unlikely that the Florida Legislature intended such discrimination or would apply the same standard differently to the same ballot. Concern over the possibility that participants in an administrative recount might apply different standards or apply the same standard differently to the same ballot is justified. The events in Florida clearly demonstrate how different canvassing boards or members of the same canvassing board may reach different conclusions regarding voter intent. The Florida Supreme Court decision in Palm Beach 126. Bush v. Gore, 531 U.S. at 119 (Rehnquist, C.J., concurring) (quoting Florida Instruction); see Epstein, supra note 6, at 632. 127. Bush v. Gore, 531 U.S. at 119 (Rehnquist, C.J., concurring) (statutory citations omitted). At least one state court (Massachusetts) has expressly rejected this argument on the basis that while voters sometimes could perhaps do a better job of expressing themselves, ballots should not automatically be disqualified because of a failure to comply strictly with announced procedures. See Delahunt v. Johnston, 671 N.E.2d 1241, 1243 (Mass. 1996). 128. See SALTMAN, supra note 10, at 35. In Florida, the automatic machine recount or retabulation statewide immediately following the election resulted in 3,583 additional votes, with Gore gaining a net of 1,784 votes. 446 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 County Canvassing Board v. Harris does not purport to decide how voter intent should be discerned; nor, however, does it find that variations in how local canvassing boards determine legal votes are immune from challenge in an election contest. It is during an election contest that irregularities in counting “legal votes” can be challenged and corrected. This issue is discussed further below. In sum, however, I believe the Florida Supreme Court correctly decided Palm Beach County Canvassing Board v. Harris. B. Albert Gore v. Katherine Harris (Round Two) The Florida Supreme Court’s Gore v. Harris decision on December 8 is much more problematic than its earlier decision in Palm Beach County Canvassing Board v. Harris. Gore v. Harris reached the court by appeal from Leon County Circuit Judge Sanders Sauls’ decision to reject Al Gore’s election contest complaint.129 Three opinions by different members of the Florida Supreme Court reflect a sharp division over the applicable law and the appropriate role of a court in fashioning relief under the circumstances that existed on December 8. A majority of four justices reversed the circuit court and ordered it to include the votes for Gore from the recounts in Palm Beach and MiamiDade counties and enter such orders as necessary to add any legal votes statewide to the candidates’ totals.130 Three justices dissented generally on the basis that Gore had failed to carry his burden at trial as a plaintiff in the election contest and that no adequate remedy could be fashioned in the time remaining if Florida’s presidential electors were to be selected by December 12, as contemplated by 3 U.S.C. § 5.131 On December 12, in Bush v. Gore, the United States Supreme Court reversed the Florida court’s ruling. Bush v. Gore also reflects a sharply divided Court. The Court’s per curiam opinion found four problems with the Florida Supreme Court order under the Fourteenth Amendment requirement of equal protection. These were as follows: 1. The Florida court’s order permitted inconsistent treatment, both among counties and within counties, in the determination of which ballots would count as legal votes in a manual recount;132 129. See Gore v. Harris, CV No. 00-2808, 2000 WL 1770257 (Fla. 2d Cir. Ct. Dec. 4, 2000). 130. Gore v. Harris, 772 So. 2d 1243, 1248 (Fla. 2000), rev’d sub nom. Bush v. Gore, 531 U.S. 98 (2000). 131. Id. at 1268-70 (Wells, C.J., dissenting); see also id. at 1270-72 (Harding, J., dissenting). 132. Bush v. Gore, 531 U.S. at 106. The concern in the per curiam opinion that standards for accepting or rejecting ballots during a recount might vary “not only from county to county but indeed within a single county from one recount team to another,” id. at 106, 2001] COUNTS, RECOUNTS, CONTESTS 447 2. The Florida court’s order permitted the inclusion of overvotes in some counties while not providing for a recounting of these ballots in other counties;133 3. The Florida court’s order permitted inclusion of a partial recount from Miami-Dade County and failed to guarantee that recounts must be complete in order to be included in any final certified result;134 and 4. The Florida court’s order failed to provide a satisfactory process for counting the votes, or even providing who would recount the ballots.135 At least six of the Justices of the United States Supreme Court joined in finding that the Florida court’s remedial order violated equal protection.136 The United States Supreme Court’s per curiam opinion in Bush v. Gore correctly identified very significant problems that could have arisen under the Florida Supreme Court’s remedial plan for counting votes in the presidential election. Even those Justices who most vigorously dissented from the United States Supreme Court’s decision expressed understandable concern over the degree of unequal treatment of votes and voters possible under the Florida court’s remedial is misplaced. Disputes over how ballots are counted within specific manual recount teams are resolved by the canvassing board. In turn, differences among counties are subject to challenge through an election contest. 133. Id. at 107-08. 134. Id. at 108. 135. Id. at 109. 136. The per curiam opinion indicated that “[s]even Justices of the Court agree that there are constitutional problems with the recount ordered by the Florida Supreme Court that demand a remedy.” Id. at 111. In a separate concurring opinion, Chief Justice Rehnquist was joined by Justices Scalia and Thomas in challenging both the November 21 and December 8 decisions of the Florida court as infringing on the power of and the statutory scheme prescribed by the Florida Legislature. Id. at 122 (Rehnquist, C.J., concurring). Justice Stevens, in a dissenting opinion joined by Justices Ginsburg and Breyer, found no substantial federal question that would entitle the Court to reverse the Florida court’s decision. Id. at 123-29 (Stevens, J., dissenting). Writing separately in a dissenting opinion in which she was joined by Justices Stevens and Breyer, Justice Ginsburg agreed that the Court lacked jurisdiction and further disputed the conclusion that a constitutionally adequate recount is impractical. Id. at 135-44 (Ginsburg, J., dissenting). In a separate dissenting opinion, Justice Souter agreed with the per curiam that the Florida Supreme Court order violated equal protection, but he concluded that the state courts should be afforded an opportunity to establish uniform standards for counting ballots and should be given an opportunity to try to manually recount all disputed undervotes before the applicable federal deadlines. Id. at 129-35 (Souter, J., dissenting). Justice Breyer’s separate dissenting opinion is less clear about whether he believes the Fourteenth Amendment had been violated, as he explained his agreement with the per curiam’s concerns in terms of how the disparities in the recount “implicate principles of fundamental fairness.” Id. at 145 (Breyer, J., dissenting). Justice Breyer would have permitted the Florida recount to continue under uniform standards. Id. at 158 (Breyer, J., dissenting). 448 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 order.137 Importantly, however, the Florida Supreme Court’s majority opinion did not expressly condone unfair or unequal treatment. The opinion directed the circuit judge to “enter such orders as are necessary to add any legal votes to the total statewide certifications.”138 The problems raised by the Justices of the United States Supreme Court in the context of the Fourteenth Amendment are precisely the type of legal and procedural questions that customarily are resolved through the adversarial process present in any state election contest. As suggested by Justice Souter, the Florida courts might well have eventually dealt with the use of differing substandards for determining voter intent in different counties employing similar systems if given the opportunity to do so.139 Rather than a holding that unequal treatment in an election contest is acceptable under the state or federal constitutions, the majority opinion of the Florida Supreme Court is a product of the extraordinary circumstances in which the court found itself on December 8. Some of the difficulties faced by the court were apparent at the time,140 including the brief period of time it had to decide complex legal issues. The majority opinion further recognized that “practical difficulties may well end up controlling the outcome of the election.”141 Despite these difficulties, a majority of the Florida court resolved to “do the best we can.”142 The court’s majority opinion reflects that ambitious, but mistaken, effort. One frequent criticism of the majority opinion is misplaced. Several commentators have urged that an election contest is in the na- 137. Id. at 126 (Stevens, J., dissenting) (admitting that “the use of differing substandards for determining voter intent in different counties employing similar voting systems may raise serious concerns”). 138. Gore v. Harris, 772 So. 2d 1243, 1262 (Fla. 2000), rev’d sub nom. Bush v. Gore, 531 U.S. 98 (2000) (emphasis added). 139. In the view of Justice Stevens, the traditional safeguard for election contests nationwide existed even under the Florida court’s remedial order—i.e., “a single impartial magistrate [that] will ultimately adjudicate all objections arising from the recount process.” Bush v. Gore, 531 U.S. at 126 (Stevens, J., dissenting). 140. The court, however, also faced less obvious difficulties created by the attorneys for the opposing parties. Secretary of State Harris opposed any manual count of votes and therefore offered no expert suggestions as to how the Court might proceed to manually count all legal votes in the state. The attorneys for Gore were so focused on obtaining an immediate addition of the votes from Palm Beach and Miami-Dade Counties and a count of the remaining 9,000 uncounted votes in Miami-Dade that they prosecuted the election contest on flawed legal theories before the circuit court and failed to provide either an adequate trial record or a legally sound remedial plan. On the other hand, the attorneys for Bush were essentially uninterested in aiding the Florida court in finding a remedy for the various disparities in treatment that the attorneys had identified. Bush was best served on appeal to the United States Supreme Court by an incomplete and insufficient opinion from the Florida Supreme Court. 141. Gore v. Harris, 772 So. 2d at 1261 n.21. 142. Id. 2001] COUNTS, RECOUNTS, CONTESTS 449 ture of a judicial review of an administrative decision.143 If this were true, the standard of review would be abuse of discretion.144 State law nationwide, however, treats an election contest as an independent judicial action specially authorized by statute. Circuit Judge Sauls applied an abuse of discretion standard to find that the Miami-Dade County board had not abused its discretion by deciding not to count the remaining 9,000 undervotes.145 The majority opinion of the Florida Supreme Court in Gore v. Harris rejected this holding.146 While dissenting from the remainder of the majority opinion, Justices Harding and Shaw also expressly rejected Circuit Judge Sauls’ finding on the standard of judicial review because he “improperly intertwined [the recount protest and election contest] and the standards applicable to each.”147 The commentators have made the same mistake. Once a judicial election contest is commenced, it is the court that is charged with deciding if legal votes have been excluded or illegal votes included.148 Cases challenging proceedings before a canvassing board, such as attempts to compel a board to conduct a recount,149 are inapposite to an election contest. Nevertheless, the majority decision in Gore v. Harris is wrongly decided. The majority correctly identified the conflicting principles— “the necessity for counting all legal votes” to effect the will of the electorate and the ultimate need for finality.150 The majority, however, weighed these principles incorrectly when it concluded that “we must do everything required by law to ensure that legal votes that have not been counted are included in the final election results.”151 This decision was inconsistent with election law in Florida and elsewhere for four reasons. First, the decision is mistaken because it essentially put the court in the position of trying to fashion a remedy despite Gore’s failed legal strategy. It was Circuit Judge Sauls who first found that the bur143. See Epstein, supra note 6, at 630-31; McConnell, supra note 6, at 668-69. 144. Epstein, supra note 6, at 630-31. 145. Transcript: Judge N. Sanders Sauls Rules Against Gore’s Contest, at 3 (Dec. 4, 2000) (copy on file with the Florida State University Law Review) [hereinafter, Transcript]. 146. Gore v. Harris, 772 So. 2d at 1252. 147. Id. at 1270 (Harding, J., dissenting). 148. See id. at 1271 (Harding, J., dissenting) (indicating that the issue of whether a canvassing board has rejected a number of legal votes sufficient to change or to place in doubt the election by virtue of cutting short a manual recount is to be determined de novo, not under an abuse of discretion standard); see also McIntyre v. Wick, 558 N.W.2d 347, 358 (S.D. 1996) (describing the scope of review in an election contest as de novo). But see id. at 1265. (Wells, C.J., dissenting) (concluding that the contest and protest statutes must be read together). 149. See, e.g., Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d 1220 (Fla. 2000), vacated sub nom. Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000); Broward County Canvassing Bd. v. Hogan, 697 So. 2d 508, 510 (Fla. 1992). 150. Gore v. Harris, 772 So. 2d at 1261. 151. Id. 450 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 den on Gore in the election contest was “to place in issue and seek as a remedy with the attendant burden of proof a review and recount of all ballots in all counties in this state with respect to the particular alleged irregularity or inaccuracy in the balloting or counting processes alleged to have occurred.”152 Gore’s failure to meet this burden should have been determinative of the outcome in the election contest regardless of any policy favoring the counting of all votes or effecting the will of the voters. While not upholding Sauls’ dismissal of the election contest, the majority agreed “that it is absolutely essential in this proceeding and to any final decision, that a manual recount be conducted for all legal votes . . . in all Florida counties where there was an undervote and, hence a concern that not every citizen’s vote was counted” and that the election should not be decided “by strategies extraneous to the voting process.”153 By remanding the election contest for a count of uncounted legal votes statewide, the majority was effectively attempting to save Gore from his own losing strategy. This was not justified. The court’s second mistake was its finding that a number of nonvotes exceeding the difference between two candidates was sufficient under section 102.168, Florida Statutes, to obtain a count of those ballots in an election contest. Significant numbers of nonvotes exist in every election. Only a fraction of such nonvotes are legal votes under even the most liberal standard for discerning voter intent. Moreover, net votes gained by a losing candidate from a manual counting of nonvote ballots in one county or precinct may be offset by votes gained by her opponent from a manual counting of nonvote ballots in other counties or precincts. Manually counting ballots throughout the county, district, or state in question is time-consuming, costly, and potentially chaotic. It is necessary to make a threshold showing that rejected legal votes exist sufficient to change or to place in doubt the outcome of an election. This threshold is not met by a mere showing that there are more nonvotes in one county or precinct than there are votes separating the two candidates.154 152. Transcript, supra note 145 (emphasis added). 153. Gore v. Harris, 772 So. 2d at 1253. 154. The presence of 9,000 uncounted ballots in Miami-Dade County does not meet the burden of showing that sufficient rejected legal votes existed to change the outcome or to place in doubt the outcome of the presidential election in Florida. Testimony offered by Gore at trial was inadequate even to show that the 9,000 ballots remaining to be manually counted in Miami-Dade County contained sufficient net votes for Gore to overcome Bush’s lead. Gore’s expert witness, Professor Nicolas Hengartner, on direct examination testified only as to the “recovery rate” of valid votes that could occur with a manual recount of the uncounted ballots. Contest Hearing R. at 176-92, Gore v. Harris, CV No. 00-2808 (Fla. 2d DCA Dec. 6-7, 2000). Gore’s petition alleged that, if a manual count of the 9,000 votes resulted in the same proportional increase in net votes as the ballots already counted by the Miami-Dade Canvassing Board, Gore would gain a net of 600 votes. Id. (R. at 326). Professor Hengartner, however, offered no testimony directly supporting this projection. The Re- 2001] COUNTS, RECOUNTS, CONTESTS 451 A third fundamental mistake by the majority was its failure to realize that the desire to count every vote had been eclipsed by the need for finality. The Illinois Supreme Court faced a very similar dilemma in 1983 when the certified results of a statewide race for Governor and Lieutenant Governor showed a difference of 5,074 votes out of 3,627,128 cast for the Democratic and Republican candidates for the these state offices.155 The Democratic candidates, Adlai E. Stevenson, Jr., and Grace Mary Stern, filed an election contest. In their petition, they claimed that after reviewing designated precincts in 70 out of 102 counties they had discovered alleged irregularities that would increase their net vote by 4,664 votes.156 Moreover, they pointed to thousands of ballots allegedly lacking the requisite initials of the election judge or bearing identifying marks made by the voter that would render those ballots illegal. In rejecting the petition, the Illinois Supreme Court pointed to the insufficiency of the pleadings and to the expense and burden of conducting the election contest. The court further explained: Until such an election contest is resolved, the political turmoil surrounding it and the fact that it will be unknown in this case whether the incumbent governor will continue to hold that office could effectively prevent the legislative and executive branches of government from dealing with the urgent problems facing this State. The State of Illinois should not be forced to endure these consequences on the mere suspicion of defeated candidates or on their belief or hope that an election contest would change the results.157 This decision brought an abrupt end to an important statewide election contest without resolving precisely which candidates received the greatest number of legal votes. 158 Sometimes the desire to publican expert witness, Dr. Laurentius Marais, countered that the projection of a gain in Gore votes from the uncounted votes in Miami-Dade County was “unreliable and inaccurate” because it was based on the false premise that the proportion of votes for Gore would be the same for the ballots that remained to be manually counted. Id. (R. at 326-27). Dr. Marais pointed out that the precincts that had been manually recounted in Miami-Dade had voted greater than 75% for Gore while the remaining precincts had voted 52% for Bush. Id. (R. at 327). He concluded that there was no basis for projecting from the heavily Democratic precincts to the remaining precincts. Id. Subsequent manual counts by the media of all undervote ballots in Miami-Dade County have shown that Bush would have gained net votes from a manual recount of the 9,000 uncounted undervotes. Fessenden & Broder, supra note 68. 155. In re Contest of the Election for the Offices of Governor and Lieutenant Governor Held at the General Election on Nov. 2, 1982, 444 N.E.2d 170, 172 (Ill. 1983). 156. Id. at 179-80. 157. Id. at 178 (emphasis added). 158. Predictably, three judges dissented from this ruling. Id. at 183 (Ward, J., dissenting). These justices claimed that the majority’s requirement for specificity in pleading in an election contest means “the candidate must do, in practical terms, the impossible” because she must accumulate sufficient evidence in fifteen days from 102 counties to show that she will win the election. Id. at 189 (Ward, J., dissenting). The dissenting judges further challenged the majority’s practical arguments by indicating that “[o]ur society does not fix af- 452 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 count every vote and to be absolutely certain of the will of the voters must, in the absence of fraud, ultimately yield to the essential need for finality.159 This point was reached in Florida in the presidential election.160 By December 8 there was no reasonable possibility that a continuation of the election contest could result in a fair, credible result within a reasonable time. Four justices of the Florida Supreme Court were unwilling to accept this reality. Finally, in its effort to achieve a count of undervotes statewide in the inadequate time available, the majority of the Florida Supreme Court abandoned the principles that are fundamental to a fair recount and election process and to an accurate outcome—that is, an adequate opportunity for candidates to identify alleged voting irregularities during the counting process and, in a contested proceeding, to have any disputed issues resolved before a single impartial arbiter. The most egregious example is the majority’s direction that the circuit court must include the additional votes for Gore from the manual counts in Miami-Dade and Palm Beach Counties. It was clear from the record that these ballots had been counted using differing standards for determining voter intent and that these standards in turn might very well be different from the standards used elsewhere in the state during the court-supervised manual review and counting of undervotes. Issues of arbitrary and disparate treatment of ballots in the other counties in the statewide count of nonvotes might theoretically have been resolved in time through court supervision in a contested judicial proceeding. However, the order that the circuit court include the Palm Beach and Miami-Dade votes for Gore in the fordable expense as a standard for doing justice or, in particular, for insuring the integrity of the electoral process.” Id. at 191 (Ward, J., dissenting). 159. An important aspect of our democracy is that power in government can readily transfer in confidence that, at the end of the term of office, it will again transfer if the election outcome is different. No outcome of any single primary or general election is sufficiently important to warrant significant disruption of the governing process, even in the interest of assuring the accuracy of the election outcome. There must be finality. The candidates, but more importantly the government and the people, must move on. See McIntyre v. Fallahay, 766 F.2d 1078, 1088 (7th Cir. 1985) (Swygert, J., dissenting) (indicating that our government is a representative democracy and that the people cannot be properly represented unless the legitimacy and authority of the elected official to represent them is finally determined). 160. Federal courts long have recognized an analogous principle of withholding a remedy when “exigent circumstances” justify conducting an election under an unlawful election system, or allowing the results of an election under an unlawful system to remain unchanged. Even as it gave federal courts the task of remedying unconstitutional state apportionment of legislative districts, the U.S. Supreme Court in Reynolds v. Sims, 377 U.S. 533, 585-87 (1964), explained that federal courts may be bound to award or to withhold relief based on the mechanics and complexities of state election laws. This principle has been applied many times since 1964. See, e.g., Upham v. Seamon, 456 U.S. 37, 44 (1982); Ely v. Klahr, 403 U.S. 108 (1971); Kilgarlin v. Hill, 386 U.S. 120, 121 (1967); Terrazas v. Clements, 537 F. Supp. 514, 537 (N.D. Tex. 1982) (three-judge panel), stay denied, 456 U.S. 902 (1982). 2001] COUNTS, RECOUNTS, CONTESTS 453 candidate’s vote total despite a lack of uniform counting substandards essentially mandated a disparity in the treatment of ballots. Moreover, this determination suggested to the circuit court that fairness and accuracy in other vote counts could have been sacrificed for the sake of expediency.161 While I do not believe that the Florida Supreme Court intended any unfairness, its December 8 order both directly and implicitly created the possibility for a result inconsistent with the law of Florida and with the fundamental principles of fairness followed in election contests nationwide. The United States Supreme Court assessed that the recount process underway in Florida in the wake of Gore v. Harris “was probably being conducted in an unconstitutional manner.”162 As a result, the Court stayed the order of the Florida Supreme Court directing the recount.163 In its subsequent opinion in Bush v. Gore, the Court concluded that “upon due consideration of the difficulties identified to this point, it is obvious that the recount cannot be conducted in compliance with the requirements of equal protection and due process without substantial additional work.”164 Those who urge that the recount process should have been allowed to proceed under the Florida Supreme Court’s remedial order, or under a new order providing uniform rules for the recount, underestimate the legal task faced on December 9 and both the number and complexity of the legal disputes that remained to be resolved. The many parties to the Gore election contest already had filed numerous motions pending ruling by the circuit court and were certain 161. As the Gore attorneys on remand before Judge Lewis pushed for an expedited recount, the already questionable legal status of the election contest deteriorated rapidly. In the interest of time, (1) the ballots manually counted in Broward and Palm Beach Counties were to be included in the vote total without any further review, (2) the partial recount that had already been completed by the Miami-Dade County Board of approximately 20 percent of the county’s undervote ballots would be left unchanged while the remaining ballots from the county would be counted by a new group of examiners, and (3) the undervotes in the remaining counties would be counted by the officials of those counties. No rules were established for use by these different counting groups. Also in the interest of time, Judge Lewis barred party observers from objecting during the manual counting process, although the observers could keep a list of disputed ballots that might serve as a basis for objections later. In effect, these procedures created a certainty that ballots would be counted according to the subjective judgment of many different persons without adequate provision for an adjudication of disputed ballots or issues by a single impartial arbiter. This publicized and broadcast proceeding suggested that a rush existed to count ballots in a seemingly arbitrary or haphazard fashion. This impression available through the media probably contributed to concerns at the United States Supreme Court that the situation was out of hand. 162. Bush v. Gore, 531 U.S. 98, 109 (2000). 163. Bush v. Gore, 531 U.S. 1046 (2000). 164. Bush v. Gore, 531 U.S. at 110. 454 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 to file many more.165 The Florida Supreme Court decision left the circuit court the task of initially determining at least the following: 1. The mechanics of the expedited statewide recount (for example, whether ballots should be manually reviewed and votes counted by local canvassing boards or all ballots should be shipped to Leon County and reviewed and counted by the circuit judge or under his supervision); 2. The logistics of the recount process (for example, the qualifications of the counters, the possible lack of personnel to conduct the count, the fatigue of the counters, and intracounty personnel issues among the state’s sixty-seven counties); 3. The procedures to be utilized to assure the opportunity of the candidates’ representatives to observe the recount process and to obtain a judicial resolution of disputes; 4. The categories of ballots to be reviewed to discern voter intent (for example, only ballots with undervotes, or also ballots with overvotes as challenged by the Bush attorneys); 5. The counties to be included in the recount (for example, only uncounted votes in counties using punch cards or also nonvote ballots in counties using other voting equipment); 6. The standards, if any, required to be used by the counters in manually reviewing ballots for voter intent; 7. Whether to manually review federal write-in ballots and to include them in the statewide tabulation even though not fully compliant with state law; 8. The merit of any challenges to the manually counted votes from Broward, Miami-Dade and Palm Beach Counties based on the use in those recounts of different standards than might be used for the remainder of Miami-Dade County and the state; 9. The number of the 9,000 undervote ballots from Miami-Dade County that constituted uncounted legal votes and the candidate for whom the voter intended to cast her ballot; 10. The resolution of disputes concerning the approximately 175,600 nonvotes statewide that could contain a legal vote; and 11. Whether it would be necessary to manually count all 6 million votes cast statewide in the presidential race to determine if any machine-counted votes should be disallowed because the presence of “hanging chad” or “dimpled chad” for a second candidate indicates that the ballot is in fact an overvote. 165. See Docket Sheet, Gore v. Harris, CV No. 00-2808 (Fla. 2d Cir. Ct. 2000), available at http://www.clerk.leon.fl.us/election_cases.html. Opposing parties in election contest litigation frequently dispute everything from the jurisdiction of the court and the standing of the contestant, to the admission of evidence, to every alleged irregularity in voting and every disputed vote that potentially can be included or excluded from the final tally. The parties in Gore v. Harris were different only in the greater number and magnitude of such disputes. 2001] COUNTS, RECOUNTS, CONTESTS 455 All such determinations were disputed166 and necessarily raised substantial questions of fairness and legality in the context of a judicial proceeding that could effectively declare a winner of the Presidency of the United States.167 Judge Lewis attempted to initially address some of these issues in an expedited fashion without proper opportunity for briefing, oral argument, or the submission of evidence.168 Fairly counting nonvotes in an election contest could come only in a protracted proceeding.169 The presidential electors from Florida could not have been timely determined170 through a legally sufficient election contest even if the Florida Supreme Court’s December 8 remedial order had been left undisturbed by the United States Supreme Court.171 166. Transcript of Motions Hearing, Gore v. Harris, 772 So. 2d 1243 (Fla. 2000) (No. SC00-2431), available at http://www.clerk.leon.fl.us/election_cases [hereinafter Transcript of Motions Hearing]. The motions hearing commenced at 8:35 p.m. and concluded at 11:39 p.m. 167. Technical problems, such as the ability to identify uncounted ballots, remained unsolved. See Bush v. Gore, 531 U.S. at 108 (indicating that any manual recount of only a portion of the ballots would require a reprogramming of the voting tabulation equipment to screen out undervotes; the distinct possibility existed under the Florida Supreme Court’s order that a statewide recount would be impossible to accomplish and that the circuit court could be asked to declare a winner on the basis of an admittedly incomplete recount). 168. See Transcript of Motions Hearing, supra note 166, at 59-65 (ruling of the Court). The ruling set in place procedures designed to begin the counting of “nonvotes or undervotes” immediately in all counties that had not previously conducted a manual review of such ballots. Legal issues, such as what constituted a legal vote, remained unsettled. 169. Historically, even relatively simple local election contests have taken months to resolve, sometimes requiring numerous hearings, appeals, and recounts of the same ballots. For example, appellate courts have been called upon to physically examine each contested ballot in an election contest to determine if the lower court has applied a correct standard. See, e.g., McIntyre v. Wick, 558 N.W.2d 347 (S.D. 1996). The prospect of the Florida Supreme Court or the United States Supreme Court possibly being asked to manually examine and to count thousands of disputed ballots on appeal is mind-boggling but not unlikely in a circumstance in which the inclusion or exclusion of only a few hundred votes could have determined the Presidency of the United States. 170. Merely manually examining the more than 170,000 nonvote ballots in a manner designed to assure fair and uniform treatment would have taken significant time and might not have yielded a convincing result. The ballot review project commissioned by eight media organizations began to organize its review of uncounted ballots in January 2001. Ford Fessenden, How the Consortium of News Organizations Conducted the Ballot Review, N. Y. TIMES, Nov. 12, 2001, at A17. The counting of ballots began in February and was completed in May. The results were available by September but were not released until November. The consortium utilized 153 ballot examiners. The consortium also benefited by avoiding the possibility of distractions caused by contesting attorneys and parties. Nevertheless, the review of ballots by the consortium took months and reached only inconclusive results given the various possible definitions of voter intent. Id. 171. The per curiam opinion of the United States Supreme Court indicated: it is obvious that the recount cannot be conducted in compliance with the requirements of equal protection and due process without substantial additional work. It would require not only the adoption (after opportunity for argument) of adequate statewide standards for determining what is a legal vote, and practicable procedures to implement them, but also orderly judicial review of any disputed matters that might arise. Bush v. Gore, 531 U.S. at 110 (emphasis added). 456 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 The United States Supreme Court effectively brought the Gore election contest to a close in a manner similar to which other much less monumental contests have been ended by state courts on the practical effects of upcoming elections.172 The perceived need for finality came to outweigh any need for absolute accuracy in the election outcome. Time, if it was ever sufficient, simply ran out. Three justices of the Florida Supreme Court would have ended the contest for these same reasons on December 8.173 It became the justifiable burden of a majority of the United States Supreme Court to do so three days later.174 IV. THE PRECEDENTIAL SIGNIFICANCE OF BUSH V. GORE A. Requirements for Equal Protection Although several writers have suggested that the decision in Bush v. Gore provides an advancement in voting rights,175 other contributors to this symposium have correctly questioned this conclusion, pointing out that the per curiam opinion itself indicates that “consideration is limited to the present circumstances.”176 One writer has cautioned that the decision will be of little precedential value because the Court itself did not take its holding seriously or engage in serious legal analysis.177 Moreover, this same writer notes that the holding in Bush v. Gore constitutes a strong break from the conservative majority’s usual approach to equal protection issues.178 The per curiam opinion in Bush v. Gore explains the scope of its holding as follows: 172. For example, as a practical matter it often is impossible to fully adjudicate recounts or election contests in party primary or runoff elections because candidates must be determined in adequate time to be placed on the runoff or general election ballot and to campaign for election. 173. Chief Justice Wells of the Florida Supreme Court explained in his dissenting opinion on December 8 that “it is inescapable that there is no practical way for the contest to continue for the good of this country and state.” Gore v. Harris, 772 So. 2d 1243, 1269 (Fla. 2000) (Wells, J., dissenting), rev’d sub nom. Bush v. Gore, 531 U.S. 98 (2000). Justice Harding, joined by Justice Shaw, similarly concluded in his dissent that the majority of the Florida Supreme Court was attempting to provide a remedy which would be “impossible to achieve” and which would “ultimately lead to chaos.” Id. at 1273 (Harding, J., dissenting). 174. Constitutional scholars correctly point out that there is a dearth of precedent for the application of the Fourteenth Amendment to the Florida Supreme Court’s remedial order interpreting and applying Florida state law. My experience as a practitioner, however, gave me a different perspective. I was not surprised that a majority of the U.S. Supreme Court was unwilling to permit the Presidency of the United States to potentially be determined by a fundamentally flawed state remedial order. A lack of precedent for federal court intervention in such a circumstance is no insurmountable barrier. 175. See, e.g., Issacharoff, supra note 3; Sunstein, supra note 6, at 769 (“On its face, the Court appears to have created the most expansive voting right in many decades.”). 176. Bush v. Gore, 531 U.S. at 109. 177. See, e.g., Hasen, supra note 5, at 387-90. 178. Id. at 390. 2001] COUNTS, RECOUNTS, CONTESTS 457 The recount process, in its features here described, is inconsistent with the minimum procedures necessary to protect the fundamental right of each voter in the special instance of a statewide recount under the authority of a single state judicial officer. Our consideration is limited to the present circumstances, for the problem of equal protection in election processes generally presents many complexities. The question before the Court is not whether local entities, in the exercise of their expertise, may develop different systems for implementing elections. Instead, we are presented with a situation where a state court with the power to assure uniformity has ordered a statewide recount with minimal procedural safeguards.179 Importantly, the Supreme Court’s per curiam opinion limits its scope to a remedial order of a judicial officer having authority to ensure uniformity in a statewide recount. By doing so, the opinion expressly limits itself to election contest types of proceedings. There is nothing in the opinion to suggest that it establishes a new equal protection requirement for statutes governing the recount structure, procedure, or standards of a state’s election process. In fact, the express wording of the opinion as quoted above appears to be intended to avoid implicating state and local control of elections. Read literally, the per curiam opinion holds only that a remedial order in a judicial election contest proceeding must assure “rudimentary requirements of equal treatment and fundamental fairness”180 for resolving disputes over the counting of votes throughout the election jurisdiction in question. Nothing is novel in this declaration. A contrary concept would have been shocking. The function of an election contest before a single judicial officer or court is to consider alleged irregularities that might affect the outcome of the election and to resolve disputes regarding the alleged inclusion of illegal votes or exclusion of legal votes in a uniform manner. State courts historically have insisted on fairness and equal treatment of voters and ballots in an election contest without needing to rely on the Fourteenth Amendment or even explicitly on comparable provisions of state constitutions.181 These state law princi- 179. Bush v. Gore, 531 U.S. at 109 (emphasis added). 180. Id. 181. See, e.g., In re Issue 27 Election of Nov. 4, 1997, 693 N.E.2d 1190 (Ohio 1998) (disallowing recount where manual counts by board of only the votes “for” a municipal proposition produced an increase in votes sufficient to change the outcome). The court reasoned that if irregularities (i.e., “hanging chad”) affected votes for the proposition, it was incumbent on the canvassing board to inspect the “no” votes as well. Id. The canvassing authority has the duty to ensure a fair and accurate recount. Id. at 1192-93; Greenwood Township Election Case, 25 A.2d 330, 332 (Pa. 1942) (revoking certification of recount on finding that erasures had caused invalidation of ballots in one precinct and ordering a check for similar tampering in fifteen other precincts); McIntyre v. Wick, 558 N.W.2d 347 (S.D. 458 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 ples of fairness have the same implications in the context of an election contest as those that arise from requirements of equal protection under Bush v. Gore.182 It was the majority of the Florida Supreme Court that strayed from these principles. B. Discerning Voter Intent The per curiam opinion in Bush v. Gore indicates: Florida’s basic command for the count of legally cast votes is to consider the “intent of the voter.” This is unobjectionable as an abstract proposition and a starting principle. The problem inheres in the absence of specific standards to ensure its equal application. The formulation of uniform rules to determine intent based on these recurring circumstances is practicable and, we conclude, necessary.183 This statement is made in the context of the recount of punch card ballots and a per curiam opinion limited to an election contest type of proceeding. Significantly, the per curiam opinion does not suggest that the guiding principle in vote counting—discerning voter intent—is inappropriate.184 Therefore, the voter intent principle prescribed by law in virtually all states is not directly threatened by the Bush v. Gore decision. The per curiam opinion does indicate, however, that “[t]he formulation of uniform rules to determine intent based on these recurring circumstances is practicable and, we conclude, necessary”185 and that the “search for [the voters’] intent can be confined by specific rules designed to ensure uniform treatment.”186 It is unclear whether the Court meant these conclusions to apply beyond the context of a judicial election contest type of proceeding. Nevertheless, whether required or not, rules or guidelines for discerning voter intent with punch card and other electronic counting systems are advisable in any circumstance in which ballots may be manually 1996) (challenging the inconsistent methods by which certain types of ballots (straight party ballots) were counted in different counties within a state legislative district). 182. Federal courts also have relied on fundamental fairness. See, e.g., Griffin v. Burns, 570 F.2d 1065 (lst Cir. 1978) (overturning Rhode Island Supreme Court decision as fundamentally unfair in interpreting state law to not allow counting of absentee ballots in a primary election). 183. Bush v. Gore, 531 U.S. at 105-06 (citation omitted). 184. Several of the contests in Congress have involved an effort by a house of that body to discern voter intent on ballots. For example, in 1925, the U.S. Senate ordered 900,000 ballots transported to Washington, D.C., in a dispute over the election of a senator from Iowa; the dispute centered on ballots allegedly counted or thrown out based on “extraneous” marks on the ballots. See H.R. REP. No. 99-58 (1985) (providing in disputed congressional election for the counting of all ballots from which the intent of the voter may be discerned without regard to technicalities). 185. Bush v. Gore, 531 U.S. at 106. 186. Id. 2001] COUNTS, RECOUNTS, CONTESTS 459 reviewed either before187 or after being machine-counted. Many states have been operating under such rules existing either in statute or through regulations or guidelines promulgated by the state’s election officer.188 Even when rules or guidelines are present, the ultimate standard often remains the intent of the voter. For example, section 127.130 of the Texas Election Code was suggested in Florida as an example of an instance in which state law established specific rules for determining what constitutes a vote without relying on the standard of “voter intent” as used by the Florida courts. Subsection (d) of section 127.130 indicates that a vote on a punch card ballot “may not be counted unless: (1) at least two corners of the chad are detached, (2) light is visible through the hole, [or] (3) an indentation on the chad from the stylus or other object is present and indicates a clearly ascertainable intent of the voter to vote.”189 The subsection goes on, however, to qualify these rules by allowing an exception when “the chad reflects by other means a clearly ascertainable intent of the voter to vote.”190 The section further subsumes all of these more specific substandards within the “intent of the voter” standard by expressly indicating that nothing in subsection (d) supersedes “any clearly ascertainable intent of the voter.”191 This Texas statute reflects the practical reality that it is not possible to fully prescribe rules for controlling a determination of voter intent because the difficulties in ascertaining voter intent are not limited to only punch card systems or to certain identifiable circum- 187. Some states review ballots before machine processing to determine if ballots have been mismarked and may not be read accurately by the counting machine. See, e.g., N.D. CENT. CODE § 16.1-15-09 (2001). Machine-readable duplicate ballots are prepared according to the voter’s intent shown on the original ballot. The original ballots are maintained so that they can be examined if necessary in a later manual recount or contest. 188. See 2001 ELECTION ADMIN. SURVEY, supra note 9. The National Commission of Election Standards and Reform adopted a series of preliminary recommendations on April 22, 2001. Essentially, the Commission recommended that elections remain under state and local control with federal regulatory control. See National Association of Counties, COUNTY NEWS, May 7, 2001, at 1-3. The Commission indicates that the federal government can best address the weakness of the system by funding improvements in equipment and administration, sponsoring research, and disseminating information. Id. The Commission had been created in January, 2001, by the National Association of County Officials (NACO). Id. In addition to recommendations for funding, the Commission recommended that states: (1) determine what constitutes a vote for each type of equipment; (2) establish clear recount procedures; (3) work to remove partisanship by election officials; and (4) provide adequate time to complete a canvas of an election prior to any recount or contest. Id. Most state legislation efforts at reform in 2001 failed. Id. at 2. 189. TEX. ELEC. CODE ANN. § 127.130(d)(1)-(3) (Vernon 2000) (emphasis added). 190. Id. § 127.130(d)(4). 191. Id. § 127.130(e). 460 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 stances.192 For example, reports from Florida have contrasted the apparent accuracy of an optical scanner for discerning a voter’s intent to vote as compared to the equipment counting punch card ballots. Nevertheless, federal government reports warn that ballots can be unreadable in optical scanners even when the voter’s intent is clearly ascertainable from the ballot by manual examination.193 It appears likely that many of the ballots treated as nonvotes in Florida’s optical scan counties would have been recognized as votes through a manual recount.194 Even when rules exist for discerning voter intent, objective and reasonable counters may disagree on the application of those standards to particular ballots. For example, in Delahunt v. Johnston195 the Supreme Court of Massachusetts found that the judge of the trial court had applied the correct standard for discerning voter intent but that “[o]n balance, we are slightly more willing to find an intention expressed on [the] ballots where the trial judge ruled there was none.”196 Just as different umpires in baseball call balls and strikes differently at the extremes of the prescribed strike zone, different judicial and administrative officials may include or exclude votes from ballots left uncounted by counting equipment even when such decisions are subject to controlling rules. Statutory or administrative rules are desirable for discerning the intent of voters with a manual review of ballots. The decision in Bush v. Gore properly places an emphasis on the utility of such rules. Nevertheless, even with such rules, subjective judgments are unavoidable. The safeguard under election circumstances is that any irregularities in how votes are counted ulti- 192. The potential for failing to accurately record voter intent exists under any voting system. The potential for error varies according to many factors, including the type and condition of the counting equipment. 193. See SALTMAN, supra note 10, at 37. 194. For example, the Orlando Sentinel reported that Gore would have picked up a net increase of 203 votes in Orange County if the undervote ballots left uncounted by the county’s optical scanning equipment had been manually counted. Roger Roy & Mike Griffin, Orange Tally Nets 203 Votes for Gore, ORLANDO SENTINEL, Feb. 10, 2001, at A1. This undercount problem with optical scanners in Florida is consistent with the experience of other states such as Texas, where 0.63 percent of the ballots in counties using optical scanners for the 2000 election were found to have nonvotes for President. See Texas Secretary of State Henry Cuellar, Report to the Texas Legislature (January 2001) (on file with the Florida State University Law Review). By comparison, the new Direct Record Electronic (DRE) systems in Texas had an even higher percentage of 0.89 percent of undervotes. Id. 195. 671 N.E.2d 1241, 1243 n.2 (Mass. 1996); see also Duffy v. Mortenson, 497 N.W.2d 437, 439-40 (S.D. 1993) (concluding that a ballot with indented chad not counted by the trial court provides evidence of clear voter intent and, when counted, results in a tie between the candidates). 196. Delahunt, 671 N.E.2d at 1243 n.2. 2001] COUNTS, RECOUNTS, CONTESTS 461 mately are reviewable by a single impartial arbiter through an election contest. C. The Single Impartial Arbiter The essential requirement for equal and fair treatment is the presence of a fair adversarial process before a single, impartial arbiter with responsibility for adjudicating contested issues and determining whether the election result reflects the will of the voters. As Justice Stevens recognized, the Supreme Court’s concerns in Bush v. Gore that different county canvassing boards used differing standards for determining voter intent should have been alleviated, if not eliminated, “by the fact that a single impartial magistrate will ultimately adjudicate all objections arising from the recount process.”197 If time had permitted a prosecution of the Gore election contest in Florida courts on a normal time frame, it is likely that those adversarial proceedings would have eventually prevented the disparities identified by the United States Supreme Court. It is because of the extraordinary circumstances that existed on December 8 that the Florida Supreme Court’s majority opinion strayed so far from assuring the fairness required in an election contest. State laws generally provide that recounts, including manual recounts, in state and federal elections are ultimately subject to resolution before a single magistrate, appellate court, state officer, or institution with authority to resolve disputes or to adjudicate contested issues. Recount disputes in an election for a subcounty or county office are initially resolved by the local canvassing board or recount committee with authority to determine what ballots should lawfully be included or excluded for precincts within the entire jurisdiction. The judicial election contest provides a check on the exercise of this authority by making the outcome of the election subject to de novo challenge before a single impartial arbiter subject to appellate review. State or federal laws also provide such an arbiter for multicounty and statewide elections. The specific officer, court, or institution varies among the states and according to the elective office in question. For example, federal elections for Congress are initially subject to potentially disparate local recount procedures or practices, but it is Congress itself that ultimately has authority to resolve any contested issues, including the inclusion or exclusion of disputed ballots from throughout the election jurisdiction. Congress has exercised this authority on several occasions in the past through an adversarial process in which committees conduct hearings, consider evidence, and 197. Bush v. Gore, 531 U.S. 98, 126 (2000) (Stevens, J., dissenting). 462 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 recommend a ruling by the affected legislative body.198 State legislative elections similarly are often subject to local recounts, but it is the state legislative bodies that ultimately act to resolve contested issues. These state institutions also generally act through adversarial proceedings with an opportunity for briefs, hearings, and the presentation of evidence. As in Congress, this process occurs before a committee that then makes recommendations to the affected legislative body.199 For nonlegislative, multicounty, or statewide elections, the states have created numerous differing alternatives for adjudicating contested issues affecting the outcome of a general election. For example, in Texas an election contest over a state office, such as Governor, is resolved by the state legislature.200 Curiously, an election contest over presidential electors in Texas is decided after an evidentiary hearing by the Governor.201 In other words, if Texas law had been applicable in Florida in 2000, it would have been Jeb Bush who would have been designated by statute to hear the Gore election contest. Primary elections present a particular problem because they are intraparty affairs and because there is limited time available between a primary election and a general election to resolve disputes in time to assure that the winning person in the primary appears as the party’s candidate on the general election ballot.202 Since local government election officials generally are responsible for conducting such primary elections, recounts occur in a similar although sometimes expedited basis. Contests, on the other hand, may end up in state court or before local or state political party officials or committees. State and federal laws effectively provide a single arbiter with authority to adjudicate disputes over the counting of ballots. Combined with the opportunity for candidates and their partisan representatives to observe recounts, the availability of a single arbiter provides a means for achieving fairness and uniformity in the counting of ballots despite the possibility of some initial disparity among county governments. 198. ANNE M. BUTLER & WENDY WOLFF, U.S. HISTORICAL OFFICE, UNITED STATES SENATE: ELECTION, EXPULSION AND CENSURE CASES 1793-1990 (1995). On several occasions the U.S. Senate has subpoenaed all ballots for manual review in Washington, D.C. As might be expected in such circumstances, a determination regarding the inclusion or exclusion of particular ballots or categories of ballots became a partisan fight. Id. 199. Junnell, supra note 50, at 1095. 200. TEX. ELEC. CODE ch. 242 (2000). 201. Id. ch. 243. 202. Often the issue in a general primary is a determination of whether a runoff election may be required and which candidates qualified for the runoff. The need for a runoff further shortens the time for resolution of recounts and election contests. 2001] COUNTS, RECOUNTS, CONTESTS 463 There is not, however, any universally accepted means for assuring that this final arbiter is impartial. Decisions left to state or federal legislative bodies, or possibly to the Governor of a state, may, as shown by the events in Florida, become viewed as partisan rather than impartial decisions. The recent experience in Florida also shows, however, that circumstances can occur under which not even the highest courts of a state or of the nation are above suspicion by some as acting for partisan reasons in an election dispute. V. GORE’S LOSING LEGAL STRATEGY It is very possible that there is nothing that Albert Gore could have done after November 7 to prevail as the certified winner of Florida’s presidential electors. Bush might have prevailed under any recount or election contest scenario simply because he had the most votes. Or, Bush might have prevailed because the combination of a Republican Governor, a Republican-controlled state legislature, a Republican chief state election officer, and ultimately a U.S. Supreme Court dominated by persons identified with the Republican Party may ultimately have been too much for Gore to overcome regardless of whether he received the most votes. Nevertheless, the legal strategy followed by the Gore attorneys significantly lessened Gore’s opportunity to prevail.203 The fundamental flaw in Gore’s legal strategy beginning November 8 was its failure to appreciate the difference in law and dynamics between an administrative recount of votes and an election contest. If the necessary uncounted votes were there, an appropriate recount could find them. If the campaign was left to pursuing an election contest, however, the chances of success were essentially nonexistent in the time available before the deadlines set by federal law for the selection of presidential electors. This difference between a vote recount and an election contest was not readily apparent on examination of the relevant Florida statutes. The election contest provisions of section 102.168, Florida Statutes, appeared seductively simple. They expressly acknowledged the “rejection of a number of legal votes sufficient to change or place in doubt the result of the election”204 as a ground for an election contest 203. I admit as a trial attorney that I am reluctant to offer these critical comments about the legal strategy followed by the Gore legal team in Florida. The Gore legal team was outstanding. Nevertheless, if the strategies of Lee at Gettysburg and Napoleon at Waterloo are susceptible to reevaluation in light of the outcome in those battles, then it would also seem appropriate to subject Gore’s strategy in Florida to reevaluation. Despite my admiration for the attorneys in question, I believe that the flaws in the Gore strategy existed at least in part because those outstanding attorneys lacked experience in the world of election administration, recounts, and the litigation of election contests. 204. FLA. STAT. § 102.168(3)(c), (2000). 464 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 and authorized the circuit court judge to “fashion such orders as he or she deems necessary.”205 It is reasonable on the face of this statute to expect that a showing by Gore of almost any number of potentially uncounted votes (much less the 9,000 or more undervotes in MiamiDade County) would have led to a counting of enough undervotes to erase the miniscule Bush lead. Based on this assumption about section 102.168, success in the administrative recount process before the local canvassing boards, while desirable, was not essential to Gore’s eventual triumph. The apparent promise of section 102.168, however, was illusory. As in other states, Florida courts historically have been reluctant to interfere in elections, particularly to overturn a certified election outcome.206 Therefore, as in other states, the Florida courts have created a presumption in favor of the validity of the certified results and evidentiary burdens for a contestant to overcome that do not necessarily appear on the face of the contest statute.207 At a minimum, Gore had to prove that legal votes had been excluded and that, if included, these legal votes would change or place in doubt the outcome of the election itself, based on the inclusion or exclusion of votes statewide, not just in selected counties. Moreover, even simple local election contests generally are marked by intense legal battles. Given the im205. Id. § 102.168(8) (amended 2001). 206. This reluctance was evident even in 2000 in the Florida Supreme Court’s rejection of challenges to the outcome of the presidential election based on the Palm Beach County butterfly ballot and the absentee balloting in Seminole and Martin Counties. See infra note 207. 207. In his dissent in Gore v. Harris, Chief Justice Wells explained that “Historically, this Court has only been involved in elections when there have been substantial allegations of fraud and then only upon a high threshold because of the chill that a hovering judicial involvement can put on elections.” 772 So. 2d 1243, 1263 (Fla. 2000) (Wells, C.J., dissenting); see, e.g., Smith v. Tynes, 412 So. 2d 925 (Fla. 1st DCA 1982) (indicating that it is not enough for a contestant to show a reasonable possibility that election results could be altered by irregularities, rather a reasonable probability that the results would have been changed must be shown). This threshold of “reasonable probability” was created by the Florida courts and did not appear explicitly in the election statutes. See also Boardman v. Esteva, 323 So. 2d 259, 268 (Fla. 1975) (indicating that where the record does not show that votes were illegal “the presumption of the correctness of the election officials’ returns stands”); Krivanek v. Take Back Tampa Political Comm., 625 So. 2d 840, 844-45 (Fla. 1993) (quoting Boardman v. Esteva, supra, for the proposition that “[i]t is certainly the intent of the constitution and the legislature that the results of elections are to be efficiently, honestly and promptly ascertained by election officials to whom some latitude of judgment is accorded, and that courts are to overturn such determinations only for compelling reasons when there are clear, substantial departures from essential requirements of law”). In Gore v. Harris, the majority of the Florida Supreme Court concluded that the “reasonable probability” standard was no longer applicable under section 102.168 as amended in 1999. 772 So. 2d at 1255. Dissenting Justices Harding and Shaw agreed with the majority that the reasonable probability requirement had not survived the 1999 amendments. Id. at 1271 (Harding, J., dissenting). Nevertheless, these two justices concluded that section 102.168 still required a contestant to show “that the number of legal votes rejected by the canvassing boards is sufficient to change or place in doubt the result of this statewide election” and that Gore had failed to carry this burden. Id. (emphasis in original). 2001] COUNTS, RECOUNTS, CONTESTS 465 portance of this particular election, it was safe to assume on November 8 that any election contest would be a monstrous, complex, and chaotic affair with multiple parties and myriad legal issues. Realistically, no candidate was likely to win such a contest in the time available. Therefore, success in the recount was Gore’s only opportunity after November 7 to prevail in the election. The automatic statewide recount produced 1,753 additional net votes for Gore. State law provided a means of obtaining a further recount, including a manual recount of the nonvote ballots, through the protest process. Although this process was certain to be vigorously contested, it was an established procedure bound by a deadline that would expire before presidential electors were to be selected or would cast their ballots. As a winner of that recount, Gore would have been entitled to certification as the winner of Florida’s electors and would have enjoyed both the legal and political benefit of having finality on his side. Such a scenario would have forced Bush to rely on an election contest of disputed ballots, a challenge in federal court, or an appeal to the state legislature, with the considerable legal and public relations burden of overturning the official outcome of the election. The dynamic of finality would have been on Gore’s side. Despite the critical nature of the recount, Gore’s strategy toward the recount started half-heartedly, with requests for manual recounts planned only for Palm Beach and Volusia counties.208 Experienced recount attorneys urged that manual recounts be sought throughout Florida. After all, as explained by The Recount Primer: If a candidate is behind, the scope should be as broad as possible, and the rules for the recount should be different from those used election night. A recount should be an audit of the election to insure the accuracy and honesty of the results.209 In other words, the trailing candidate is advised to look for voting or tabulation errors wherever they might exist. Eventually Gore’s attorneys also requested manual recounts in Broward and Miami-Dade Counties,210 but no timely request was filed in the other sixty-three counties.211 208. DEADLOCK, supra note 53, at 71, 78. Gore’s selective requests for recount were characterized by some as “cherry-picking,” “mining for votes,” or “gamesmanship.” Touchston v. McDermott, 234 F.3d 1133, 1143, 1150, 1152 (11th Cir. 2000) (Tjoflat, J., dissenting). 209. DOWNS, supra note 28, at 5. 210. Attorneys for Gore were successful in obtaining a manual recount in at least one other county. See DEADLOCK, supra note 53, at 158. 211. One reason for not requesting a manual recount except in these four counties was that the attorneys for Gore foresaw a difficulty in timely requesting manual recounts statewide since a written protest would have to be filed in each county. Such a county-by- 466 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:425 Gore found the Florida Supreme Court receptive to a manual counting of ballots to discern voter intent on ballots on which no vote previously had been recorded by the counting machines. On November 16, the court ordered that the manual recounts could continue212 despite the statutory deadline for the receipt of county election results by the Secretary of State. The next day, the court sua sponte enjoined the Secretary of State from certifying the election results.213 At oral argument on November 20, the members of the court evidenced their desire for counting all votes and asked the Gore attorneys how long it would take to complete the manual recounts. Tellingly, Gore’s attorneys had no answer. The Florida Supreme Court extended the deadline to November 26.214 Nevertheless, only one county, Broward, was able to complete its recount within the period of this extension. Therefore, on November 26 the Gore attorneys confronted the worst possible situation. The extension essentially had been wasted, and Bush had been certified as the official winner. Gore’s attorneys now had ten fewer days to prepare and successfully prosecute an election contest. Gore’s failure to appreciate the difference in law and dynamics between a recount and an election contest also pervaded the candidate’s strategy in the election contest. Relying on the position that section 102.168, Florida Statutes, required only that Gore show that there were enough uncounted votes in Miami-Dade to “place in doubt the [outcome] of the election,”215 the apparent strategy was to treat the election contest as essentially a continuation of the aborted recounts and to push the case through Judge Sauls to the Florida Supreme Court as quickly as possible. As a result, Gore presented only two witnesses and made no apparent effort to place the election itself in doubt beyond the potential of added votes from a few selected counties. This lack of a winning election contest theory or a compelling evidentiary record ultimately sealed Gore’s fate. county process is manageable. This author in 1978 oversaw the first statewide recount in the history of Texas. State law at the time provided for four to six different methods for obtaining a recount based on the type of balloting or voting equipment used within each county. To obtain recounts statewide in the gubernatorial general election, we simply prepared a generic form for each type of county voting system and sent the appropriate form to a party representative within each county with written instructions for how to particularize the form to the specific county and how to proceed. The logistical problems of timely filing recount requests in most of Texas’s 254 counties in 1978 were great but not overwhelming. In an age of e-mail, fax, and cell phones, the logistical problems of filing such requests in Florida surely are more manageable. 212. Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d 1220, 1225 (Fla. 2000), vacated sub nom. Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000). 213. Id. at 1227. 214. Id. at 1240. 215. FLA. STAT. § 102.168(3)(c) (2000). 2001] COUNTS, RECOUNTS, CONTESTS 467 If any bona fide opportunity existed for Al Gore to prevail in Florida after November 7, it was through the manual recount of ballots by the local government canvassing boards. It was in the interest of Bush and Republican Party officials to delay and to confuse those recounts. Nevertheless, the chance of success for Gore was greater at the recount stage than in an election contest. Gore’s legal strategy failed to recognize the necessity of winning the recount battle. VI. CONCLUSION The events following the November 7, 2000, presidential election were extraordinary. The legal opinions resulting from those events, however, are of limited precedential value in part because they were written without time for a full development of the facts or the law. The decision of the United States Supreme Court in Bush v. Gore is not novel for its notion of fairness and equal treatment in the context of an election contest. State election contests before a single impartial arbiter historically have been the means by which disputes over the outcome of an election have been resolved and meaningful irregularities in an election have been corrected. Fairness in the treatment of candidates and voters is an essential principle of such contests. Any contrary holding would have been shocking and unacceptable. In Gore v. Harris, the majority of the Florida Supreme Court lost sight of these fundamental principles. Although the intervention of the United States Supreme Court in the Florida election controversy may be unprecedented, its decision to bring an end to the recount ordered by the Florida Supreme Court was consistent with the historic need for finality in state election contest proceedings and was justified under the existing circumstances. It was Gore’s flawed legal strategy that reduced his opportunity for winning the election in Florida and that ultimately brought his legal battle to an unsuccessful end. A PLACE AT THE TABLE: BUSH V. GORE THROUGH THE LENS OF RACE* SPENCER OVERTON** I. COUNTING VOTES AND ASSUMPTIONS ABOUT DEMOCRACY................................ II. MERITOCRACY THROUGH THE LENS OF RACE ..................................................... A. Race Exposes the Shortcomings of the Merit-Based Vision’s Individualized Focus .................................................................................... B. Race Exposes Particular Expressive Components of Merit-Based Vision .. III. MERIT AND THE EXCLUSION OF US ALL .............................................................. CONCLUSION ........................................................................................................ 473 479 479 484 489 491 In the 2000 presidential election, African Americans made up only 16% of the voting population in Florida but cast 54% of the ballots rejected in automatic machine counts (“machine-rejected ballots”).1 Across the state, automatic machines rejected 14.4% of the ballots cast by African Americans, but only 1.6% of the ballots cast by others.2 Racial disparities appeared even when the same voting technol* Cf. Langston Hughes, I, Too, in THE NORTON ANTHOLOGY OF AFRICAN AMERICAN LITERATURE 1258 (Henry Louis Gates Jr. & Nellie Y. McKay eds., 1997) (“I am the darker brother. They send me to eat in the kitchen when company comes. . . . Tomorrow, I’ll be at the table when company comes. Nobody’ll dare say to me, ‘Eat in the kitchen,’ then. . . . I, too, am America.”). ** Acting Professor of Law, University of California, Davis. Several people read earlier drafts of this Essay and provided helpful comments, including Richard Banks, Roger Fairfax, Floyd Feeney, Heather Gerken, Lani Guinier, Bill Hing, Kevin Johnson, Tom Joo, Pamela Karlan, Kenneth Mack, Cynthia Overton, Leslie Overton, Marc Spindelman, Madhavi Sunder, and Tobias Wolff. This Essay also benefited from the author’s conversations with Diane Amann, Holly Doremus, Frank Michelman, Martha Minow, Reggie Oh, Joseph Singer, and Terry Smith, as well as from the outstanding research assistance of Russell Johnson and Johanna Berta. Special thanks to Jim Rossi, Richard Hasen, Jason Kellogg, and Amanda Keener. 1. U.S. COMM’N ON CIVIL RIGHTS, VOTING IRREGULARITIES IN FLORIDA DURING THE 2000 PRESIDENTIAL ELECTION 2 (2001) [herinafter U.S. COMM’N ON CIVIL RIGHTS] (“Approximately 11 percent of Florida voters were African American; however, African Americans cast about 54 percent of the 180,000 spoiled ballots in Florida during the November 2000 election based on estimates derived from county-level data.”); Josh Barbanel & Ford Fessenden, Racial Pattern in Demographics of Error-Prone Ballots, N.Y. TIMES, Nov. 29, 2000, at A25 (reporting that “black voters made up 16% of the vote on Election Day”). In compiling the data, the U.S. Commission on Civil Rights relied on the percentage of registered voters in Florida who were African American (11%), U.S. COMM’N ON CIVIL RIGHTS, supra, at 2, whereas The New York Times relied on the percentage of those who voted on election day who were African American (16%), Barbanel & Fessenden, supra, at A25. 2. Katharine Q. Seelye, Divided Civil Rights Panel Approves Election Report, N.Y. TIMES, June 9, 2001, at A8 (reporting on a study conducted by Allan J. Lichtman, a history professor at American University and an elections expert); U.S. COMM’N ON CIVIL RIGHTS, supra note 1, at 2 (showing that across Florida, ballots cast by African Americans were almost ten times more likely to be rejected than the ballots of whites); see also U.S. to Look Into Possible Irregularities at the Polls, CHIC. TRIB., Dec. 4, 2000, at 9 (reporting on a computer analysis finding that “the more black and Democratic a precinct, the more likely a high number of presidential votes was not counted”). Studies have shown that racial disparities in uncounted votes also exist outside of Florida. See, e.g., David Stout, Study Finds Ballot Problems are More Likely for Poor, N.Y. TIMES, July 9, 2001, at A9 (reporting on a 469 470 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:469 ogy was used. For example, counting machines rejected punch card ballots in predominantly African-American precincts in Miami-Dade County at twice the rate they rejected ballots in predominantly Latino precincts, and four times the rate they rejected ballots in predominantly white precincts.3 In their discussions of Bush v. Gore,4 legal academic commentators have not grappled with the significance of the racial disparities reflected in machine-rejected ballots. Despite the fact that the U.S. Supreme Court permanently halted the manual count of these ballots,5 doctrinal analysis employing the facts as framed by the Justices has, by and large, commanded the most attention. study conducted by the U.S. House of Representatives Government Reform Committee that found “the votes of poor people and members of minorities were more than three times as likely to go uncounted in the 2000 presidential election.”). 3. Barbanel & Fessenden, supra note 1, at A25. The impact of these differences on the outcome will never be known but their potential magnitude is evident in Miami-Dade County, where predominantly black precincts saw their votes thrown out at twice the rate as Hispanic precincts and nearly four times the rate of white precincts. In all, [one] out of [eleven] ballots in predominantly black precincts were rejected, a total of 9,904. Id.; see also Kim Cobb, Black Leaders Want Action on Florida Vote Complaints, HOUS. CHRON., Nov. 30, 2000, at A24 (“U.S. Rep. Corrine Brown, D-Jacksonville, said that 16,000 of the 27,000 ballots left uncounted in Duval County were from predominantly black precincts.”); U.S. to Look into Possible Irregularities at the Polls, supra note 2, at 9 (“In MiamiDade, the state’s most populous county, about 3 percent of ballots were excluded from the presidential tally. But in precincts with a black population of 70 percent or more, about 10 percent were not counted.”). The disparity between African Americans and whites with regard to machine-rejected ballots was higher than racial disparities in the use of punch card technology. Barbanel & Fessenden, supra note 1, at A25 (observing that “64[%] of the state’s black voters live in counties that used the punch cards while 56[%] of whites did so.”). But see Stephen Ansolabehere, Voting Machines, Race, and Equal Protection, 1 ELECTION L.J. (forthcoming 2001) (arguing that nationally, no significant correlation exists between race and punch card machine-rejected ballots). 4. 531 U.S. 98 (2000). 5. The discussion of race has been more extensive outside of the legal academy. See, e.g., Common Cause v. Jones, No. 01-03470 (C.D. Cal. filed Apr. 24, 2001) (alleging disparate voting procedures disadvantage racial minorities in violation of equal protection); U.S. COMM’N ON CIVIL RIGHTS, supra note 1 (2001); Ansolabehere, supra note 3 (political scientist’s study of race and voting technology); Barbanel & Fessenden, supra note 1, at A25. A few legal commentators have touched upon the relevance of race. See, e.g., Heather Gerken, New Wine in Old Bottles: A Comment on Richard Hasen’s and Richard Briffault’s Essays on Bush v. Gore, 29 FLA. ST. U. L. REV. 407, 422-23 (2001) (“The Court’s failure to wrestle with these questions—what does equality mean, and how far should we go to attain it when the twin problems of race and poverty permeate our democratic structures?— gives an unwarranted patina of legitimacy to the election system.”); Pamela S. Karlan, Nothing Personal: The Evolution of The Newest Equal Protection from Shaw v. Reno To Bush v. Gore, 79 N.C. L. REV. 1345, 1366-67 (2001). There is credible evidence that systems that disproportionately reject votes both have a racially disparate impact and are more often used in the populous jurisdictions in which minority voters are concentrated. Thus, the newest equal protection once again vindicates the interests of middle-class, politically potent voters, while ignoring the interests of the clause’s original beneficiaries. Id.; Richard A. Posner, Florida 2000: A Legal and Statistical Analysis of the Election Deadlock and the Ensuing Litigation, 2000 SUP. CT. REV. 1, 14 (engaging in a statistical analy- 2001] A PLACE AT THE TABLE 471 Without a consideration of race, however, the conversation about Bush v. Gore remains woefully incomplete. Politics and race in the United States have characteristics that sometimes overlap.6 Issues of racial identity and racial differences necessarily evoke questions of representation in the political process, particularly among groups that have been historically excluded. Because of the unique role of race in American politics, an examination of race yields important insights that might otherwise go unnoticed. While this short Essay does not comprehensively analyze all of the components of race in Bush v. Gore, the piece does use race to address normative assumptions about democracy embedded in the opinion.7 The use of a racial framework shows how these assumptions adversely impact racial minorities and other Americans as well. Professor Briffault acknowledges that the five U.S. Supreme Court Justices who voted to discontinue manual counting of the ballots in Bush v. Gore deviated from the Court’s trend of including previously excluded groups in the political process.8 In a similar spirit, Professor sis and observing that “the larger the black population and the lower the literacy level, the higher the incidence of undervotes even after other factors are taken into account.”). 6. Cf. LANI GUINIER & GERALD TORRES, THE MINER’S CANARY (forthcoming Feb. 2002) (manuscript at 9, on file with author) (“[W]e begin by reclaiming the idea of race from its current, artificially limited conception as an exclusively individualistic form of personal identity. Rather than see race as merely denoting the biological facts of ancestry, we seek to deploy race as a proxy for political status.”); Vikram David Amar & Alan Brownstein, The Hybrid Nature of Political Rights, 50 STAN. L. REV. 915, 976 (1998) (analyzing recent jurisprudence, and observing that “[t]he Supreme Court has dismissed the group dimension of political rights only, it appears, where race is involved”); Jerome McCristal Culp, Jr., Colorblind Remedies and the Intersectionality of Oppression: Policy Arguments Masquerading as Moral Claims, 69 N.Y.U. L. REV. 163, 191 (1994) (“The race of the voters matters in North Carolina precisely because the black voters have voted consistently against the racial politics of North Carolina’s Congressional Club and Republican Party.”); Lani Guinier, [E]racing Democracy: The Voting Rights Cases, 108 HARV. L. REV. 109, 130 (1994) (“In other words, minority group representation is not purely cultural, historical, or biological; it also has a political component. Group members may identify collectively along a common axis and organize to promote common interests in ways similar to other political associations.”); Martha Minow, Not Only for Myself: Identity, Politics, and Law, 75 OR. L. REV. 647, 697 (1996) (asserting that “[i]dentity politics have been crucial and perhaps inevitable responses to perceived oppressions”). 7. Cf. BELL HOOKS, FEMINIST THEORY FROM MARGIN TO CENTER xvi (2d ed. 2000) (asserting that a view from the “margin” allows one to understand both the center and the margin of society, and provides a sense of wholeness); Charles R. Lawrence III, Two Views of the River: A Critique of the Liberal Defense of Affirmative Action, 101 COLUM. L. REV. 928, 950-51 (2001) (“Critics of liberal theory, including critical race theorists, have offered another way to think about promoting equality and human dignity, one that reflects the perspective of the subordinated.”); Spencer Overton, Fannie Lou Hamer Wouldn’t Like This, L.A. TIMES, Mar. 29, 2001, at B11 (discussing the Fannie Lou Hamer standard, which considers the campaign finance system from the perspective of a poor woman of color like Fannie Lou Hamer). 8. Richard Briffault, Bush v. Gore as an Equal Protection Case, 29 FLA. ST. U. L. REV. 325, 347-49, 372 (2001) (describing Bush v. Gore as quite different from the Court’s earlier inclusionary equal protection cases). Professor Gerken observes that the majority mistakenly believes itself to be agnostic, and she identifies many of the problems that re- 472 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:469 Hasen asserts that Bush v. Gore’s break from past cases may “ease the way for future Supreme Court majorities to pursue their own visions of political equality without much thought about whether that vision is supported by existing case law.”9 This Essay agrees with Professors Briffault and Hasen to the extent that they suggest that Bush v. Gore rejected more inclusionary assumptions about democracy articulated in earlier cases, but also asserts that the Court embraced merit-based assumptions that conditioned political recognition on an individual voter’s capacity to produce a machine-readable ballot.10 The use of race reveals how both the focus on individual responsibility and the expressive harm of exclusion that accompany the merit-based vision pose unique problems in the context of voting. Though some might argue that taking race into consideration is inappropriate in a “colorblind” society,11 a consideration of race need not entail the employment of a “race card” that trumps all other concerns and singularly insists on race-specific solutions. Instead, just as decisionmakers balance such concerns as individual rights, economic efficiency, and general welfare,12 race can be used as one analytical tool to be considered in conjunction with other factors. Some might assert that race is irrelevant to an analysis of the machinerejected ballots, preferring instead to attribute responsibility to voter inexperience, voter illiteracy, and substandard voting equipment in particular jurisdictions.13 These explanations, however, are not prepolitical or randomly distributed throughout society but disproporsult from the Court’s failure to explicitly anchor its decision in a concrete, normative theory of democracy. Gerken, supra note 5, at 415. This Essay attempts to identify some of the unstated assumptions underlying the perspective of the majority per curiam and its defenders and begins to address some of the thornier normative issues embedded in the decision. 9. Richard L. Hasen, Bush v. Gore and the Future of Equal Protection Law in Elections, 29 FLA. ST. U. L. REV. 377, 380 (2001). 10. See infra Part I (discussing meaning of inclusionary and merit-based visions of democracy). 11. See City of Richmond v. J.A. Croson Co., 488 U.S. 469, 521 (1989) (Scalia, J., concurring) (citing Plessy v. Ferguson, 163 U.S. 537, 559 (1896) (Harlan, J., dissenting)), for the proposition that “our Constitution is color-blind.” But see Regents of Univ. of Cal. v. Bakke, 438 U.S. 265, 336 (1978) (Brennan, White, Marshall, & Blackmun, JJ., dissenting in part and concurring in part) (noting that “no decision of this Court has ever adopted the proposition that the Constitution must be colorblind”). 12. See JOSEPH SINGER, PROPERTY LAW: RULES, POLICIES, AND PRACTICES 279-82 (2d ed. 1997) (describing the use of various policy considerations as analytical tools in lawmaking). 13. See, e.g., Ansolabehere, supra note 3 (arguing that nationally, no significant correlation exists between race and punch card machine-rejected ballots and that racial disparities are explained by a higher percentage of less reliable punch card technology in AfricanAmerican precincts); Abigail Thernstrom & Russell G. Redenbaugh, The Florida Election Report: Dissenting Statement, at http://www.manhattan-institute.org/html/final_ dissent.htm (asserting that the problems encountered during the Florida election were caused by bureaucratic inefficiencies, inexperienced voters, illiterate voters, substandard voting technology, and other issues unrelated to race). 2001] A PLACE AT THE TABLE 473 tionately impact certain populations due in part to past statesponsored racial discrimination.14 A consideration of race allows scholars and legal decisionmakers to avoid the pitfalls of the “colorblind card,” an ideological extreme that mechanically trumps historical considerations, silences discussion, removes relevant issues from the table, and ignores important problems. Part I of this Essay reviews two opposing visions of democracy that emerged in Bush v. Gore. The Florida Supreme Court’s more inclusionary vision prompted it to order that the ballots rejected by machines be counted manually, while the U.S. Supreme Court’s more merit-based vision motivated it to prohibit a manual count of the imperfectly marked ballots. Part II uses race to reveal many of the shortcomings of the merit-based vision of democracy. Although the Court’s facially neutral, merit-based criteria focus on individual responsibility, they interfere primarily not with individual rights, but with the ability of groups of voters like African Americans to identify with one another as a political community, to create alliances with others of different backgrounds, and to use the vote to enact political change. Further, the lens of race exposes how merit-based criteria convey an expressive harm of exclusion that carries particular potency in light of a history of poll taxes, literacy tests, and other devices used to suppress political participation by African Americans. While the merit-based vision’s adverse impact on African Americans should prompt concern in and of itself, Part III explores how the shortcomings of the merit-based vision adversely impact other Americans.15 I. COUNTING VOTES AND ASSUMPTIONS ABOUT DEMOCRACY Other commentators have recognized that one’s choices regarding the law of the political process are shaped by one’s assumptions about democracy, which reflect the individual’s understanding about particular cultural, professional, and social realities of politics.16 Two 14. Cf. Gaston County v. United States, 395 U.S. 285, 291 (1969) (invalidating county’s literacy test in part because the county’s previous maintenance of a de jure segregated school system had “deprived its black residents of equal educational opportunities, which in turn deprived them of an equal chance to pass the literacy test”). 15. This Essay concentrates primarily on African Americans due to the high rate of machine-rejected ballots among African Americans. As developed in Part III, however, the experience of African Americans can be used as a lens to reveal structural problems with the merit-based assumptions of the Court that impact many others, including but not limited to Latinos, the elderly, and the poor. 16. See Frank I. Michelman, Conceptions of Democracy in American Constitutional Argument: Voting Rights, 41 FLA. L. REV. 443, 444 (1989) (observing that with regard to issues “soaked with political interest[,] . . . legal argument and judicial explanation . . . unselfconsciously reflect underlying assumptions about actual and potential social relations, and about the institutional arrangements and forms of political life fit for those relations as they are and are capable of becoming”); see also James A. Gardner, 474 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:469 different understandings of democracy animate the judicial opinions in Bush v. Gore—an inclusionary vision and a merit-based vision. The inclusionary vision of democracy values widespread participation and looks to remove criteria or conditions that act as barriers to such participation.17 Under this vision, political participation is a right, and courts and democratic decisionmakers have a responsibility to create an environment that allows for, and even encourages, participation by all citizens.18 Professor Briffault, for example, lists the legislative apportionment cases and the invalidation of the poll tax to illustrate the inclusionary nature of the Court’s jurisprudence prior to Bush v. Gore.19 The Voting Rights Act of 1965, which bans literacy tests and fluency in English as prerequisites for voting,20 and Liberty, Community and the Constitutional Structure of Political Influence: A Reconstruction of the Right to Vote, 145 U. PA. L. REV. 893, 897 (1997) (“We can hardly expect to figure out what voting—or ‘fair’ voting, or ‘meaningful’ voting—means without some conception of what voting is for, what purpose it serves within a larger regime of democratic self-government.”); Spencer Overton, Rules, Standards, and Recounts: Form and the Law of Democracy, 37 HARV. C.R.-C.L. L. REV. (forthcoming 2001) (observing that one’s assumptions about democracy shape one’s preference for using rules or standards to allocate discretion); Richard H. Pildes, Democracy and Disorder, 68 U. CHI. L. REV. 695, 696 (2001) (describing judicial culture as “the empirical assumptions, historical interpretations, and normative ideals of democracy that seem to inform and influence the current constitutional law of democracy”). 17. See Pamela S. Karlan, Undoing the Right Thing: Single-Member Offices and the Voting Rights Act, 77 VA. L. REV. 1, 45 (1991) (describing the “inclusionary understanding of democracy” in amended section 2 to the Voting Rights Act); cf. Gardner, supra note 16, at 904 (describing democratic theory in which “exclusion from the electoral process is exclusion in the deepest possible sense from the essence of American society”). 18. See Kendall Thomas, Racial Justice: Moral or Political?, in LAW’S CENTURY (Austin Sarat ed., forthcoming 2001) (observing that, within a political conception of racial justice, American democracy is charged with three tasks: to maintain “equal and meaningful access for vulnerable racial publics” to institutions in which political identity is formed, to “modify the participatory practices through which” the political opinions of vulnerable racial publics “can be framed and communicated,” and to “insure that the interests of vulnerable racial publics are represented in institutional arenas in which binding collective choices are discussed and made”). 19. Briffault, supra note 8, at 347 nn.103-04 (citing Hill v. Stone, 421 U.S. 289 (1975) (concluding that a Texas provision impermissibly disenfranchised otherwise qualified voters solely because they had not rendered their property for taxation); City of Phoenix v. Kolodziejski, 399 U.S. 204 (1970) (invalidating a provision excluding nonproperty owners from voting in an election to approve general-obligation bonds); Cipriano v. City of Houma, 395 U.S. 701 (1969) (concluding that a Louisiana provision limiting the right to vote on the issuance of revenue bonds to taxpayers violated the Equal Protection Clause); Harper v. Va. Bd. of Elections, 383 U.S. 663 (1966) (invalidating poll taxes). But cf. Ball v. James, 451 U.S. 355 (1981) (concluding that a state could rationally limit voting in a water district election to landowners and that each vote could be weighted respective to the amount of land each voter owned); Buckley v. Valeo, 424 U.S. 1, 49 n.55 (1976) (rejecting the argument that Harper allows Congress to restrict political expenditures); Salyer Land Co. v. Tulare Lake Basin Water Storage Dist., 410 U.S. 719, 726-30 (1973) (concluding that a property-based scheme for electing the governing board of water reclamation district does not violate equal protection). 20. Voting Rights Act of 1965, 42 U.S.C. § 1973aa (1994) (suspending use of literacy tests nationwide); id. § 1973b(f)(1)-(2) (2001) (prohibiting English-only elections and other voting qualifications or prerequisites intended to deny language minorities the right to 2001] A PLACE AT THE TABLE 475 the Twenty-fourth Amendment, which prohibits exclusion from federal elections “by reason of failure to pay any poll tax or other tax,”21 also embrace inclusionary notions of democracy. Proposals for public financing of politics fit into this category as well.22 The majority of the Florida Supreme Court embraced inclusionary assumptions about democracy in its conclusion that a “legal vote” constituted any ballot upon which the clear intent of the voter could be ascertained.23 While the Florida court’s articulation of the “clear intent standard” was based on state statutory language,24 the state vote). But cf. Lassiter v. Northampton County Bd. of Elections, 360 U.S. 45, 51 (1959) (upholding a literacy test that did not have a discriminatory effect because “[t]he ability to read and write . . . has some relation to standards designed to promote intelligent use of the ballot”); Hasen, supra note 9, at 397 (“[Lassiter] is of questionable value following cases like Harper. . . . But if Lassiter remains good law, it stands for the proposition that the state can condition the franchise on voters’ ability to follow instructions—thereby insuring that only educated voters vote.”). 21. U.S. CONST. amend. XXIV, § 1. 22. See, e.g., Richard L. Hasen, Clipping Coupons for Democracy: An Egalitarian/Public Choice Defense of Campaign Finance Vouchers, 84 CAL. L. REV. 1, 20-27 (1996). This Article argues for a new system of campaign finance. . . . Under this plan, each voter would have the opportunity to contribute vouchers to candidates or to interest groups in every federal election cycle. The interest groups would use the vouchers to contribute to candidates or to organize independent expenditure campaigns. With limited exceptions, only funds from the voucher system could be spent to support or oppose candidates for elected federal offices. Id. at 6; Bruce Ackerman, Crediting Voters: A New Beginning for Campaign Finance, 13 AM. PROSPECT 71, 78-79 (1993); Edward B. Foley, Equal-Dollars-Per-Voter, 94 COLUM. L. REV. 1204, 1204 (1994) (“The Constitution of the United States should contain a principle, which I shall call ‘equal-dollars-per-voter,’ that would guarantee to each eligible voter equal financial resources for purposes of supporting or opposing any candidate or initiative on the ballot in any election held within the United States.”). 23. See Briffault, supra note 8, at 372 (arguing that although it was not constitutionally mandatory, the Florida court’s order was consistent with the inclusionary thrust of the United States Supreme Court’s prior application of the Equal Protection Clause in the voting rights area); Pamela S. Karlan, Unduly Partial: The Supreme Court and the Fourteenth Amendment in Bush v. Gore, 29 FLA. ST. U. L. REV. 587, 598-99 (2001) (asserting author’s belief that “as a matter of Florida law, the ‘clear intent of the voter’ standard meant that many of the ballots that were out of strict compliance with Florida law were nonetheless legal votes,” but noting that if such ballots did “not contain legally cast votes, then a recount process that includes them might infringe upon the voting rights of those citizens who did comply with the state’s requirements”). 24. In ordering a manual recount of ballots on which automatic machine recounts had failed to detect a vote for President and which had not yet been manually recounted, the Florida Supreme Court embraced phrasing established by the Florida Legislature. The statute provided that “[n]o vote shall be declared invalid or void if there is a clear indication of the intent of the voter as determined by the canvassing board.” FLA. STAT. § 101.5614(5) (2000), amended by 2001 Fla. Laws ch. 40, § 37, at 144, 145. According to the Florida Supreme Court’s interpretation of the provision, legitimate votes included not only ballots completely punched through, but also all ballots that expressed the clear intent of the voter. The Florida Supreme Court stated that the “clear message from” the legislature was “that every citizen’s vote be counted whenever possible,” thereby imputing an inclusionary understanding of democracy to the Florida Legislature. Gore v. Harris, 772 So. 2d 1243, 1254 (Fla. 2000), rev’d sub nom. Bush v. Gore, 531 U.S. 98 (2000). Thus, the court required canvassing boards and officials to count a vote if there was a “clear indication of the intent of the voter” on the ballot, unless it was “impossible to determine the elector’s 476 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:469 court noted that it had, in the past, “pointed to the ‘will of the voters’ as the primary guiding principle” in resolving election disputes.25 In prior cases it “repeatedly held . . . that so long as the voter’s intent may be discerned from the ballot, the vote constitutes a ‘legal vote’ that should be counted.”26 Rather than emphasizing the shortcomings of voters, the Florida court focused on the responsibility of the state and mentioned that the margins of error for punch card machines might be so significant as to require a reevaluation of the use of the machines.27 In short, a majority of the Florida Supreme Court interpreted the Florida statutory scheme as containing a broad, inclusive definition of a vote that put responsibility on state officials to manually review ballots that lacked machine-readable markings. United States Supreme Court Justices Breyer, Ginsburg, Souter, and Stevens either agreed with or were prepared to defer to this interpretation and would have allowed a manual counting of the ballots.28 In contrast to the inclusionary vision of democracy, the meritbased vision conditions the right to political participation on a citizen’s ability to comply with a particular set of criteria.29 A meritchoice.” Id. (citing FLA. STAT. § 101.5614(5)-(6) (2000)). The Florida Supreme Court may have focused on the legislative standard cognizant of earlier questions expressed by the U.S. Supreme Court Justices as to whether the Florida Supreme Court based an earlier holding on state constitutional provisions rather than state legislative provisions or, through interpretation, “changed” the law in violation of Article II, Section 1, Clause 2 of the U.S. Constitution. Bush v. Gore, 531 U.S. 98, 145 (2000) (Breyer, J., dissenting) (suggesting that “[i]n light of our previous remand, the Florida Supreme Court may have been reluctant to adopt a more specific standard than that provided for by the legislature for fear of exceeding its authority under Article II”). Note that inclusionary objectives were not absent in the consideration of the statutory structure containing section 101.5614(5), Florida Statutes. See, e.g., Letter from Reubin O’D. Askew, Governor, State of Florida, to Members of the Florida Senate and House of Representatives (Mar. 15, 1977) (on file with the Florida State Archives) (observing shrinking voter participation nationally, and suggesting that electoral reform was needed that will “heighten public interest and participation” and “enhance[ ] the ability of citizens to exercise their right to vote”). 25. Gore v. Harris, 772 So. 2d at 1254. 26. Id. at 1256; see also id. at 1261 n.20 (observing that “[t]his presidential election has demonstrated the vulnerability of what we believe to be a bedrock principle of democracy: that every vote counts”). 27. Id. at 1261; see also id. at 1254 (asserting that the right to vote is not just the right to participate and to speak, “but more importantly the right to be heard ” ) (emphasis in original). 28. Bush v. Gore, 531 U.S. at 127 (Stevens, J., dissenting, joined by Breyer and Ginsburg, JJ.) (“Thus, nothing prevents the majority, even if it properly found an equal protection violation, from ordering relief appropriate to remedy that violation without depriving Florida voters of their right to have their votes counted.”). Justice Souter stated that he would defer to the Florida Supreme Court’s interpretation of a legal vote, establish uniform counting standards, and allow the State to count the machine-rejected ballots. Id. at 13135. (Souter, J., dissenting, joined by Breyer, J.). 29. Cf. Jacob Katz Cogan, The Look Within: Property, Capacity, and Suffrage in Nineteenth-Century America, 107 YALE L.J. 473 (1997) (monitoring the nineteenth century shift in locating a person’s capacity for political participation externally in material things like property to internal characteristics such as literacy, and the continued disenfranchisement of women and African Americans); Michelman, supra note 16, at 480-85 (discussing en- 2001] A PLACE AT THE TABLE 477 based vision is individualist to the extent that an individual citizen rather than government has a responsibility to secure or meet the conditions necessary for his or her political participation.30 The merit-based vision of democracy also enhances societal well-being, the argument goes, because better political decisions arise from an electorate made up of citizens who are either competent enough or care enough to meet the criteria.31 Examples of devices that have been considered legitimate under merit-based assumptions about franchisement on the basis of competence); James Thomas Tucker, Affirmative Action and [Mis]representation: Part II—Deconstructing the Obstructionist Vision of the Right to Vote, 43 HOW. L.J. 405, 452-55 (2000) (discussing the merit-based approach as applied to voting). While the merit-based and inclusionary visions differ, they are not mutually exclusive, and both might be embraced in varying degrees. For example, while the Florida Supreme Court employed a more inclusionary vision than the U.S. Supreme Court, the state court’s order would have been even more inclusionary had it, in addition to ordering a manual review of undervotes, ordered a manual review of overvotes. Also, many merit-based practices may not have been repudiated due to a rejection of merit, but because the practices arbitrarily excluded many who were believed just as competent or interested in making political judgments. In other words, the argument goes, the devices were not sufficiently precise in their task of allocating membership in political community based on merit. See Daniel R. Ortiz, The Democratic Paradox of Campaign Finance Reform, 50 STAN. L. REV. 893, 896 (1998) (“We have rejected [poll taxes, property qualifications, and literacy tests] not because we have come to believe their aim of ensuring the independent exercise of political judgment is not worth pursuing [but] . . . because we have come to think that some people had misappropriated these practices to unjustly exclude groups that were just as capable as the rest of us of exercising this kind of judgment. Their central democratic aim remains untarnished.”). 30. Cf. R. Richard Banks, Meritocratic Values and Racial Outcomes: Defending ClassBased College Admissions, 79 N.C. L. REV. 1029, 1036 (2001) (“Meritocracy is individualist insofar as it seeks to distribute opportunities and resources on the basis of the conduct or attributes of individuals. It is productivity-oriented to the extent that it distributes opportunities and resources based on predictions of future performance that will enhance societal well-being.”). 31. Cf. 1 WILLIAM BLACKSTONE, COMMENTARIES 171 (“The true reason of requiring any qualification, with regard to property, in voters, is to exclude such persons as are in so mean a situation that they are esteemed to have no will of their own. If these persons had votes, they would be tempted to dispose of them under some undue influence or other.”); Harper v. Va. Bd. of Elections, 383 U.S. 663, 684-85 (1966) (Harlan, J., dissenting): [I]t is certainly a rational argument that payment of some minimal poll tax promotes civic responsibility, weeding out those who do not care enough about public affairs to pay $1.50. . . . [I]t was probably accepted as sound political theory . . . that people with some property have a deeper stake in community affairs, and are consequently more responsible, more educated, more knowledgeable, more worthy of confidence, than those without means, and that the community and Nation would be better managed if the franchise were restricted to such citizens. Id.; Lassiter v. Northampton County Bd. of Elections, 360 U.S. 45, 52 (1959) (“Yet in our society where newspapers, periodicals, books, and other printed matter canvass and debate campaign issues, a State might conclude that only those who are literate should exercise the franchise. It was said last century in Massachusetts that a literacy test was designed to insure an independent and intelligent exercise of the right of suffrage.”). 478 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:469 democracy include poll taxes, literacy tests, and a privately-funded campaign finance system.32 As acknowledged by Professors Briffault and Hasen, the five U.S. Supreme Court Justices who voted to discontinue manual counting of the ballots did not adopt the Florida Supreme Court’s inclusionary language.33 Instead, Chief Justice Rehnquist and Justices Kennedy, O’Connor, Scalia, and Thomas adopted a more merit-based interpretation that put responsibility on the voter to produce a ballot that could be read by a properly functioning counting machine. The per curiam opinion repeatedly emphasized the failure of those who cast machine-rejected ballots in its reference to punch cards that “have not been perforated with sufficient precision” and were “not punched in a clean, complete way by the voter” due to either “error or deliberate omission.”34 Upon its conclusion that the clear intent standard lacked uniformity in its application and violated equal protection, the Court did not order a manual count of ballots based on a uniform standard. Instead, the Court prohibited any further recognition of the imperfectly marked ballots by asserting that the Florida Legislature preferred to submit the state’s presidential electors by December 12.35 A separate concurrence by Chief Justice Rehnquist and Justices 32. Some have asserted that racial and gender exclusions were merit-based devices that excluded those who were not deemed sufficiently independent and competent to make political decisions. See Ortiz, supra note 29, at 908-09 (“Many believed that freed blacks were uniquely vulnerable to their former masters, employers, or opportunistic whites, and women were thought to be easily swayed by their husbands.”). 33. Briffault, supra note 8, at 372 (describing Bush v. Gore as quite different from the Court’s earlier inclusionary equal protection cases). 34. Bush v. Gore, 531 U.S. 98, 104-05 (2000); cf. Oral Argument Tr. at 58, Bush v. Gore, 531 U.S. 98 (2000) (No. 00-949) (indicating a question of Justice O’Connor as “Well, why isn’t the standard the one that voters are instructed to follow, for goodness sakes? I mean, it couldn’t be clearer. I mean, why don’t we go to that standard?”). 35. The Court stated that the Florida Legislature preferred to conclude the vote tabulation process by December 12 in order to secure a federal statutory guarantee that Congress would not challenge its election results. Bush v. Gore, 531 U.S. at 109-11. But see Michael W. McConnell, Two-and-a-Half Cheers for Bush v. Gore, 68 U. CHI. L. REV. 657, 675 (2001) (observing that the two Florida Supreme Court opinions cited by the Court do not “supply any authoritative pronouncement that December 12 is the absolute deadline for state law purposes”). The majority reasoned that Florida could not possibly tabulate the votes in accordance with minimal constitutional requirements by the deadline, and thus reversed the Florida Supreme Court’s manual recount order. Bush v. Gore, 531 U.S. at 109-11. If imperfectly marked ballots contained votes as legitimate as perfectly marked ballots, then presumably equal protection guarantees would have required a manual review, and such constitutional concerns would have trumped any federal statutory deadline. Cf. Briffault, supra note 8, at 359 (“If equal protection guarantees applied to imperfectly marked ballots, then presumably even ‘the press of time’ would not have justified the failure to review them.”); id. (“[T]here was nothing in the per curiam opinion that indicated that a voter who casts any particular sort of imperfectly marked ballot had any substantive entitlement to have that ballot treated as a valid vote.”). But see Hasen, supra note 9, at 389: Nonetheless, the Court held that the Florida Legislature’s interest . . . in taking advantage of the “safe harbor” provisions of federal law for counting the state’s 2001] A PLACE AT THE TABLE 479 Scalia and Thomas explicitly concluded that the Florida statutory scheme required counting only the ballots on which chads had been punched completely by voters.36 According to the merit-based assumptions of political participation underlying the majority’s opinion, voter intent was less important than voter compliance. If a voter did not “properly” mark his or her ballot in a manner that machines could read, more pressing concerns outweighed a manual review of the ballot. The Court extended a conceptual understanding of formal equality to invalidate the use of the clear intent standard. At the same time, the Court used the seemingly natural and logical cultural values of merit to limit the protection afforded by its particular brand of equality to machinereadable ballots.37 Capacity to punch a ballot so as to completely remove a chad constituted a relevant criterion that the Court used to define the political community. II. MERITOCRACY THROUGH THE LENS OF RACE By employing a seemingly neutral, merit-based qualifier to identify those ballots that deserve recognition, the Court in fact avoids deeper and more difficult normative questions about structural inequalities in our political process. This Part employs race to illuminate some of the more troubling implications of the merit-based vision. A. Race Exposes the Shortcomings of the Merit-Based Vision’s Individualized Focus The lens of race reveals that the merit-based vision’s individualized focus overlooks both the collective nature of political participation and the structural nature of racial disadvantage. electoral votes trumped the rights of all Florida voters to have valid votes counted. It is not self-evident that such a state interest was compelling and trumped the right, recognized in Reynolds but ignored by the Court in Bush v. Gore, to have every vote count . . . . 36. Bush v. Gore, 531 U.S. at 120-22 (Rehnquist, C.J., concurring) (concluding that “there is no basis for reading the Florida statutes as requiring the counting of improperly marked ballots” based on an opinion by Secretary of State Katherine Harris and an argument that the statutory provision was inapplicable because it allegedly applied only to damaged or defective ballots, not ballots imperfectly marked by voters). The concurrence reasoned that the clear intent of the voter was irrelevant and that the Florida Supreme Court’s flawed interpretation changed Florida election laws in violation of Article II of the U.S. Constitution. See id. at 114-15 (Rehnquist, C.J., concurring). 37. The Court did not extend its conceptual understanding of equality to require that all voters have access to similar types of voting machinery. See id. at 109 (“The question before the Court is not whether local entities, in the exercise of their expertise, may develop different systems for implementing elections.”). 480 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:469 In the absence of a racial analysis, the Court’s merit-based criteria may appear neutral, reasonable, and benign. Because the Court gave all eligible Florida citizens the formal opportunity to vote on roughly the same terms, each Floridian had an equal opportunity to participate. The fact that some individuals were better able to follow instructions and perform the simple tasks necessary to produce a machine-countable ballot reflected differences in voter motivation, voter education, or voter experience.38 Individuals with greater education and wealth are more likely to participate in politics generally,39 one might argue, and it is not surprising that these individuals were more likely to cast machine-readable votes. From this perspective, no outcast or disenfranchised groups existed that required protection. Instead, there were only responsible individuals whose political entitlement was threatened by the claims of those who failed to exhibit an appropriate amount of personal responsibility. While the simple task of punching a ballot may not appear to be a significant barrier for any individual voter, the merit-based vision fails to recognize that politics involves not simply individual rights but also associational and structural concerns.40 Although individuals 38. Cf. Dana Canedy, Florida Governor Calls Commission Report on Election Biased, N.Y. TIMES, June 6, 2001, at A22. Gov. Jeb Bush’s office sent a scathing letter to the United States Commission on Civil Rights today denouncing its preliminary findings on the problemplagued presidential race in Florida last November. The letter dismissed the investigation’s findings as irresponsible and biased. . . . Mr. Bush’s letter took aim at many of the findings, including the issue of widespread disenfranchisement among minority voters. The letter said that, as with all voters, minorities could have been affected by a number of variables that the commission failed to take into account, such as “the voter’s education level, the voter’s experience with voting, the ballot design and the voting machine used.” Id. 39. See U.S. CENSUS BUREAU, VOTING AND REGISTRATION IN THE ELECTION OF NOVEMBER 1998, at 7 (2000), available at http://www.census.gov/prod/2000pubs/ p20-523.pdf (“In 1998, citizens who had bachelor’s degrees were nearly twice as likely (58 percent) to report that they voted as those who had not completed high school (30 percent). At each level of educational attainment from high school completion and above, voting rates increase significantly.”); John Green, Paul Herrnson, Lynda Powell & Clyde Wilcox, Individual Congressional Campaign Contributors: Wealthy, Conservative and ReformMinded, Individual Donors and Campaign Finance 13 (1998), at http://www. opensecrets.org/pubs/donors/donors.htm (providing that, of contributors to the 1996 congressional elections who responded to an academic survey funded by the Joyce Foundation, 81% had annual incomes over $100,000 and 20% had annual incomes higher than $500,000). 40. Compare Reynolds v. Sims, 377 U.S. 533, 561 (1964) (describing voting rights as “individual and personal in nature”), with Lani Guinier, Groups, Representation, and Race-Conscious Districting: A Case of the Emperor’s Clothes, 71 TEX. L. REV. 1589, 1595 (1993) (suggesting “that the one-person, one-vote doctrine is consistent with both group and individual conceptions of voting”), and Samuel Issacharoff & Pamela S. Karlan, Standing and Misunderstanding in Voting Rights Law, 111 HARV. L. REV. 2276, 2282 n.30 (1998) (asserting that one-person, one-vote cases like Reynolds “should be viewed as cases about group political power . . . rather than purely about individual rights”). 2001] A PLACE AT THE TABLE 481 cast votes, individual voters enact political change by associating with political groups.41 Under a constitutive understanding of voting, the experience of participation in politics is valued “as a process of formation or field of exertion of self or community” through which “persons or communities (or both, reciprocally) forge identities.”42 Voting is seen as a “vehicle for self-development and identification, and a means for creating alliances and thus a community among individuals so engaged.”43 The Court’s seeming unwillingness to recognize these values in Bush v. Gore is especially evident when one considers race. In its focus on individual responsibility,44 either the Court fails to consider or is indifferent to the manner in which its merit-based criteria interfere with the ability of voters like African Americans to identify with one another as a political community, create alliances with others of 41. Cf. Davis v. Bandemer, 478 U.S. 109, 167 (1986) (Powell, J., concurring in part and dissenting in part) (“The concept of ‘representation’ necessarily applies to groups: groups of voters elect representatives, individual voters do not.”); Alexander M. Bickel, The Supreme Court and Reapportionment, in REAPPORTIONMENT IN THE 1970S, at 57, 59 (Nelson W. Polsby ed., 1971) (“We have, since Madison, realized that people tend to act politically not so much as individuals as in groups.”); Anthony A. Peacock, Voting Rights, Representation, and the Problem of Equality, in AFFIRMATIVE ACTION AND REPRESENTATION: SHAW V. RENO AND THE FUTURE OF VOTING RIGHTS 17 (Anthony A. Peacock ed., 1997) (“Although representational politics is necessarily group oriented—groups of voters electing representatives, not individuals—the individual right to vote must be respected in any system of representation.”); Heather K. Gerken, Understanding the Right to an Undiluted Vote, 114 HARV. L. REV. 1663, 1742 (2001) (“[I]f we are going to recognize an aggregate harm like dilution, we must take into account its group-like qualities.”); Samuel Issacharoff, Groups and the Right to Vote, 44 EMORY L.J. 869, 884 (1995) (“[T]he right to effective voting is incomprehensible without that conception of the group.”). 42. Michelman, supra note 16, at 451. 43. Ellen D. Katz, Race and the Right to Vote After Rice v. Cayetano, 99 MICH. L. REV. 491, 513 (2000). 44. The merit-based vision’s focus on individual responsibility is not inconsistent with the concentration by some Justices on the individual characteristics of voting. See Shaw v. Hunt, 517 U.S. 899, 917 (1996) (“To accept that [a remedial] district may be placed anywhere implies that the claim, and hence the coordinate right to an undiluted vote (to cast a ballot equal among voters), belongs to the minority as a group and not to its individual members. It does not.”); Miller v. Johnson, 515 U.S. 900, 911 (1995) (explaining that a state may not create voting districts on the basis of race because “[g]overnment must treat citizens as individuals”); Shaw v. Reno, 509 U.S. 630, 647-49 (1993) (describing harms that arise from districting based upon groups and not individuals); see also Amar & Brownstein, supra note 6, at 917 (criticizing “the Court’s exclusively individualistic perspective” in the voting and jury contexts); Gerken, supra note 41, at 1665-66 (describing the “highly individualistic view of rights developed by the Rehnquist Court”); Guinier, supra note 40, at 1601 (referring to “the efforts of some members of the Court to characterize representation as an exclusively individual notion”); Pamela S. Karlan & Daryl J. Levinson, Why Voting Is Different, 84 CAL. L. REV. 1201 (1996) (criticizing the Court’s application of its general equal protection doctrine, which focuses on individual rights, to the voting context); Tucker, supra note 29, at 410 (claiming that the majority in Shaw v. Reno “assumed that the right to vote was an individual, and not a group right”). 482 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:469 different backgrounds, and use the vote instrumentally to enact political change.45 By limiting the relevant political community to those who exhibited the capacity to create a machine-readable ballot, the Court’s decision diluted the political choices of both African Americans whose ballots were rejected by machines and African Americans who properly punched their ballots but identified politically with those whose ballots were rejected.46 African Americans who exhibited the capacity to punch ballots were no longer allowed to aggregate their preferences with those excluded by the Court’s ruling.47 While the Court justifies its decision on a lapse in individual responsibility, its decision penalizes all African Americans who identify as part of a political group equally, with no distinction between those who did or did not completely punch their ballots. The Court did not individually reprimand those who failed to punch the ballot properly and who most probably do not even know of their transgression. Whereas a more inclusionary vision would have allowed African Americans to more freely forge a common identity and exert some degree of collective self-determination in improving their lives through shaping the political environment,48 the merit-based vision disabled a critical device used to engage in these activities. The Court’s failure to recognize this dilution of political strength as illegitimate arises, in part, from a related problem of the meritbased vision’s focus on individuals. The merit-based vision fails to adequately appreciate that racial disadvantage arises not simply from isolated, intentional actions of malicious individuals but also 45. Cf. Roberts v. Wamser, 679 F. Supp. 1513, 1532 (1987), rev’d on standing grounds, 883 F.2d 617 (8th Cir. 1989) (concluding that election board’s failure to manually review punch card ballots rejected by tabulating equipment constituted a violation of the Voting Rights Act because such a failure results “in the City’s black voters having less opportunity than other members of the City’s electorate to participate in the political process and to elect representatives of their choice”). 46. Terry Smith, A Black Party? Timmons, Black Backlash and the Endangered TwoParty Paradigm, 48 DUKE L.J. 1, 51 (1998) (“Blacks have historically functioned as a party within a party. Even during the era of limited black enfranchisement, blacks formed ‘satellite’ or ‘parallel’ parties to advance their interests within the two-party structure.”). 47. Gerken, supra note 41, at 1669-70 (distinguishing vote dilution claims from claims based on conventional individual rights by observing that with regard to voting: “fairness is measured in group terms; an individual’s right rises and falls with the treatment of the group; and the right is unindividuated among members of the group”); Tucker, supra note 29, at 414 (“When an electoral scheme systematically prevents the collective exercise of voting rights for particular groups, the individual right to vote is diminished accordingly.”). 48. Even after reviewing all of the imperfect votes, no undisputed winner of the 2000 presidential election emerged. Dennis Cauchon & Jim Drinkard, Florida Voter Errors Cost Gore the Election: Bush Still Prevails in Recount of All Disputed Ballots, Using Two Most Common Standards, USA TODAY, May 11, 2001, at 1A (finding that had all disputed ballots been counted by hand, George W. Bush would have won under two of the most widely used standards for counting votes, Al Gore would have won under the two least used, and that overall, most voters intended to vote for Gore). 2001] A PLACE AT THE TABLE 483 from structural factors that fail to account for context and history.49 The merit-based vision seems to assume that constitutionally sufficient equity exists in the status quo—that the baseline has some degree of fairness. Under these assumptions, differences in political outcomes that result from ballots not counted by machines do not reflect past discrimination but rather differences between autonomous individuals. Whereas the inclusionary vision actively seeks and removes context-specific obstacles to political participation, the meritbased vision fails to question whether its facially neutral criteria have a disparate impact upon certain populations. The merit-based vision either ignores or tolerates that factors such as lower education, a greater percentage of first-time voters, a greater reluctance to ask for assistance, segregated residential patterns, and substandard voting equipment and assistance at the polls in predominantly African-American neighborhoods are not fully pre-political or meritbased50 but stem in part from illegitimate factors such as past statesponsored racial discrimination.51 49. See Charles R. Lawrence III, Foreword: Race, Multiculturalism, and the Jurisprudence of Transformation, 47 STAN. L. REV. 819, 824-25 (1995) (“The substantive approach sees the disestablishment of ideologies and systems of racial subordination as indispensable and prerequisite to individual human dignity and equality. The nonsubstantive approach sees the individual right to be treated without reference to one’s race as primary.”); cf. Introduction to CRITICAL RACE THEORY: THE KEY WRITINGS THAT FORMED THE MOVEMENT xiii, xxiv (Kimberlé Crenshaw et al. eds., 1995) (“The debate that ensued in light of this different orientation engendered an important [Critical Race Theory] theme: the absolute centrality of history and context in any attempt to theorize the relationship between race and legal discourse.”). 50. Cf. Kimberlé Williams Crenshaw, Race, Reform, and Retrenchment: Transformation and Legitimation in Antidiscrimination Law, 101 HARV. L. REV. 1331, 1379 (1988) (“The rationalizations once used to legitimate Black subordination based on a belief in racial inferiority have now been reemployed to legitimate the domination of Blacks through reference to an assumed cultural inferiority.”); id. at 1370 (“Throughout American history, the subordination of Blacks was rationalized by a series of stereotypes and beliefs that made their conditions appear logical and natural.”). 51. Cf. S. REP. NO. 97-417, at 29 (1982), reprinted in 1982 U.S.C.C.A.N. 177, 206 (describing one factor used to determine a violation of section 2 of the Voting Rights Act as “the extent to which members of the minority group in the state or political subdivision bear the effects of discrimination in such areas as education, employment and health, which hinder their ability to participate effectively in the political process . . . .”); Lawrence, supra note 7, at 953 (“[The liberal theory] makes no effort to inquire into the ways that current facially neutral practices may have a foreseeable and unjustifiable discriminatory impact or to account for unconscious bias in their administration.”); Spencer Overton, Voices from the Past: Race, Privilege, and Campaign Finance, 79 N.C. L. REV. 1541 (2001) (observing that the existing distribution of property upon which the privatelyfunded campaign finance system is based is not pre-political, but is shaped in part by past state discrimination). Judges and other decisionmakers’ assumptions about the democratic process are likely shaped, in part, by preferences and judgments about race. If one subscribes to a colorblind ideology and believes that race is irrelevant outside of intentional actions by malicious individuals, merit-based assumptions about democracy that happen to exclude people of color may seem more logical. If one has a more expansive view that appreciates the structural nature of race, one might adopt more inclusionary assumptions about the nature of democracy. 484 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:469 In short, the merit-based vision isolates broader structural concerns, both about historical disadvantage and contemporary racial identities, because its individualized focus is not sufficiently expansive to take the broader problems into account.52 The narrowness of the merit-based vision, however, should not suggest that it is a neutral or impartial tool completely removed from political outcomes.53 B. Race Exposes Particular Expressive Components of Merit-Based Vision In addition to exposing the merit-based vision’s failure to appreciate the unique characteristics of voting that extend beyond the individual, a consideration of race allows one to more broadly comprehend the expressive effect of the Court’s decision in Bush v. Gore. Voting’s expressive component “rests on the message the electoral system is understood to disseminate, and accordingly represents something bestowed on the political community.”54 To members of the majority of the Court, the Florida Supreme Court’s manual recount order sent a message of haphazardness and arbitrariness that was “not well calculated to sustain the confidence that all citizens must 52. Cf. Tucker, supra note 29, at 453 (“The individual portrait . . . is very ill-suited for the political landscape in which voting occurs. . . . [P]oliticians are keenly aware of the racial, social, political, and economic characteristics of voters, and redraw district boundaries to comport with those groups they believe will best enhance the political strength of themselves or their party.”). 53. See Martha Minow, Foreword: Justice Engendered, 101 HARV. L. REV. 10, 32 (1987) (“The unstated point of comparison is not neutral, but particular, and not inevitable, but only seemingly so when left unstated.”); see also e.g., Banks, supra note 30, at 1034 (“Merit is a functional concept—no quality or characteristic is inherently meritorious. Merit is necessarily defined with respect to particular contexts, goals, and values.”); Jamin B. Raskin, Affirmative Action and Racial Reaction, 38 HOW. L.J. 521, 551 (1995) (“Merit is neither self-defining nor self-revealing; it is an ever-changing concept that is historically, socially, and institutionally contingent—and often contested. It is impossible to define merit without asking what kinds of institutions we want to have and for what purposes.”); Daria Roithmayr, Deconstructing the Distinction Between Bias and Merit, 85 CAL. L. REV. 1449, 1503 (1997) (“Reason and merit are culturally and ideologically specific constructs that depend on a particular ideological discourse and can adjudicate only for those who subscribe to that ideology.”). 54. Katz, supra note 43, at 513 n.119; see Elizabeth S. Anderson & Richard H. Pildes, Expressive Theories of Law: A General Restatement, 148 U. PA. L. REV. 1503, 1520 (2000) (“On the rights and equality side of constitutional law, [expressive] theories assert that state action is required to express the appropriate attitudes toward persons.”). Some might suggest that the consideration of race in analyzing the Court’s decision to exclude the imperfectly marked ballots is dangerous because such an analysis might send the message that individuals are less competent in political participation simply because of race. Cf. Shaw v. Reno, 509 U.S. 630, 657 (1993) (“Racial classifications of any sort pose the risk of lasting harm to our society. They reinforce the belief, held by too many for too much of our history, that individuals should be judged by the color of their skin.”). But see Lawrence, supra note 49, at 838 (“The colorblind race baiter completes his white supremacist wizardry by blaming affirmative action itself for creating hostility, resentment, and racial divisiveness.”). 2001] A PLACE AT THE TABLE 485 have in the outcome of elections.”55 The message sent by such a recount would cast “a cloud” over the “legitimacy” of the election and might threaten democratic stability.56 The Court, however, did not explore the message sent by its merit-based qualifier that resulted in imperfectly marked ballots being disregarded. The expressive harm that the Court’s exclusion of the ballots generated is especially potent when one examines the problem through the lens of race. Recognizing that a “voter is a full member of the political community,”57 the dismissal of the imperfectly marked ballots sends a message of exclusion from the political community.58 Exclusion from the political process conveys a form of second-class citizenship on those who are excluded.59 Historically, election administra55. See Bush v. Gore, 531 U.S. 98, 109 (2000) (“The contest provision, as it was mandated by the State Supreme Court, is not well calculated to sustain the confidence that all citizens must have in the outcome of elections.”). 56. Cf. Bush v. Gore, 531 U.S. 1046, 1046 (2000) (Scalia, J., concurring with the Court’s order to stay the manual count of ballots). The counting of votes that are of questionable legality does in my view threaten irreparable harm to petitioner, and to the country, by casting a cloud upon what he claims to be the legitimacy of his election. Count first, and rule upon legality afterwards, is not a recipe for producing election results that have the public acceptance democratic stability requires. Id. 57. Karlan, supra note 17, at 5 (noting that voting “announces that the voter is a full member of the political community”); see also KENNETH L. KARST, BELONGING TO AMERICA: EQUAL CITIZENSHIP AND THE CONSTITUTION 93 (1989) (“Voting . . . is an assertion of belonging to a political community.”); QUIET REVOLUTION IN THE SOUTH: THE IMPACT OF THE VOTING RIGHTS ACT 1965-1990, at 15-16 (Chandler Davidson & Bernard Grofman eds., 1994) (noting that the right to vote “confer[s] full citizenship on the members of the group”); JUDITH N. SHKLAR, AMERICAN CITIZENSHIP: THE QUEST FOR INCLUSION 26 (1991); Gardner, supra note 16, at 906 (“To seek the vote is to seek formal recognition as a full member of society; to be denied the vote is to be either excluded altogether from membership in the community or consigned to some kind of second-class citizenship.”); Pamela S. Karlan, Maps and Misreadings: The Role of Geographic Compactness in Racial Vote Dilution Litigation, 24 HARV. C.R.-C.L. L. REV. 173, 179-82 (1989) [hereinafter Karlan, Maps and Misreadings] (discussing the symbolic value of civic inclusion). 58. Cf. Katz, supra note 43, at 512-13 (“Denial of the vote is tantamount to exclusion from the community or relegation to second-class citizenship, with the message of exclusion being the primary harm produced.”). 59. See Lani Guinier, Voting Rights and Democratic Theory: Where Do We Go From Here?, in CONTROVERSIES IN MINORITY VOTING: THE VOTING RIGHTS ACT IN PERSPECTIVE 283, 284-85 (Bernard Grofman & Chandler Davidson eds., 1993). The Voting Rights Act: is premised on a broad vision of political equality and empowerment. The vision of empowerment anticipated an electorate actively participating in policy reform, not merely reconfigured districts that ensure legislative voting privileges for a few black elected officials. The vision imagined a transformative politics that would value political participation for its own sake in order to recognize the autonomy and dignity of black voters. Participation would affirm their status as first-class citizens in a democracy. Id.; KARST, supra note 57, at 94 (“Voting is the preeminent symbol of participation in the society as a respected member, and equality in the voting process is a crucial affirmation of the equal worth of citizens.”); SHKLAR, supra note 57, at 2-3 (contrasting slavery to voting, and arguing that “the ballot has always been a certificate of full membership in society,” an indicator of “social standing” that has the “capacity to confer a minimum of social dig- 486 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:469 tors used merit-based qualifications, such as literacy tests, understanding and character clauses, property ownership, and poll taxes, to unfairly secure political power through the exclusion of African Americans and others.60 The dramatic drop in voter registration and participation among African Americans in the thirty years following passage of the Fifteenth Amendment’s prohibition on racial discrimination in voting illustrates the effectiveness of these tools.61 The percentage of voting-age African Americans who participated in Florida gubernatorial elections plummeted from 87% in 1884 to just 5% in 1896.62 Just as facially race-neutral poll taxes and literacy tests disseminated the message that African Americans were to be disproportionately excluded from the political process, the Court’s prohibition on a review of the imperfectly marked ballots, the majority of which came from the African-American community, conveyed a message of racial exclusion.63 nity” on those who have it, and describing the vote as “a demand for inclusion in the polity, an effort to break down excluding barriers to recognition”); Ronald Dworkin, What is Equality? Part 4: Political Equality, 22 U.S.F. L. REV. 1, 4 (1987) (voting is a means by which a “community confirms an individual person’s membership, as a free and equal citizen . . . .”); Pamela S. Karlan, The Rights to Vote: Some Pessimism About Formalism, 71 TEX. L. REV. 1705, 1710 (1993) (“The primary value underlying the participation cases . . . is an aspect of . . . civil inclusion: ‘a sense of connectedness to the community and of equal political dignity; greater readiness to acquiesce in governmental decisions and hence broader consent and legitimacy’”) (quoting Karlan, Maps and Misreadings, supra note 57, at 180); see generally CASS R. SUNSTEIN, THE PARTIAL CONSTITUTION (1993); David M. Estlund, Who’s Afraid of Deliberative Democracy? On the Strategic/Deliberative Dichotomy in Recent Constitutional Jurisprudence, 71 TEX. L. REV. 1437 (1993); Miriam Galston, Taking Aristotle Seriously: Republican-Oriented Legal Theory and the Moral Foundation of Deliberative Democracy, 82 CAL. L. REV. 331 (1994). 60. See Harman v. Forssenius, 380 U.S. 528, 543 (1965) (“The Virginia poll tax was born of a desire to disenfranchise the Negro.”); DERRICK BELL, RACE, RACISM AND AMERICAN LAW § 4.4.1-.2, at 191-93 (3d ed. 1992) (discussing white primaries); SAMUEL ISSACHAROFF ET AL., THE LAW OF DEMOCRACY: LEGAL STRUCTURE OF THE POLITICAL PROCESS 78 (1998); see also J. MORGAN KOUSSER, THE SHAPING OF SOUTHERN POLITICS: SUFFRAGE RESTRICTIONS AND THE ESTABLISHMENT OF THE ONE-PARTY SOUTH, 1880-1910 (1974); Armand Derfner, Racial Discrimination and the Right to Vote, 26 VAND. L. REV. 523, 524 (1973); Emma Coleman Jordan, Taking Voting Rights Seriously: Rediscovering the Fifteenth Amendment, 64 NEB. L. REV. 389, 397 (1985) (observing that “[i]n the South, state and local governments began to use gerrymandering, poll taxes, literacy tests, ‘grandfather clauses,’ white primaries, malapportionment, residency requirements, property ownership requirements, fraud, and violence to bring about the total disenfranchisement of Black voters”). 61. See ISSACHAROFF ET AL., supra note 60, at 68. 62. KOUSSER, supra note 60, at 91-103 (discussing the racially exclusionary impact of Florida poll taxes and eight-box provisions designed to disenfranchise illiterate African Americans, and the decline in African-American voter turnout); see also Katz, supra note 43, at 512-13 n.117 (discussing the expressive harm generated by post-Reconstruction disenfranchisement and citing LEON F. LITWACK, BEEN IN THE STORM SO LONG: THE AFTERMATH OF SLAVERY 531-556 (1979), and GLENDA ELIZABETH GILMORE, GENDER AND JIM CROW 123-24 (1996)). 63. The Court was complicit in the messages of disenfranchisement sent to African Americans in the 1800s and early 1900s. See Giles v. Harris, 189 U.S. 475 (1903) (denying jurisdiction of a federal court over claim brought by African-American resident of Alabama 2001] A PLACE AT THE TABLE 487 The Court’s use of merit-based criteria also has special meaning in light of recent controversies regarding affirmative action. As in the affirmative action context, some might interpret the Court’s reliance on merit with regard to the law of democracy simply as a pretext to exclude African Americans, or at least as a misplaced set of priorities in which seemingly neutral criteria are valued over the presence and participation of Americans from all backgrounds in public institutions.64 Indeed, merit has been used as an argument to invalidate race-conscious measures that secure the political rights of historically disadvantaged voters.65 Some might claim that the merit-based vision sends the message that “voters must follow the rules of voting.” Such a message, however, loses its appeal when one recognizes that Florida officials included many votes that did not comply with a literal and generally undisputed interpretation of the law, including but not limited to improperly submitted absentee votes.66 Recognizing this inconsistency, one message of the Court interpreted through the lens of race is that “members of politically disfavored groups must follow the rules of voting.” Another possible message, the substance of which is consistent with merit-based criteria, conveys that “those with the foresight, to compel local board of registrars to enroll his name upon the voting lists of the county); Williams v. Mississippi, 170 U.S. 213 (1898) (concluding that literacy tests do not facially discriminate on the basis of race and do not violate equal protection); United States v. Reese, 92 U.S. 214, 221-22 (1875) (invalidating legislation providing for punishment of a Kentucky election inspector who refused to receive and count the votes of African Americans). But see Guinn v. United States, 238 U.S. 347 (1915) (concluding that Oklahoma grandfather clause violated the Fifteenth Amendment). 64. See generally Charles R. Lawrence III, The Id, the Ego and Equal Protection: Reckoning With Unconscious Racism, 39 STAN. L. REV. 317 (1987). 65. Cf. Rice v. Cayetano, 528 U.S. 495, 517 (2000) (striking down a state law that allowed only native Hawaiians to vote for trustees of public agency that managed programs designed to benefit native Hawaiians and asserting that “[o]ne of the principal reasons race is treated as a forbidden classification is that it demeans the dignity and worth of a person to be judged by ancestry instead of by his or her own merit and essential qualities”); Katherine Inglis Butler, Affirmative Racial Gerrymandering: Fair Representation for Minorities or a Dangerous Recognition of Group Rights?, 26 RUTGERS L.J. 595, 621 n.72 (1995) (arguing that “[o]nce we recognize that racial groups are entitled to be represented as racial groups, we are well on our way to enthroning group fairness, rather than individual merit, as the basis for all societal decisions”). But cf. Tucker, supra note 29, at 454-55 (describing the “merit-based approach” of critics of affirmative action as inappropriate in the voting rights context). 66. Cf. Jacobs v. Seminole County Canvassing Bd., 773 So. 2d 519 (Fla. 2000) (holding that a county supervisor’s decision to allow representatives of one political party to add information to absentee ballot request forms in violation of Florida law did not invalidate requests); David Barstow & Don Van Natta Jr., How Bush Took Florida: Mining the Overseas Absentee Vote, N.Y. TIMES, July 15, 2001, at 1 (reporting that “[u]nder intense pressure from the Republicans, Florida officials accepted hundreds of overseas absentee ballots that failed to comply with state laws”). While some might assert that the neutrality of a merit-based vision of democracy is necessary when resolving a disputed election (as opposed to prospectively setting forth the processes for running an election), the selective application of merit-based criteria establishes the transparency of such neutrality. 488 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:469 determination, resources, and skill to manipulate rules will be rewarded.” The lens of race also exposes the democratic instability that arises from the expressive harm of exclusion. Whereas an inclusionary vision conveys government’s respect for all of its citizens and lends legitimacy to government decisions, the same cannot be said about the merit-based vision. A political victory based on merit-based assumptions sends an exclusionary message that undermines the confidence necessary to ensure the voluntary consent of citizens. In such an atmosphere, it is even more difficult to build coalitions, accommodate diverse viewpoints in political discourse and government policy, and engage in constructive dialogue across racial lines.67 Reconciliation and healing are almost impossible because African Americans have not been fully included in decisionmaking.68 Instability in the wake of exclusion arises not only from the perceived illegitimacy of political results, but also from the perceived illegitimacy of the institutions that led to those results.69 African Americans overwhelmingly vote against conservative politicians, and five Justices appointed by Republican presidents prohibited the counting of thousands of ballots, the majority of which were from African-American precincts.70 The perception that the Court attempted to secure its political objectives by ignoring the political choices of African Americans is potentially destabilizing.71 It taps into a reservoir 67. Cf. Guinier, supra note 6, at 131 (“Exclusion of a racial or language minority group exposes a deep fissure in the American democratic bargain, which purportedly reconciles majoritarian preferences with minority interests.”). 68. Cf. Guinier, supra note 59, at 285 (“A transformative politics would also ensure government legitimacy because it would give disadvantaged groups a substantive basis for lending their consent to government decisions.”). 69. Cf. Bush v. Gore, 531 U.S. 98, 157-58 (2000) (Breyer, J., dissenting) (“And, above all, in this highly politicized matter, the appearance of a split decision runs the risk of undermining the public’s confidence in the Court itself. That confidence is a public treasure. . . . [W]e do risk a self-inflicted wound—a wound that may harm not just the Court, but the Nation.”). 70. Others might raise suspicions about the Florida Supreme Court’s political motives in including the undervotes. In the context of history which has excluded African Americans from political, educational, and economic spheres, however, the high federal court’s message of exclusion may convey a message to some that extends past mere political posturing. 71. Others, including but not limited to the Justices themselves, have addressed whether political motivations influenced the Court. Compare Bush v. Gore, 531 U.S. at 128-29 (Stevens, J., dissenting) (“Although we may never know with complete certainty the identity of the winner of this year’s Presidential election, the identity of the loser is perfectly clear. It is the Nation’s confidence in the judge as an impartial guardian of the rule of law.”), and William Marshall, The Supreme Court, Bush v. Gore, and Rough Justice, 29 FLA. ST. U. L. REV. 787 (2001) (accepting that the Court’s decision was political, and inquiring as to whether it was appropriate), and Ginsburg Recalls Florida Recount Case, N.Y. TIMES, Feb. 4, 2001, at A25 (reporting that Justice Ginsburg stated that “[w]hatever final judgment awaits Bush v. Gore in the annals of history, I am certain that the good work and good faith of the U.S. federal judiciary as a whole will continue to sustain public 2001] A PLACE AT THE TABLE 489 of suspicion among African Americans that they do not count as citizens within the political community.72 The indifference of others who either rationalize or idly tolerate such exclusion only compounds the anger, resentment, and frustration felt by those who are excluded.73 In effect, the merit-based vision’s message of exclusion can promote racial distrust and detachment, factionalism, and political instability generally.74 III. MERIT AND THE EXCLUSION OF US ALL The problems associated with the merit-based vision are especially visible through the lens of race due to the unique interaction of history, political identity, and race in the United States. While racial confidence at a level never beyond repair”), with Bush v. Gore, 531 U.S. at 111 (“None are more conscious of the vital limits on judicial authority than are the members of this Court, and none stand more in admiration of the Constitution’s design to leave the selection of the President to the people, through their legislatures, and to the political sphere.”), and Bill Rankin, The Ruling: Was it Politics? That, Too, is in Dispute, ATLANTA J. & CONST., Dec. 14, 2000, at 2B (reporting that in response to questions about political motives of the Court following Bush v. Gore, Justice Thomas stated that one should not “apply the rules of the political world” to the Court, that the Justices “have no axes to grind” but simply protect the Constitution, and that he had never heard any discussion “of partisan politics among members of the court”). 72. Cf. Lawrence D. Bobo, Michael C. Dawson, & Devon Johnson, Enduring TwoNess, PUB. PERSP., May/June 2001, at 12 (reporting the results of the National African American Election Survey that show blacks are politically alienated even when compared to white Democrats). 73. Cf. DERRICK BELL, FACES AT THE BOTTOM OF THE WELL: THE PERMANENCE OF RACISM 111 (1992) (“Isn’t this the point of Invisible Man . . . where Ralph Ellison depicts blacks as a category of human beings whose suffering is so thoroughly ignored that they, and it, might as well not exist?”) (citing RALPH ELLISON, INVISIBLE MAN 261-80 (1972)); PATRICIA J. WILLIAMS, THE ALCHEMY OF RACE AND RIGHTS 56 (1991) (observing that invisibility results from “not being part of the larger cultural picture”); Martin Luther King, Jr., Letter from Birmingham Jail, reprinted in 26 U.C. DAVIS L. REV. 835, 843 (1993) (“We will have to repent in this generation not merely for the hateful words and actions of the bad people but for the appalling silence of the good people.”). 74. As recently as 1997, the Court recognized that “[s]tates . . . have a strong interest in the stability of their political systems,” and political stability has most often been invoked to justify ballot access laws that favor the two major parties to the disadvantage of other parties. Timmons v. Twin Cities Area New Party, 520 U.S. 351, 366 (1997) (“States also have a strong interest in the stability of their political systems.”); Rutan v. Republican Party of Ill., 497 U.S. 62, 107 (1990) (Scalia, J., dissenting) (“The stabilizing effects of such a [two-party] system are obvious.”); Davis v. Bandemer, 478 U.S. 109, 144-45 (1986) (O’Connor, J., concurring). There can be little doubt that the emergence of a strong and stable two-party system in this country has contributed enormously to sound and effective government. The preservation and health of our political institutions, state and federal, depends to no small extent on the continued vitality of our two-party system, which permits both stability and measured change. Id.; Storer v. Brown, 415 U.S. 724, 736 (1974) (“Splintered parties and unrestrained factionalism may do significant damage to the fabric of government. It appears obvious to us that the [provision at issue] furthers the State’s interest in the stability of its political system. We also consider that interest as not only permissible, but compelling . . . .”) (citations omitted). 490 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:469 disadvantage in and of itself warrants concern, it is also important to note that the vulnerability of African Americans to the Court’s remedy serves as a diagnostic tool that exposes structural faults in a merit-based vision of democracy that impact others.75 Many Americans of various backgrounds, including but not limited to the elderly, the poor, the language and religious minorities, and the disabled, use voting as a means to maintain communities of identity and to exert collective self-determination in shaping their world through the political process.76 Exclusionary, merit-based assumptions about democracy that ignore context and history interfere with the ability of these Americans to identify with one another and ally themselves with others in the political sphere.77 Further, a diverse group of Americans suffer the expressive harms associated with an exclusionary, merit-based vision of democracy, whether the message is “you are too old and senile to vote,” or “English is our primary language,” or “you are poor and don’t deserve the same equipment as other voters.” By confronting merit-based assumptions about democracy embedded in the majority’s reasoning in Bush v. Gore, one discovers that there is more to fix in American democracy than vote-counting machinery. Indeed, a focus on better machines and more uniformity may push other, more weighty questions about the meaning of democracy to the background. Significant disparities in participation and even the counting of votes are likely to continue under a system of formal equality that fails to account for context-specific barriers to political inclusion.78 Some commentators may make seemingly persuasive ar75. Cf. GUINIER & TORRES, supra note 6 (manuscript at 1): Those who are racially marginalized are like the miner’s canary. It is easy enough to think that when we sacrifice this canary the only harm is to communities of color. Yet if those who are racially marginalized do function as the miner’s canary, others ignore problems that converge around racial minorities at their own peril. We are ignoring the symptoms that tell us we are all being poisoned. Id. 76. Cf. KOUSSER, supra note 60, at 49 (noting that poll taxes and literacy tests disenfranchised poor whites in addition to African Americans). 77. See Dana Canedy, Vote Spices Up Bubbling Ethnic Stew, N.Y. TIMES, Nov. 11, 2000, at A13 (describing similarities between Jews and African Americans in Florida, and reporting that many “elderly Jewish voters in Palm Beach County said confusing ballots had caused them to mistakenly vote for Patrick J. Buchanan . . . who is perceived by some as anti-Semitic” and that many Jews “believe they lost out on a chance to elect the first Jewish vice president”). 78. For example, a state might provide matching funds to every county for the purchase of optical scanner voting systems. This proposal, however, might fail to result in meaningful inclusion and might even increase racial disparities, as poorer counties might not be able to afford the systems. Similarly, every county might have optical scanners, but the majority of spoiled ballots might still come from precincts of color. Rather than responding with indifference to these situations, reasoning that every voter had an equal opportunity, decisionmakers should seek strategies that maximize inclusion. Cf. Jordan, 2001] A PLACE AT THE TABLE 491 guments that educational and economic markets should reward compliance with uniform merit-based criteria with no regard for context or history.79 With regard to democratic exchange and governance, however, the need for commitment from a diverse, broad base of perspectives suggests that we should not limit political community with narrow, merit-based assumptions.80 In challenging the merit-based vision, perhaps the most difficult questions ask how far courts and legislatures should go to promote inclusion. Should decisionmakers consider every context-specific issue that impacts any individual in interpreting and restructuring election laws? At what point does this analysis become too unmanageable and impracticable? These questions reflect inevitable tensions between administrative convenience on the one hand and important substantive values related to inclusiveness and participation on the other. This Article does not purport to identify the proper place to draw a bright line between the two, as further discussion is necessary to determine how to balance the competing values. It is clear, however, that mechanical rules that prioritize administrative convenience over a meaningful view of participation value the time of administrators over citizens’ interests in democratic inclusion. Administrative convenience alone cannot be the primary basis of a democracy that purports to reflect the will of the people.81 CONCLUSION Examining race allows us to see more clearly the shortcomings of merit-based assumptions of democracy harbored by the majority in supra note 60, at 397 (noting that majority rule and formal equality through the Fourteenth Amendment “have created havens for racial and political gerrymandering, while at the same time providing inadequate protection for the rights of representation of Blacks and other discrete and insular minorities”). 79. See, e.g., STEPHAN THERNSTROM & ABIGAIL THERNSTROM, AMERICA IN BLACK AND WHITE: ONE NATION, INDIVISIBLE 171-80, 393-461 (1997) (criticizing affirmative action programs); see also SHELBY STEELE, THE CONTENT OF OUR CHARACTER: A NEW VISION OF RACE IN AMERICA (1990). 80. Cf. ABIGAIL M. THERNSTROM, WHOSE VOTES COUNT? 242 (1987) (admitting that “[a] white denied a seat on a city council cannot claim entitlement on the ground of ‘merit.’ . . . Qualification for office is not measured by meritocratic standards in the customary sense”); Karlan & Levinson, supra note 44, at 1202 (arguing that “the Court’s attempt to integrate voting rights law into its more general approach to affirmative action is both misguided and incoherent . . . because government decisionmaking with respect to voting, at least in its functional sense of rationing and apportioning the power to govern, is different from other governmental decisionmaking”); Tucker, supra note 29, at 454-55 (criticizing the merit-based approach as applied to voting, distinguishing redistricting from employment, contracting, and college admissions decisions). 81. Cf. Dunn v. Blumstein, 405 U.S. 330, 351 (1972) (“States may not casually deprive a class of individuals of the vote because of some remote administrative benefit to the State.” (quoting Carrington v. Rash, 380 U.S. 89, 96 (1965))). 492 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:469 Bush v. Gore, and to understand why this limited vision of democracy inadequately protects the political rights of racial minorities and other Americans as well. By conditioning political participation on the capacity to create a machine-readable ballot, the merit-based vision interferes with the ability of groups of voters to identify with one another in order to form a political community, create alliances with others of different backgrounds, and use the vote to enact political change. The merit-based vision also conveys an expressive harm of exclusion that may lead to democratic instability. Improved dialogue and a more comprehensive understanding of democratic visions and election procedures generally, as well as Bush v. Gore specifically, require the integration of race as an important analytical tool. JUDGING JUDGING: THE PROBLEM OF SECONDGUESSING STATE JUDGES’ INTERPRETATION OF STATE LAW IN BUSH V. GORE HAROLD J. KRENT* I. BUSH V. GORE AND JUDICIAL ENFORCEMENT OF ARTICLE II, SECTION 1 .......... II. SECOND-GUESSING STATE COURT CONSTRUCTION OF STATE LAW TO PROTECT FEDERAL RIGHTS ................................................................................................. A. Bouie and its Progeny................................................................................... B. Contracts Clause........................................................................................... III. RETREAT FROM ENFORCEMENT OF BOUIE AND JUDICIAL IMPAIRMENT DOCTRINES ........................................................................................................... A. The Bouie Doctrine ....................................................................................... B. Contracts Clause........................................................................................... C. Enforcement of Article II, Section 1 ............................................................. CONCLUSION ........................................................................................................ 497 502 502 504 511 512 521 526 533 That the election for President in 2000 was decided by a court surprised everyone. That the election was decided by a federal court was even more surprising. But, the fact that three Justices supported that decision based on second-guessing a state court’s construction of state law possibly was more astonishing still. Prior to the 2000 presidential election, no one could have foreseen the pivotal role played by the United States Supreme Court. Never had an election of that magnitude hinged on court intervention. The Supreme Court’s dramatic decision, Bush v. Gore,1 doomed the hopes of former Vice President Gore and precipitated his concession. But, perhaps almost as importantly, the decision immediately cast a pall upon the institution of the Court itself, for the decision appeared to many as a crass political move designed to ensure the election of the Justices’ personal choice for office. The reasoning of the Justices was labeled conclusory, unintelligible, and worse.2 Conservative members of the Court who champion federalism appeared willing in Bush v. Gore to ignore years of precedent by secondguessing the Florida Supreme Court’s construction of state law. The U.S. Supreme Court on countless prior occasions had held that state * Professor and Associate Dean, Chicago-Kent College of Law. I would like to thank Trevor Morrison and Steve Siegel for their excellent comments on an earlier draft, and I would like to thank the Florida State University College of Law for taking the initiative to host this symposium. 1. 531 U.S. 98 (2000). 2. For a sampling of the editorials, see Fred Barbash, A Brand New Game, WASH. POST, Dec. 17, 2000, at B1; Alan Dershowitz, The Supreme Court and the 2000 Election, SLATE, July 2-9, 2001, at http://slate.msn.com/?id=111313; Anthony Lewis, Legitimacy: Supreme Court Decision Undercuts Americans’ Belief in Judiciary, DALLAS MORNING NEWS, Dec. 12, 2000, at 27A; Kirk Loggins & John Shiffman, Supreme Court’s Decision to Stop Recount Spurs Ethical Political Debate, THE TENNESSEAN, Dec. 14, 2000, at 1A. 493 494 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 supreme courts are the final expositors of the meaning of state law.3 Under our system of governance, federal courts have no role in overseeing or participating in evolution of state law. Nonetheless, three United States Supreme Court Justices in Bush v. Gore would have held that the Florida Supreme Court’s decision ordering a recount violated Article II, Section 1’s requirement that “[e]ach State shall appoint, in such Manner as the Legislature thereof may direct” electors for President and Vice President.4 They reasoned that the Florida state court’s “interpretation of the Florida election laws impermissibly distorted them beyond what a fair reading required”5 and thereby violated Article II, Section 1 of the Constitution.6 Indeed, that was the Bush team’s principal argument in its brief: the Florida court in effect made new law in interpreting the contest provisions and thus trampled upon the state legislature’s federal constitutional right to choose the manner in which electors are selected.7 The three Justices, therefore, would have secondguessed the Florida Supreme Court’s interpretation of Florida law to protect a federal constitutional guarantee.8 That conclusion seems to turn federalism principles on their head and foretells an increased role for the Supreme Court in subsequent disputes over the interpretation and application of state law. 3. E.g., Mullaney v. Wilbur, 421 U.S. 684 (1975); Murdock v. City of Memphis, 87 U.S. (20 Wall.) 590 (1874) (holding that Congress had not authorized U.S. Supreme Court review of state law issues). Despite those precedents, the Court’s earlier decision in Bush v. Palm Beach County Canvassing Board, 531 U.S. 70 (2000), signaled its intent to review the Florida court’s interpretation of its own state’s law, and the question of whether the Florida court had acted properly in construing state law occupied the headlines leading to the second U.S. Supreme Court decision. Interestingly enough, however, the Supreme Court prior to 1863 deferred to state court construction of state law but had not ruled out its own role as a categorical matter. See Green v. Neal’s Lessee, 31 U.S. (6 Pet.) 291, 298 (1832) (finding that state tribunals had no “power to bind this Court”). The Court indicated change in Leffingwell v. Warren, 67 U.S. (2 Black) 599, 603 (1863). 4. U.S. CONST. art. II, § 1, cl. 2; see also McPherson v. Blacker, 146 U.S. 1, 35 (1892) (stating that Article II confers “plenary power to the state legislatures in the matter of the appointment of electors”). 5. Bush v. Gore, 531 U.S. at 115. 6. This Article assumes the correctness of the Court’s exposition of Article II, which the dissenting Justices largely did not dispute. 7. Brief for Petitioner at 19-33, Bush v. Gore, 531 U.S. 98 (2000) (No. 00-949). Indeed, some have speculated that the concurrence was originally slated to be the majority opinion. See David G. Sauvage & Harvey Weinstein, Right to Vote Led Justices to 5-4 Ruling, L.A. TIMES, Dec. 14, 2000, at A1. At oral argument, Justice O’Connor twice remarked that “it just does look like a very dramatic change made by the Florida court,” and Justice Kennedy offered a similar observation. Oral Argument Transcript at 38-39, Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000) (No. 00-836), available at http:// www.supremecourtus.gov/florida.html. 8. Chief Judge Richard Posner in his recent book wrote that the changed law argument would have provided the Court with the most solid legal ground upon which to base its result. According to Posner, the argument at a minimum was “respectable.” RICHARD POSNER, BREAKING THE DEADLOCK: THE 2000 ELECTION, THE CONSTITUTION, AND THE COURTS 152 (2001). 2001] JUDGING JUDGING 495 Despite the seeming affront to federalism, however, the three Justices were on firm ground in holding that the U.S. Supreme Court can review a state court’s construction of state law in order to protect a federal right.9 In a variety of circumstances, that Court has scrutinized state court decisionmaking to ensure that state judges had not altered state law so as to defeat federal rights. For instance, in Bouie v. City of Columbia,10 the Court held that South Carolina’s interpretation of a state trespass statute violated federal due process guarantees by changing the law in effect at the time of the defendant’s offense.11 Similarly, the Court in Indiana ex rel. Anderson v. Brand12 reviewed the Indiana Supreme Court’s determination that a public school teacher did not enjoy any contract rights under Indiana law so that it could ascertain whether subsequent legislation had impaired her rights under the Contracts Clause.13 The Court has similarly reviewed state court construction of state laws in treaty cases whenever the federally protected right may be lost by a state court’s overly creative interpretation of state law.14 The three Justices thus had ample precedent for reviewing state judges’ interpretation of state law to ensure protection for the constitutionally protected interest in state legislative selection of electors.15 Moreover, the Court in both the Bouie and Contracts Clause contexts has inquired whether state judicial decisions have so changed state law as to deprive an individual or firm of the settled expectations protected under the Constitution. In order to protect federal rights safeguarding against excessive retroactivity, the Court has not 9. This is not to suggest, however, that the Court should have intervened in the dispute in the first instance. Rather, I contend that the three Justices’ consideration of whether the Florida Supreme Court impermissibly changed state law is neither novel nor unwarranted. 10. 378 U.S. 347 (1964). 11. Id. at 349. 12. 303 U.S. 95 (1938). 13. Lower courts have also assessed whether state courts have changed state law governing election disputes. For an insightful analysis, see Richard H. Pildes, Judging “New Law” in Election Disputes, 29 FLA. ST. U. L. REV. 691 (2001); see also POSNER, supra note 8, at 159 (“Nothing is more infuriating than changing the election rules after the outcome of the election, conducted under existing rules, is known.”). 14. See Fairfax’s Devisee v. Hunter’s Lessee, 11 U.S. (7 Cranch) 603, 626-28 (1813) (concluding that state court erred in determining whether 1789 ejectment order constituted a confiscation under a 1783 peace treaty with Great Britain); see also NAACP v. Alabama ex rel. Patterson, 357 U.S. 449, 456 (1958) (second-guessing state court’s application of state procedural rule that had the effect of barring review of a federal claim). 15. An interesting exception is posed in takings cases, where—despite the concurrence’s implication to the contrary, see Bush v. Gore 531 U.S. 98, 115 n.1 (2000) (Rehnquist, C.J., concurring)—the Court has steadfastly refused to hold that judicial changes in the law can effect takings of property. See, e.g., Brinkeroff-Faris Trust & Sav. Co. v. Hill, 281 U.S. 673 (1930); see also Barton H. Thompson, Jr., Judicial Takings, 76 VA. L. REV. 1449 (1990) (addressing the Court’s failure to permit changes in judicial doctrine to constitute takings under the Constitution). 496 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 only second-guessed determinations by the state’s highest court on state law, but asked whether state judicial pronouncements changed prior law. Irrespective of the doctrinal pedigree, however, the Rehnquist concurrence is extraordinary in applying the above principles. When the Court has reviewed state law decisionmaking in other contexts, it has afforded wide latitude to state court decisionmaking. It has only disputed state court construction of state law on rare occasions, and it has concluded that state court decisions made “new” law even more rarely. The Justices’ willingness in Bush v. Gore to hold that the state court interpretation of state law changed Florida’s law flies in the face of the Court’s reticence to examine state court decisions closely, even in the contexts in which federal rights undeniably are at stake. The Court has taken great pains in other settings not to substitute its own view of state law for that of state courts. In Bush v. Gore, the instrument of last resort became a frontline weapon of attack. Part I of this Article examines the state law issues in Bush v. Gore and the Rehnquist concurrence’s willingness to substitute its own interpretation for that of the Florida Supreme Court. The Rehnquist concurrence focused principally on the Florida court’s refusal to defer to the certification of election results and its determination that the intent of the voter standard should govern. In contrast, the dissenting Justices painted a very different picture of Florida’s legislative terrain. And, they lambasted the concurrence for failing to defer to the state court’s interpretation. Part II of this Article sketches two contexts in which the U.S. Supreme Court has reviewed judicial interpretations to determine whether the interpretations “changed” or “altered” preexisting law. Under the Bouie doctrine, federal courts have assessed state court construction of state law to determine whether any change in doctrine violated the defendant’s due process rights. Similarly, under the Contracts Clause, federal courts in the past questioned state court articulation of state contract law in order to protect the federal right that the state not impair the obligation of contracts. Underlying decisions in the two contexts swirl notions of fair notice, predictability, and mistrust for lower court judicial power. To some extent the Article II, Section 1 setting shares these concerns. State court judges might alter state law in a way that favors their own political leanings or futures. Furthermore, stability and predictability are particularly important values when elections are at stake. In the absence of review, state courts, by interpreting state law adventurously, could nullify the directive in Article II that it is the province of state legislatures to determine the manner in which elec- 2001] JUDGING JUDGING 497 tors to the presidential election are to be chosen. The Rehnquist concurrence was on solid ground in asserting the power to review. Yet, with limited exceptions, the Supreme Court in practice has upheld state court interpretation of state law in the Bouie and Contracts Clause settings. Part III sketches some of the means by which federal courts have avoided concluding that state law had “changed.” Through a mixture of fictions and judicial legerdemain, federal judges have concluded that state judicial interpretations in both the Bouie and Contracts Clause contexts were foreshadowed in some way. Indeed, in the Contracts Clause setting, the federal courts subsequently withdrew from the field altogether. The Court is now willing to defer completely to state judicial construction of the state law determinants of the federally protected right. Part III then returns to Bush v. Gore through the prism of these precedents. Although the principle underlying the Rehnquist concurrence is sound, the application is bewildering. The Florida Supreme Court’s construction of Florida law, while in no way dictated by precedent or the plain language of the statutory scheme, was at a minimum, plausible. The U.S. Supreme Court has consistently failed to disturb far more questionable state court decisions in the Bouie and Contracts Clause contexts. There is nothing in the Article II setting that demands a more searching inquiry than in the Bouie or Contracts Clause contexts. Sound policy reasons support the Court’s prior reluctance to second-guess state court interpretations of state law. First, as a matter of judicial administration, permitting challenges to judicial decisions that “change” the law is highly problematic because many losing parties will have the incentive to challenge state court judgments in the Article II, Contracts Clause, and Bouie contexts in federal court. Second, given that judges always have made law interstitially through decisions, distinguishing which interpretations change the law more dramatically than others is quite daunting. Finally, federalism concerns plainly counsel against exacting scrutiny of state court interpretations of state law, even where federal interests are at stake. Thus, in comparison to decisions in related fields, the three Justices’ conclusion that the Florida Court’s decision was so unforeseeable that it changed law is nothing short of startling. I. BUSH V. GORE AND JUDICIAL ENFORCEMENT OF ARTICLE II, SECTION 1 Florida’s legislature, like that in every state today, has determined that presidential electors should be appointed by direct elec- 498 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 tion. Its legislatively designed mechanism for selecting electors in 200016 was relatively detailed, though far from clear. Under the statute, “[v]otes cast for the actual candidates for President and Vice President shall be counted as votes cast for the presidential electors supporting such candidates.”17 The legislature delegated to county canvassing boards the responsibility of administering elections.18 The legislature also directs the canvassing boards to provide results to the state Elections Canvassing Commission, comprised of the Governor, the Secretary of State, and the Director of the Division of Elections.19 After the election, the county canvassing boards were to receive returns from precincts, count the votes, and conduct a mandatory recount if a candidate lost by less than .5% of the vote.20 In addition, the boards were to certify election returns with the Secretary of State by 5 p.m. on the seventh day following the election, and the Elections Canvassing Commission subsequently certified the results of the election.21 A losing candidate could have contested the certified election results if, among other things, there was “[r]eceipt of a number of illegal votes or rejection of a number of legal votes sufficient to change or place in doubt the result of the election.”22 The statute further provided that “[t]he circuit judge to whom the contest is presented may fashion such orders as he or she deems necessary to ensure that each allegation in the complaint is investigated, examined, or checked, to prevent or correct any alleged wrong, and to provide any relief appropriate under such circumstances.”23 On November 26, 2000, the Florida Elections Canvassing Commission certified the results of the election in favor of Governor Bush. The next day, Vice President Gore filed a complaint in Leon County Circuit Court contesting the certification. A two-day evidentiary hearing focused on the so-termed undervote in counties in which machines tabulated votes on punch cards. Vice President Gore introduced evidence suggesting that the tabulation missed a sufficient number of votes that would have swung Florida’s election in his favor. Governor Bush defended by arguing that no sound statistical method existed for determining the extent of the undervote, and he 16. The Florida Legislature passed wholesale amendments to its elections procedures in the 2001 session. See Florida Election Reform Act of 2001, 2001 Fla. Laws ch. 40, at 11773. 17. FLA. STAT. § 103.011 (2000) (amended 2001). 18. Id. § 102.141 (amended 2001). 19. Id. § 102.111 (amended 2001). 20. Id. § 102.141(4) (amended 2001). 21. Id. §§ 102.111-.112 (amended 2001). 22. Id. § 102.168(3)(c). 23. Id. § 102.168(8) (repealed 2001). 2001] JUDGING JUDGING 499 asserted that it was not clear whether a different tabulation would have tipped the election.24 Leon County Circuit Court Judge Sanders Sauls agreed, holding that Vice President Gore had failed to meet the required burden of proof.25 Judge Sauls ruled that the circuit court could not overrule the canvassing commission’s certification absent an abuse of discretion. Gore’s burden, according to Judge Sauls, was to prove, “but for the irregularity, or inaccuracy claimed, the result of the election would have been different, and he or she would have been the winner. . . . [A] reasonable probability that the results of the election would have been changed must be shown.”26 He concluded that “there [was] no credible statistical evidence, and no other competent evidence to establish by a preponderance of a reasonable probability that the results of the statewide election in the State of Florida would be different from the result which had been certified . . . .”27 Judge Sauls did not, however, issue any specific findings of fact.28 On appeal, the Florida Supreme Court reversed in part. The court determined first that Judge Sauls had erred in extending deference to the certification of the state election authorities.29 Rather, it was for the circuit court itself to determine whether “a number of illegal votes” had been counted or whether the state had rejected “a number of legal votes sufficient to change or place in doubt the result of the election.”30 Furthermore, the court held that the contestant did not need to show a “reasonable probability” that the election would have resulted differently if proper votes had been counted. Instead, a contestant need only demonstrate that “available data shows that a number of legal votes would be recovered from the entire pool of the subject ballots, which, if cast for the unsuccessful candidate, would change or place in doubt the result of the election.”31 A possibility, in other words, would have sufficed. In addition, the Florida Supreme Court concluded that Gore satisfied the threshold requirement by demonstrating that, upon consideration of the thousands of undervote ballots presented—those ballots with no registered vote—the number of legal votes therein were sufficient to at least place in doubt the result of the election.32 To 24. See Gore v. Harris, 772 So. 2d 1243, 1247 (Fla. 2000), rev’d sub nom. Bush v. Gore, 531 U.S. 98 (2000). 25. Gore v. Harris, CV No. 00-2808, 2000 WL 1770257 (Fla. 2d Cir. Ct. Dec. 4, 2000), rev’d, 772 So. 2d 1243 (Fla. 2000), rev’d sub nom. Bush v. Gore, 531 U.S. 98 (2000). 26. Gore v. Harris, 772 So. 2d at 1255. 27. Id. 28. Id. at 1247. 29. Id. at 1252. 30. Id. at 1253 (emphasis omitted) (quoting FLA. STAT. § 102.168(3)(c) (2000)). 31. Id. at 1256. 32. Id. 500 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 reach its conclusion, it construed the statutory term “rejection” to include instances where the county canvassing board failed to count legal votes; that is, votes in which the intent of the voter could be discerned, irrespective of whether the ballots had been mismarked.33 The court relied on a neighboring statutory provision, section 101.5614(5), to reject the view of the Secretary of State—reflected in an administrative opinion34—and the Bush camp that “legal vote” referred only to a vote “properly executed in accordance with the instructions provided to all registered voters.”35 As a remedy, the court determined that there should be a recount of the undervote in all Florida counties, not just those challenged by Gore in his contest petition.36 In dissent, Chief Justice Wells retorted that the majority’s reading of the contest provisions “has no foundation in the law of Florida as it existed on November 7, 2000 . . . .”37 He would have shown deference to the county canvassing boards and would have held that a contest could only be granted if the plaintiff demonstrated a “substantial noncompliance with election laws,” which Gore had not done.38 The Rehnquist concurrence agreed with Chief Justice Wells. The opinion charged that the Florida Supreme Court decision “empties certification of virtually all legal consequence during the contest, and in doing so, departs from the provisions enacted by the Florida Legislature.”39 The concurring Justices asserted that the election code vests discretion in the county canvassing boards to determine whether to recount, and thus deference should have been afforded to the boards,40 as Circuit Judge Sauls had held. Moreover, the concurrence stated that “Florida statutory law cannot reasonably be thought to require the counting of improperly marked ballots.”41 Thus, the concurrence, unlike the Florida Supreme Court, would have held that Florida law does not turn on an intent of the voter standard when a ballot had been mismarked. The statewide recount ordered by the Florida court similarly found no support in Florida law. According to the concurrence, the Florida court had not merely misread the legislative scheme but rearranged it beyond recognition: “the Florida Supreme Court’s interpretation of the Florida election 33. Id. at 1257. 34. See Bush v. Gore, 531 U.S. 98, 119 (2000) (Rehnquist, C.J., concurring) (addressing 00-13 Fla. Op. Div. of Elec. (2000)). 35. Gore v. Harris, 772 So. 2d at 1257. 36. Id. at 1261-62. 37. Id. at 1263 (Wells, C.J., dissenting). 38. Id. at 1264-68. 39. Bush v. Gore, 531 U.S. at 118 (Rehnquist, C.J., concurring). 40. Id. at 117-18 (Rehnquist, C.J., concurring). 41. Id. at 118-19 (Rehnquist, C.J., concurring). 2001] JUDGING JUDGING 501 laws impermissibly distorted them beyond what a fair reading required . . . .”42 No deference was due because, “[t]o attach definitive weight to the pronouncement of a state court, when the very question at issue is whether the court has actually departed from the statutory meaning, would be to abdicate our responsibility to enforce the explicit requirements of Article II.43 The concurrence thus advocated active, if not de novo, review of state law requirements.44 In dissent, Justice Stevens concluded that the Florida Supreme Court had not made “any substantive change in Florida electoral law.45 Rather, “[i]t did what courts do—it decided the case before it in light of the legislature’s intent to leave no legally cast vote uncounted.46 Justice Souter’s dissent similarly concluded that “[n]one of the state court’s interpretations is unreasonable to the point of displacing the legislative enactment quoted.”47 The opinion noted that “other interpretations were of course possible, and some might have been better than those adopted by the Florida court’s majority,”48 but concluded that the state court’s view was “within the bounds of reasonable interpretation . . . .”49 Given that the statute did not define “legal vote,” Justice Souter explained that the state court’s adoption of the intent of the voter standard from a neighboring provision was highly plausible.50 Similarly, its reading of “rejection” to mean not counted was reasonable—there was no requirement to read into rejection some kind of machine malfunction.51 Moreover, Justice Souter found no clear Florida statutory basis for determining whether the contest standard required a plaintiff to demonstrate a probability as opposed to a possibility that the outcome would be different if the undervotes were tabulated.52 Similarly, Justice Ginsburg’s separate dissent retorted that “disagreement with the Florida court’s interpretation of its own State’s law does not warrant the conclusion that the justices of that court have legislated.”53 Justice Breyer’s dissenting opinion also decried the analysis in the Rehnquist concurrence: “I do not see how one 42. Id. at 115 (Rehnquist, C.J., concurring). 43. Id. (Rehnquist, C.J., concurring). 44. At one point the concurrence suggested that the Article II, Section 1 context was an area “in which the Constitution requires the Court to undertake an independent, if still deferential, analysis of state law.” Id. at 114. (Rehnquist, C.J., concurring). The level of deference intended is unclear. 45. Id. at 127-28 (Stevens, J., dissenting). 46. Id. at 128 (footnote ommited) (Stevens, J., dissenting). 47. Id. at 131 (Souter, J., dissenting). 48. Id. (Souter, J., dissenting). 49. Id. (Souter, J., dissenting). 50. Id. at 132 (Souter, J., dissenting). 51. Id. (Souter, J., dissenting). 52. Id. at 132-33 (Souter, J., dissenting). 53. Id. at 136 (Ginsburg, J., dissenting). 502 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 could call its plain language interpretation of a 1999 statutory change so misguided as no longer to qualify as judicial interpretation or as a usurpation of the authority of the State legislature.”54 Both the concurring and dissenting opinions accepted the premise that Article II, Section 1 prohibits state courts from changing the manner in which presidential electors are selected. Moreover, the opinions also agreed that claims asserting that state courts had changed state law were subject to U.S. Supreme Court oversight.55 They disagreed, however, over the extent to which Florida’s highest court changed Florida law and the measure of deference that the Court should afford to state court construction of state law. II. SECOND-GUESSING STATE COURT CONSTRUCTION OF STATE LAW TO PROTECT FEDERAL RIGHTS The Rehnquist concurrence’s second-guessing of state court interpretations of state law is not unique. In a variety of other contexts, the Supreme Court has scrutinized state court interpretation of state law to ensure protection for federal rights. Accordingly, this Part focuses on two areas—the Bouie doctrine and Contracts Clause—in which the Court at times has exercised the power to review state court interpretations of state law. A. Bouie and its Progeny Prior to Bouie v. City of Columbia,56 judicial change in the criminal law violated no federal constitutional guarantee. Although the Ex Post Facto Clause prohibited retroactive imposition of legislative change, members of the Court had long opined that judicial change did not fall within the Ex Post Facto Clause.57 In Bouie, however, the Court reversed a trespass conviction on the ground that the state conviction rested on an unexpected construction of the state trespass statute by the South Carolina Supreme Court.58 The state supreme court had construed a trespass statute prohibiting “entry upon the lands of another . . . after notice from the owner or tenant prohibiting such entry” to apply to African-American 54. Id. at 152 (Breyer, J., dissenting). Neither Justice O’Connor nor Justice Kennedy weighed in on the issue. Although one can speculate as to reasons for their reticence, the upshot is that no majority of the Court agreed on the Article II issue. 55. Justice Ginsburg’s opinion, however, calls for almost total deference to the state court construction. Id. at 135-44 (Ginsburg, J., dissenting). 56. 378 U.S. 347 (1964). 57. See, e.g., James v. United States, 366 U.S. 213, 224 (1961) (Black, J., dissenting) (“[T]he ex post facto [clause] . . . has not ordinarily been thought to apply to judicial legislation.”); Ross v. Oregon, 227 U.S. 150, 161 (1913) (holding that the Ex Post Facto Clause “is a restraint upon legislative power and concerns the making of laws, not their construction by the courts”). 58. Bouie, 378 U.S. at 350. 2001] JUDGING JUDGING 503 demonstrators at a lunch counter who entered the lunch counter premises before the owner asked them to leave.59 Although the Supreme Court noted that the South Carolina Supreme Court’s construction of the statute—applying it to individuals refusing to leave another’s property after being so requested—was possible, it held that retroactive application of the interpretation violated due process because the statute did not “give fair warning of the conduct that it makes a crime.”60 The Court acknowledged that it was reading ex post facto principles into the Due Process Clause. As the Court explained in greater depth: an unforeseeable judicial enlargement of a criminal statute, applied retroactively, operates precisely like an ex post facto law. . . . If a state legislature is barred by the Ex Post Facto Clause from passing such a law, it must follow that a State Supreme Court is barred by the Due Process Clause from achieving precisely the same result by judicial construction.61 Given that the state court judges apparently strained to affix some criminal liability upon the civil rights protestors, prohibiting retroactive application of unforeseeable judicial constructions may have had the salutary effect of improving future state judicial decisionmaking. When evaluating the deeds of an accused, judges may wish to bend the law to ensure that the accused is punished by including the objectionable conduct within a criminal enactment. After Bouie, South Carolina and other courts might hesitate before again interpreting criminal provisions merely to ensure punishment for the offender before the court. Lower courts have extended the Supreme Court’s due process analysis to unforeseeable judicial changes in sentencing structure as well,62 making the theoretical reach of the Due Process and Ex Post Facto Clauses congruent. 59. Id. at 349-50. 60. Id. at 350. For a contextual analysis of Bouie as part of the Court’s response to the civil rights movement, see Jack Greenberg, The Supreme Court, Civil Rights and Civil Dissonance, 77 YALE L.J. 1520, 1530 (1968); Michael Klarman, An Interpretive History of Modern Equal Protection, 90 MICH. L. REV. 213, 273-76 (1991). 61. Bouie, 378 U.S. at 353-54. The Court adopted similar reasoning several years later in Rabe v. Washington, 405 U.S. 313, 316 (1972). See also Marks v. United States, 430 U.S. 188 (1977) (holding that the Due Process Clause prevented retroactive application of a new judicial test for ascertaining pornography restrictions); Douglas v. Buder, 412 U.S. 430, 432 (1973) (determining that the judge’s construction of the term “arrest” to include traffic citation was unforeseeable); cf. Helton v. Fauver, 930 F.2d 1040, 1042 (3d Cir. 1991) (holding that New Jersey’s construction of jurisdictional statute affecting prosecution of juveniles was unforeseeable); Moore v. Wyrick, 766 F.2d 1253, 1257 (8th Cir. 1985) (holding that the state court’s change in felony murder doctrine was not foreseeable). 62. See, e.g., Green v. Catoe, 220 F.3d 220 (4th Cir. 2000), abrogated on other grounds by Bell v. Jarvis, 236 F.3d 149, 160 (4th Cir. 2000); Dale v. Haeberlin, 878 F.2d 930 (6th 504 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 Because interpretation is not a mechanical exercise, judicial constructions—particularly of broadly worded texts—seem like conventional lawmaking. Judicial interpretations that alter the substantive definition of a crime can imperil liberty to the same extent as legislative change that would violate the ex post facto doctrine. Surprising judicial decisions deprive an offender of notice to the same extent as a new legislative enactment enlarging the scope of a criminal provision.63 B. Contracts Clause The Constitution provides that “[N]o State shall . . . [pass any] law impairing the obligation of contracts . . . .”64 The Contracts Clause operates as a substantial protection against state interference in private contractual obligations. For instance, in Allied Structural Steel v. Spannaus65 the Court struck down a Minnesota statute that safeguarded certain workers’ expectations of receiving pensions. The Court reasoned that the law “nullifies express terms of the company’s contractual obligations and imposes a completely unexpected liability in potentially disabling amounts.”66 The Supreme Court has held that the question of whether the contract was impaired, as in Spannaus, is a federal question. Given that the Constitution introduces the term “impairing,” that conclusion is logical. But a more intriguing issue is whether federal courts, in order to protect federal rights, can second-guess a state court’s determination whether a contract exists. Consider Indiana ex rel. Anderson v. Brand.67 There, a public school teacher asserted that she had earned tenure under the existing statutory framework.68 Her contract had contained the clause: “[I]t is further agreed by the contracting parties that all of the provisions of the Teachers’ Tenure Law, approved March 8, 1927, shall be in full force and effect in this contract . . . .”69 Upon termination in Cir. 1989); Knapp v. Cardwell, 667 F.2d 1253 (9th Cir. 1982); Foster v. Barbour, 613 F.2d 59 (4th Cir. 1980). 63. Indeed, the concern for constructive notice may play a larger role in restraining judicial, rather than legislative, retroactivity. When the legislature changes a standard of conduct retroactively, there is little concern for augmenting the power of police and prosecutors. In contrast, when judges fashion new doctrine and apply it to the case at bar, police and prosecutors as opposed to legislators have already selected the particular defendant for punishment. In any common law of crimes system, prosecutors and police play a more important role than with legislatively defined crimes. 64. U.S. CONST. art. I, § 10, cl. 1. For a history, see BENJAMIN FLETCHER WRIGHT, JR., THE CONTRACT CLAUSE OF THE CONSTITUTION 3-26 (1938). 65. 438 U.S. 234 (1978). 66. Id. at 247. 67. 303 U.S. 95 (1938). 68. Id. at 97. 69. Id. at 97. 2001] JUDGING JUDGING 505 July 1933, she sued in state court, alleging a breach of contract. The Indiana Supreme Court upheld her dismissal on the ground that the Teachers Tenure Law of 1927 had been repealed with respect to township public school teachers and that, under Indiana law, no tenure attached and therefore that no tenure rights could currently be enjoyed.70 The U.S. Supreme Court acknowledged that the question posed was “one primarily of state law,” but continued that, “in order that the constitutional mandate may not become a dead letter, we are bound to decide for ourselves whether a contract was made, what are its terms and conditions, and whether the State has, by later legislation, impaired its obligation.”71 The Court explained that “[u]ntil its decision in the present case the Supreme Court of the State had uniformly held that the teacher’s right to continued employment by virtue of the indefinite contract created pursuant to the Act was contractual.”72 The Supreme Court therefore rejected the state court’s apparent flip-flop, and held that the teacher’s contract had been impaired by the 1933 law. In order to protect the petitioner’s contract rights, the Supreme Court thus rejected the state Supreme Court’s analysis of whether a contract existed under state law. More recently, the Court in General Motors Corp. v. Romein73 clarified that, although “ultimately we are bound to decide for ourselves whether a contract was made,” the Court “accord[s] respectful consideration and great weight to the views of the State’s highest court.”74 Federal courts will second-guess state courts’ determination of whether contractual rights exist under state law. But the Contracts Clause context also highlights another problem: what if it is a state judicial decision as opposed to state legislation which impairs the obligation of contract? In Brand, for instance, it may have not been the 1927 law that stripped away the teacher’s expectations but the judicial decision construing the law’s requirements. Should judicial changes in statutory interpretation be treated the same as a legislative amendment? As in the ex post facto context, the language of the clause seems directed to state legislatures as opposed to state judiciaries, but the impact of any impairment would be identical on the firm whose contract rights had been impaired.75 Judicial change can be as devastating as legislative alteration. 70. Id. at 98. 71. Id. at 100. 72. Id. at 105. 73. 503 U.S. 181 (1992). 74. Id. at 187 (quoting Brand, 303 U.S. at 100). 75. For a review of the original understanding, see Barton H. Thompson, Jr., The History of the Judicial Impairment “Doctrine” and Its Lessons for the Contract Clause, 44 STAN. L. REV. 1373, 1384-87 (1992). 506 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 The mid-nineteenth century case of Gelpcke v. City of Dubuque76 is representative. In 1800s Dubuque and elsewhere, taxpayers were stuck holding the bag when municipalities could not afford to honor their obligation on railway bonds. The predicted economic boom did not come to pass. Iowa’s courts had previously sustained the constitutionality of the municipalities’ authority to issue the bonds, but taxpayers nevertheless sued the City of Dubuque, renewing their argument. This time, the state court held that the city had no such power under Iowa’s Constitution.77 Bondholders, seeing the handwriting on the wall, sued the city in federal court for recovery of interest on the bonds. The case eventually was heard by the U.S. Supreme Court. The Supreme Court, relying on its diversity jurisdiction,78 held that federal courts were bound by decisions of state courts on state law at the time the contract was concluded, but that subsequent state decisions as to the substance of state law were not binding. Accordingly, the Court relied on the earlier Iowa judicial pronouncements and concluded that the city was obligated to pay the bondholders. Changes in judicially fashioned law could only be applied prospectively: “otherwise . . . rights acquired under a statute may be lost by its repeal.”79 The Court concluded that “[w]e shall never immolate truth, justice, and the law, because a State tribunal has erected the altar and decreed the sacrifice.”80 As Chief Justice Waite later commented in Douglass v. Pike County,81 [a]fter a statute has been settled by judicial construction, the construction becomes, so far as contract rights acquired under it are concerned, as much a part of the statute as the text itself, and a change of decision is to all intents and purposes the same in its effect on contracts as an amendment of the law by means of a legislative enactment.82 76. 68 U.S. (1 Wall.) 175 (1863). 77. State ex rel. Burlington & Mo. R.R. v. County of Wapello, 13 Iowa 388, 407 (1862). 78. Under the regime of Swift v. Tyson, 41 U.S. (16 Pet.) 1, 18 (1842), federal courts exercised the power to interpret state law to protect the out-of-state party. The Court in Gelpcke and its progeny, however, appeared to base its rulings as well on the power to resolve the Contracts Clause claim. See, e.g., Carondelet Canal & Navigation Co. v. Louisiana, 233 U.S. 362, 375 (1914); Douglas v. Kentucky, 168 U.S. 488, 498 (1898); WRIGHT, supra note 64, at 239-41; Thompson, supra note 75, at 1412-13. The Court in Township of Pine Grove v. Talcott, 86 U.S. (19 Wall.) 666, 678 (1873), for instance, stated that “the National Constitution forbids the States to pass laws impairing the obligations of contracts. In cases properly brought before us that end can be accomplished unwarrantably no more by judicial decisions than by legislation.” 79. Gelpcke, 68 U.S. (1 Wall.) at 206. 80. Id. at 206-07. 81. 101 U.S. 677 (1879). 82. Id. at 687. The Court continued that “[t]he true rule is to give a change of judicial construction in respect to a statute, the same effect in its operation on contracts and exist- 2001] JUDGING JUDGING 507 The Gelpcke doctrine likely arose from both ideological and economic considerations. First, federal courts feared that state court elected justices would side with local interests and repudiate the obligations. As the Supreme Court noted in Wade v. Travis County,83 absent federal court review, state decisions “would enable the State to set a trap for its creditors by inducing them to subscribe to bonds, and then withdrawing their own security.”84 Bondholders such as Gelpcke often lived outside the jurisdiction. Second, federal judges may have believed that stability in the capital markets was critical to continued growth of the country. The views expressed in the municipal bond cases are all the more notable given the prevailing view at the time that judges applied or found the law, not made it.85 The Bouie and Contracts Clause contexts are analogous in many respects. Two overlapping rationales justified their invocation. First, in both cases, the Supreme Court was deeply concerned about process. In Bouie, the Court feared that state judges—who largely were elected—would adopt strained interpretations of state statutes and common law in order to penalize a particular offender.86 Such overreaching would redound to their political benefit. Conversely, those subject to criminal sanctions constituted an easy target given their relative poverty and inability to sway elections. The concern for judicial lawlessness was particularly acute in the civil rights era. Similarly, the U.S. Supreme Court applied the Contracts Clause to judicial impairments in large part due to fears that state judges would protect municipalities within their jurisdictions at the expense of bondholders, who may well have lived outside the jurisdiction. Indeed, a number of state legislatures in the years prior to Gelpcke had lessened judges’ independence, which could well have sparked federal judicial concern.87 Retroactive decisionmaking could target unpopular or politically powerless individuals and firms. Moreover, as in the Bouie example, those injured could not easily influence the political process if they were outside the jurisdiction.88 The two doctrines therefore serve as a check on state judicial power in order to ensure that federal constitutional rights are protected. Second, the Bouie and Gelpcke doctrines privileged concerns for settled expectations. In Bouie, the Court reasoned that all criminal ing contract rights that would be given to a legislative amendment; that is to say, make it prospective, but not retroactive.” Id. 83. 174 U.S. 499 (1899). 84. Id. at 509; see also Ga. Ry. & Power Co. v. Decatur, 262 U.S. 432 (1923); McCullough v. Virginia, 172 U.S. 102 (1898). 85. Thompson, supra note 75, at 1419-22. 86. See Bouie v. City of Columbia, 378 U.S. 347, 535 (1964). 87. See Thompson, supra note 75, at 1396-97. 88. In comparison to criminal defendants, however, bondholders could influence elections more effectively in light of their comparable economic clout. 508 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 defendants should have notice of the criminality and consequences of conduct before the fact.89 The rule of law demands that criminal prohibitions be clear before the disputed conduct takes place. Similarly, in Gelpcke, the Court believed that bondholders should understand the scope of their rights and obligations at the time they concluded the contract.90 Economic stability depended heavily on settled expectations. Concerns underlying the Article II, Section 1 directive that state legislatures have exclusive power to direct the manner in which presidential electors are chosen91 similarly militate for federal court oversight. In the absence of review, Congress, governors, or state courts could manipulate the selection process so as to rob Article II of its meaning. Article II, Section 1 reflects the view that individual citizens have no federal constitutional right to vote for electors unless the state legislature so directs. At times state legislatures exercised the power themselves to appoint electors.92 When states vest that right in their citizens, as they now all do, other governmental actors must abide by that choice. Under the Constitution, it remains the province of the state legislature not only to choose who selects electors but also the manner in which the selection takes place.93 As James Madison noted, “The State Judiciarys had not & he presumed wd. Not be proposed as a proper course of appointment of the Presidential electors.”94 A concern for predictability, as in the Bouie and Contracts Clause contexts, is absolutely vital in the election context. When the rules of the game change midstream for criminal punishment or contract enforcement, individual rights undoubtedly suffer. But, when the rules of an election change, faith in the integrity of the election process itself may be compromised. The Constitution mandates that any such change come from the state legislature.95 Moreover, the U.S. Supreme Court scrutinized state court decision making in the Bouie and Contract Clause contexts as a check on judicial power. That concern applies as well in the Article II, Section 1 89. Bouie, 378 U.S. at 352. 90. Gelpcke v. City of Dubuque, 68 U.S. (1 Wall.) 175, 206 (1863). 91. U.S. CONST. art II, § 1, cl. 2. 92. See McPherson v. Blacker, 146 U.S. 1, 28-33 (1892). 93. U.S. CONST. art. II, § 1, cl. 2. 94. 2 THE RECORDS OF THE FEDERAL CONVENTION OF 1787, at 110 (Max Farrand ed., 1966). 95. U.S. CONST. art. II, § 1, cl. 2. Congress has also limited the legislature’s flexibility somewhat by fashioning the safe harbor provision. 3 U.S.C. § 5 (1994) (informing state legislatures that changes after the fact in the manner in which legislators are to be selected may be challenged in Congress). At some point, congressional interference in the state legislative decision might itself violate Article II, Section 1. 2001] JUDGING JUDGING 509 context. Some oversight is needed to check state court jurists’ ability to affect a presidential election. To the extent that history is relevant, the Framers wished to ensure both that the President not become beholden to the electing body and that the electing body not be subject to untoward political pressures. The Framers plausibly sought a compromise between vesting the selection power in an entity that was too powerful and might abuse the power to appoint electors and vesting it in an entity that was not accountable to the people for any manipulation. Unlike state executives, state legislatures were more numerous, less stable, and therefore unlikely to form a coalition over time that could extract promises from a President.96 And, unlike Congress, state legislatures would be more likely receptive to state concerns.97 The state legislature is directly accountable to the electorate and arguably is less likely subject to presidential influence than a governor hoping for a cabinet position (or other influence) or than a state judge hoping for elevation to the federal bench (or other position). Individual judges might be susceptible to untoward pressures to affect the manner in which presidential electors are selected. Nor were judges necessarily subject to sufficient majoritarian checks. The choice of state legislature over national legislature, state executives, and state judiciary was significant but apparently was more a selection among lesser evils. Irrespective of the reason for the constitutional choice, fidelity to the Constitution presupposes some oversight of the elector selection process to ensure that it is the state legislature, as opposed to executive or judiciary, that is exercising the constitutionally entrusted power. There is one important difference to consider. The purpose of Supreme Court review under Bouie and the Contracts Clause is to protect individuals, not the legislative domain. The Clause protects individuals only indirectly to the extent that vesting the power to determine the manner of selecting electors in the legislature preserves their interests. One might argue, therefore, that the structural pro- 96. As Justice Story commented, “The appointment of the president is not made to depend upon any preexisting body of men, who might be tampered with beforehand to prostitute their votes.” 3 JOSEPH STORY, COMMENTARIES ON THE CONSTITUTION 745 (1987). 97. As James Madison explained, “the candidate would intrigue with the Legislature . . . and be apt to render his administration subservient to its views,” 2 JAMES MADISON, JOURNAL OF THE CONSTITUTIONAL CONVENTION 109, and that the national legislature, “being standing bodies, they could and would be courted, and intrigued with by the candidates by their partisans.” Id. at 110. Butler summed up the problem: “The two great evils to be avoided are cabal at home, and influence from abroad. It will be difficult to avoid either if the Election be made by the National Legislature.” Id. at 112; see also 46 CONG. REC. 4398 (1879) (reprinting letter from Madison to Jefferson averting to dangers of congressional interference in Electoral College). 510 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 tection should be enforced less rigorously under Article II, Section 1 than the individual rights under Bouie and the Contracts Clause. Indeed, many might believe that enforcement of Article II should be left completely in Congress’s hands. Congress retains the right under Article II and the Twelfth Amendment to determine whether to count electoral votes from particular states. During the 1876 election, Congress received two slates of electors from Florida and two other states and eventually had to determine which slate to count.98 Congress at other times has determined whether to count contested votes from individual electors.99 The Constitution, therefore, may be seen as vesting enforcement of the Article II, Section 1 right in Congress which, after all, is comprised of representatives from each state. Congress possesses the ultimate power to determine if a state judge or governor has impermissibly infringed the state legislative perrogative to prescribe the manner in which presidential electors are selected. The Article II, Section 1 challenge might, therefore, be viewed as a political question. Similarly, in Luther v. Borden,100 the Court held that the Constitution commits the Republican Guarantee Clause101 to congressional enforcement. In the Borden case, two different governments claimed lawful authority in Rhode Island, and rather than decide between the two, the Court opined that Congress must necessarily decide what government is established in the state before it can determine whether it is republican or not. And when the senators and representatives of a state are admitted into the councils of the union, . . . [a state’s] republican character[] is recognized by the proper constitutional authority. And its decision is binding on every other department of the government . . . .102 Subsequent republican guarantee claims have been held nonjusticiable.103 The Article II claim in Bush v. Gore similarly may be committed to Congress’s determination.104 If Congress accepts the electoral 98. See Day by Day: What Happened in the Electoral College Controversy of 1876-77, HARPWEEK, at http://elections.harpweek.com/9Controversy/events-controversy.htm (last visited Dec. 10, 2001). 99. See LAWRENCE D. LONGLEY & NEIL PIERCE, THE ELECTORAL COLLEGE PRIMER 2000, at 113 (1999). 100. 48 U.S. (7 How.) 1 (1849). 101. U.S. CONST. art. IV, § 4. 102. Borden, 48 U.S. (7 How.) at 42. There was little judicial oversight over federal elections at all prior to Baker v. Carr, 369 U.S. 186 (1962). 103. See, e.g., Pacific States Tel. & Tel. Co. v. Oregon, 223 U.S. 118, 151 (1912); Coyle v. Smith, 221 U.S. 559, 579 (1911). 2001] JUDGING JUDGING 511 votes, then it must have determined that the state legislature appropriately set the manner in which the electors were selected. Moreover, vesting that decision completely in Congress may prevent delays and ensure sufficient time within which to conduct an orderly transition to the new administration. As Justice Breyer stated in the Bush v. Gore dissent, “[C]ongress is the body primarily authorized to resolve remaining disputes.”105 Justice Breyer further noted that “there is no reason to believe that federal law either foresees or requires resolution of such a political issue by this Court.”106 Nevertheless, the Rehnquist concurrence plausibly concluded that at least some review is appropriate.107 In the absence of oversight, a judge, governor, or Congress could meddle in the state’s choice of how electors should be selected. Given that one of the reasons Article II vests the selection power in state legislatures was to prevent congressional interference, it would make little sense—given contemporary understanding of judicial review—to vest all oversight in Congress. As the Framers warned, Congress could become too powerful if it exercised a direct role in selecting presidential electors. Article II, Section 1, therefore, should be seen as an aspect of federalism woven into the Constitution, and accordingly subject to some type of Supreme Court enforcement. Thus, as in the Bouie and Contract Clause contexts, concerns for predictability and accountability support the Supreme Court’s decision to review state court interpretation of state law in the presidential elector setting. The question of whether state legislators have established the manner in which presidential electors are selected arguably does not pose a nonjusticiable political question. But, as Part III discusses, the availability of judicial review should not lead to the conclusion that exacting scrutiny is appropriate. III. RETREAT FROM ENFORCEMENT OF BOUIE AND JUDICIAL IMPAIRMENT DOCTRINES The Justices’ conflict in Bush v. Gore could not have been more clear. Three Justices were convinced that the Florida court’s construction of state law was unmoored from the statute, while the four Justices in dissent were just as emphatic that the Florida court’s opinion reasonably reflected and implemented the intent of the legislature. What neither side fully grappled with, however, was the theoretical and doctrinal difficulty of distinguishing reasonable from un104. Cf. Nixon v. United States, 506 U.S. 234 (1993) (holding that the constitutional command in Article I, Section 3 that the Senate shall “try all Impeachments” is for the Senate alone to resolve). 105. Bush v. Gore, 531 U.S. 98, 154 (2000) (Breyer, J., dissenting). 106. Id. at 155 (Breyer, J., dissenting). 107. Id. at 113 (Rehnquist, C.J., concurring). 512 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 reasonable constructions of law. Examination of the Bouie and Contracts Clause cases provides a cautionary tale. Despite the doctrinal clarity of the Bouie and judicial impairment precedents, the Supreme Court, shortly after articulating the doctrines, began whittling away at their core. That Court and lower courts became extremely wary of second-guessing state court construction of state law and therefore devised a series of devices to avoid the need to upset state court construction of state law. The Rehnquist concurrence in Bush v. Gore simply ignored this history. Chief Justice Rehnquist’s declaration that “[t]o attach definitive weight to the pronouncement of a state court, when the very question at issue is whether the court has actually departed from the statutory meaning, would be to abdicate our responsibility to enforce the explicit requirements of Article II,”108 is undercut by generations of experience in the two analogous contexts. A. The Bouie Doctrine The Rehnquist concurrence noted that the Supreme Court had equated judicial interpretation with change in Bouie v. City of Columbia;109 and that the Bouie line of cases therefore supported its conclusion of a federal constitutional violation in Bush v. Gore.110 The concurrence was half right. Bouie fully supports the concurrence’s position but as this section discusses, developments after Bouie have dramatically limited its reach, undercutting the concurrence’s prop. Just weeks before deciding Bush v. Gore, the Court heard oral argument in a Bouie case. In Rogers v. Tennessee, the defendant brutally knifed a friend after a card game.111 Prosecutors indicted Rogers for attempted murder. The victim, reduced to a vegetative state, lingered for over a year until he died from a kidney infection as a complication of his injuries.112 Accordingly, prosecutors obtained a new indictment, charging Rogers with first-degree murder. A jury subsequently convicted Rogers of second-degree murder.113 On appeal, Rogers argued that preexisting Tennessee law provided that a killing could only be prosecuted as a homicide if the death arose within one year and a day.114 The Tennessee Supreme Court agreed with the defendant’s assessment of Tennessee homicide 108. Id. at 115 (Rehnquist, C.J., concurring). 109. 378 U.S. 347 (1964). 110. Bush v. Gore, 531 U.S. at 115 (Rehnquist, C.J., concurring). 111. State v. Rogers, 992 S.W.2d 393 (Tenn. 1999), aff’d, Rogers v. Tennessee, 121 S. Ct. 1693 (2001). 112. Id. at 395. 113. Id. 114. Id. 2001] JUDGING JUDGING 513 law but abolished the year-and-a-day rule.115 It noted that “[m]odern pathologists are able to determine the cause of death with much greater accuracy than was possible in earlier times.”116 Because the reasons underlying the rule no longer existed, the court decided to overrule it judicially. With respect to whether to apply the rule retroactively, the state court held that abolition of the rule was “not an unexpected and unforseen [sic] judicial construction of a principle of criminal law.”117 The Tennessee court further reasoned that “the rule has fallen into disfavor and had been legislatively or judicially abrogated by the vast majority of jurisdictions which had recently considered the issue.”118 Developments in other jurisdictions had foreshadowed Tennessee’s own change. Moreover, the court pointed out that the rule had not been applied recently within the state.119 In a sense, the Tennessee court changed the law and expanded the zone of criminal liability. When Rogers committed his offense, the law of causation in Tennessee arguably precluded prosecutions for murder if the victim did not die within a year and a day.120 Indeed, the Tennessee highest court, unlike the Florida Supreme Court in Gore v. Harris,121 admitted that it was altering the preexisting law.122 No Tennessee court had ever called the year-and-a-day rule into question. After Bush v. Gore, therefore, one might imagine that the Justices would easily conclude that retroactive application of the Court’s decision violates Bouie. The Supreme Court, however, hardly blinked. Justice O’Connor’s opinion for the Court reinterpreted Bouie to wrest it from its roots in ex post facto doctrine.123 To apply the same strictures to the courts as to the legislature “would evince too little regard for the important institutional and contextual differences between legislating and common law decisionmaking . . . .”124 She continued that “[s]trict application of ex post facto principles in that context would unduly impair the incremental and reasoned development of precedent that is the foundation of the common law system.”125 Only state law interpretations that are “unexpected and indefensible by reference to the law which had been expressed prior to the conduct at issue” should be 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. Id. at 401. Id. Id. at 402. Id. at 397. Id. at 402. Id. at 395. 772 So. 2d 1243 (Fla. 2000), rev’d sub nom. Bush v. Gore., 531 U.S. 98 (2000). State v. Rogers, 992 S.W.2d at 400-01. Rogers v. Tennessee, 121 S. Ct. 1693 (2001). Id. at 1695. Id. at 1700. 514 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 overturned.126 Under that standard, the Court then held, as had the Tennessee Supreme Court, that the year-and-a-day rule “was not unexpected and indefensible.”127 The Court noted that the rule widely had been viewed as “an outdated relic,” that it had been abolished in “the vast majority of jurisdictions recently to have addressed the issue,” and that the rule had not been applied recently, if at all, in Tennessee.128 In dissent, Justice Scalia could only wonder what had happened to Bouie. Justice Scalia persuasively asserted that Buoie had not focused on “the unexpected and indefensible” language relied on by the majority, and he concluded that the majority opinion could not be squared with Bouie: “Today’s opinion produces . . . a curious Constitution that only a judge could love. One in which (by virtue of the Ex Post Facto Clause) the elected representatives of all the people cannot retroactively make murder what was not murder when the act was committed, but in which unelected judges can do precisely that.”129 Rogers, however, can be reconciled with developments post-Bouie. Few Bouie claims over the past twenty-five years have been successful.130 Courts routinely apply new interpretations of statutory language to criminal defendants appearing before them, even when those interpretations are not dictated by the plain language of the statutory provision. Bouie principles seldom prevent courts from arriving at novel interpretations of statutory terms or from changing prior interpretations of the statutory language. The Bouie cases stand in stark contrast to the Rehnquist concurrence’s analysis. This Part next examines the doctrinal formulations that courts have used to block Bouie claims. As in Rogers, courts have used foreseeability as the key to the Bouie inquiry, upholding judicial alterations in statutory interpretation as long as the changes were foreseeable. When courts make law, retroactive application arguably can be as unjust as in the legislative context,131 but when courts engage in the evolutionary interpretive process, their decisions are foreseeable and thus should be applied retroactively without concern for the individual rights at stake. 126. Id. (citation omitted). 127. Id. at 1701. 128. Id. 129. Id. at 1703 (Scalia, J., dissenting). 130. I have adapted the following section from Harold J. Krent, “Should Bouie Be Buoyed?” Judicial Retroactive Lawmaking and the Ex Post Facto Clause, 3 ROGER WILLIAMS U. L. REV. 35, 57-77 (1997). 131. For instance, U.S. Constitution Article I, Section 10, Clause 1, the Ex Post Facto Clause, would plainly prevent a state legislature from retroactively rescinding the yearand-a-day rule. 2001] JUDGING JUDGING 515 Courts, however, have construed the objective foreseeability requirement quite laxly. They have considered a wide range of factors in determining that a judicial change was foreseeable, stressing the ambiguity of a statute or regulation,132 precedents from other jurisdictions,133 prior administrative interpretation of the statute,134 and whether the Supreme Court had granted certiorari in a similar case.135 No one indicator is dispositive, leaving wide room for the courts to deny a Bouie challenge. Courts, in other words, have acted as if judges almost never make law. For instance, in State v. Mummey,136 the defendant assaulted a man outside a bar, in part by bludgeoning him with tennis shoes. Under the pertinent statute distinguishing felony from misdemeanor assault, the defendant could only be convicted if he used a weapon, which was defined as an instrument “readily capable of being used to produce death or serious bodily injury.”137 The court held that tennis shoes so qualified, and that the defendant should have anticipated such a construction of the statute.138 The conclusion of foreseeability is perplexing—tennis shoes in the public eye are not weapons, let alone “readily capable of being used to produce death or serious bodily injury.” Moreover, if tennis shoes and other everyday articles of clothing can be considered weapons, then the difference between Montana’s punishment scheme for misdemeanor assault (without the weapon) and felonious assault (with a weapon) is difficult to perceive. Mummey is far from unique. As the Ninth Circuit earlier opined, as long as the state court “construed the sentencing scheme in accordance with the principles of statutory construction and its conclusion is certainly not ‘unexpected’ or ‘indefensible’ no Bouie claim should be successful.”139 132. See McSherry v. Block, 880 F.2d 1049 (9th Cir. 1989). 133. See Hagan v. Caspari, 50 F.3d 542 (8th Cir. 1995); State v. Mummey, 871 P.2d 868 (Mont. 1994). 134. See, e.g., Lustgarden v. Gunter, 966 F.2d 552 (10th Cir. 1992). 135. See United States v. Russotti, 780 F. Supp. 128, 134 (S.D.N.Y. 1991) (holding that the change in the interpretation of the criminal statute was foreseeable because the Supreme Court had granted certiorari in a case raising a similar interpretive question before the defendant had committed the crime). 136. 871 P.2d 868 (1994). 137. MONT. CODE ANN. § 45-2-101(76) (1994). 138. Mummey, 871 P.2d at 871. 139. Aponte v. Gomez, 993 F.2d 705, 708 (9th Cir. 1993). Consider as well the Bouie challenge raised in United States v. Lanier, 520 U.S. 259 (1997). Even though Lanier addressed a challenge to judicially imposed change in federal law, the analysis is representative. The same need exists to determine whether judges have changed the law. Prosecutors alleged that the defendant state court judge assaulted several women in judicial chambers. They charged him with violating a civil rights statute, 18 U.S.C. § 242 (1994), on the ground that he had deprived the victims of “rights and privileges which are secured and protected by the Constitution and the laws of the United States, namely the right not to be deprived of liberty without due process of law, including the right to be free from wilful 516 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 Moreover, courts have held that judges make no law even when overruling controlling administrative interpretations. Consider the Eleventh Circuit’s decision in Metheny v. Hammonds.140 There, four Georgia state prisoners challenged retroactive application of a rule by the Georgia Board of Pardons and Paroles that prisoners convicted of a fourth felony were ineligible for parole. The pertinent statute on its face was clear, providing that, upon conviction of the fourth felony, the individuals “shall not be eligible for parole until the maximum sentence has been served.”141 Nevertheless, two Georgia Attorneys General had issued advisory opinions that the statute unconstitutionally infringed on the Board’s power under the Constitution,142 and thus the Board had permitted parole for four-time offenders. In 1994, however, the Georgia Supreme Court in Freeman v. State143 signaled that the legislature could ban parole for a particular category of offenders. In response to Freeman, a new Attorney General decision directed the Board to stop granting parole for four-time offenders,144 and the Board accordingly changed its policy. Those convicted of a fourth offense prior to Freeman, when the Board had been granting parole, argued that Freeman’s change could not be applied retroactively. The Metheny court, however, responded: “[T]hat lower state courts had struck down a few other attempts to limit the Board’s power or that the Attorney General had issued an opinion stating that the statute was unconstitutional is not decisexual assault.” Id. at 262 (quoting trial court indictment). After conviction, the defendant appealed, arguing that the federal statute did not criminalize sexual assaults by state officials, and that, if it did, then retroactive application of that novel interpretation violated his due process rights. United States v. Lanier, 73 F.3d 1380, 1394 (6th Cir. 1996). The Sixth Circuit Court of Appeals agreed with him, setting aside the conviction, by holding that “[t]he indictment in this case for a previously unknown, undeclared and undefined constitutional crime cannot be allowed to stand.” Id. In reversing, the Supreme Court did not take issue with defendant’s assertion that application of the civil rights statute to sexual assault was entirely novel. Yet the Court held that novelty was not the key; rather, it stated that the “touchstone is whether the statute, either standing alone or as construed, made it reasonably clear at the relevant time that the defendant’s conduct was criminal.” 520 U.S. at 267. The Court counseled that, in considering whether the conduct’s criminality was “reasonably clear,” the Court itself need not have considered a similar claim, nor need it identify any lower court precedents involving substantially similar facts. Indeed, the Court clarified that criminal liability can attach merely if, “in light of pre-existing law, the unlawfulness [under the Constitution is] apparent.” Id. at 271-72 (quoting Anderson v. Creighton, 483 U.S. 635, 640 (1987)). The Court accordingly remanded the case to the court of appeals for application of the new standard. The court of appeals never reached the issue on remand because Lanier failed to turn himself in to federal authorities. Accordingly, the appeal was dismissed with prejudice. United States v. Lanier, No. 99-5983, 2000 U.S. App. LEXIS 28047 (6th Cir. Nov. 6, 2000). 140. 216 F.3d 1307 (11th Cir. 2000). 141. GA. CODE. ANN. § 17-10-7(c) (2000). 142. Metheny, 216 F.3d at 1308. 143. 440 S.E.2d 181 (Ga. 1994). 144. Metheny, 216 F.3d at 1309. 2001] JUDGING JUDGING 517 sive.”145 The court continued that, because the Georgia Supreme Court had yet to decide the issue, its subsequent decision in Freeman was not “unforeseeable.”146 Indeed, the Eleventh Circuit stressed “[t]hat the erroneous interpretations were later declared invalid and corrected by the state supreme court does not entitle Plaintiffs to the benefit of those mistaken interpretations.”147 It seems to blink reality to suggest that an inmate is not entitled to rely upon the official, articulated position of the state parole authority with respect to the construction of a statute it is charged with administering, particularly when two Attorney General opinions supported that position.148 The argument against foreseeable change is likely greatest when courts explicitly overrule judicial as opposed to administrative precedent. In that context, courts act more overtly like legislatures in changing the law directly. Courts, however, have held that reversal of prior case law does not necessarily make the decision unforeseeable. For example, in Dale v. Haeberlin,149 the Sixth Circuit rejected a due process challenge to the Kentucky Supreme Court’s decision in Dale v. Commonwealth,150 which upheld the prosecution’s use of the same felony both to prove underlying guilt (felon in possession of a weapon charge) and then to enhance his sentence for armed robbery as a persistent felony offender. The Sixth Circuit acknowledged that the Kentucky Supreme Court’s Dale opinion overruled a prior Kentucky Supreme Court decision but concluded that the offender had fair warning that his sentence for robbery could have been twenty- 145. Id. at 1312-13. 146. Id. at 1313. 147. Id. at 1314. 148. The Tenth Circuit’s decision in Lustgarden v. Gunter, 966 F.2d 552 (10th Cir. 1992), is similar. The court rejected a prisoner’s due process challenge to the Colorado Supreme Court’s construction of a parole statute that limited parole opportunities for certain sexual offenders. The statute targeted “any person sentenced for conviction of a sex offense, as defined in section 16-13-202(5) . . . [and] any person sentenced as a habitual criminal.” Id. at 554 (citing COLO. REV. STAT. § 17-2-201 (5)(a) (1990)). The Colorado Parole Board had construed that statute to apply to habitual offenders and only those sexual offenders sentenced under the Sex Offenders Act, of which section 16-13-202(5) was a part. The state supreme court later disagreed, however, reading the statute to cover anyone convicted of any charge within the broader category of sexual offenses. Thiret v. Kautzky, 792 P.2d 801, 806 (Colo. 1990). The Tenth Circuit in Lustgarden conceded an ambiguity in the statute, namely whether the reference to section 202(5) was intended to limit the impact to those convicted under the Sexual Offenders Act or rather, as the state supreme court in Thiret had determined, just to furnish a definition of “sex offense.” The federal court held, however, that the state supreme court’s change was not unforeseeable, given that it accorded with the most logical construction of the statute. Lustgarden, 966 F.2d at 553. 149. 878 F.2d 930 (6th Cir. 1989). 150. 715 S.W.2d 227 (Ky. 1986). 518 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 five years, even if the same prior felony could not have been used at the time to establish one charge and enhance the other.151 Whether Dale could have anticipated a twenty-five year sentence or not is beside the point. He arguably should not have anticipated the state supreme court’s abrupt change as to whether the same prior felony could be used both at the liability and enhancement stages.152 And, he certainly was likely to receive a lighter sentence without the enhancement. In United States v. Newman153 a defendant who pleaded guilty to bank robbery charges requested sentencing credit for time spent on pretrial release at a residential drug treatment program.154 Under Ninth Circuit precedent, Newman argued, credit was available both at the time Newman committed his offense and when he made the decision to participate in the drug treatment program.155 Shortly after Newman committed his offense, however, the United States Supreme Court in Reno v. Koray156 held that credit should not be available under the pertinent federal sentencing statute.157 The par151. Dale, 878 F.2d at 935. In another case, Commonwealth v. Santiago, 681 N.E.2d 1205 (Mass. 1997), the Supreme Judicial Court of Massachusetts held that Santiago could be tried for first degree murder on the theory that “where the defendant chooses to engage in a gun battle with another with the intent to kill or do grievous bodily harm and a third party is killed, the defendant may be held liable for the homicide even if it was the defendant’s opponent who fired the fatal shot.” Id. at 1215. Prior to that case, however, the Massachusetts court arguably had held that first degree murder could not be demonstrated in such circumstances. See Commonwealth v. Balliro, 209 N.E.2d 308, 314 (1965) (requiring that defendant be the one who pulled the trigger). The Court was not troubled by the change. 152. More generally, courts have held that litigants should anticipate that courts will overrule prior precedents from lower courts. For instance, in Hagan v. Caspari, 50 F.3d 542 (8th Cir. 1995), the petitioner argued that the Missouri Supreme Court’s rejection of his double jeopardy claim was unforeseeable. Id. at 544. He had been charged both with second degree robbery for forcibly grabbing the keys to a van and then for stealing the van. Id. He argued that the two crimes merged because of Missouri’s single larceny rule, which defined the stealing of several articles of property during the same scheme or course of conduct as a single offense under Missouri Revised Statutes, § 570.050 (1986). Id. at 546. Under similar facts, a Missouri appellate court had found that conviction on both a robbery and theft charge would violate the Double Jeopardy Clause. See State v. Lewis, 633 S.W.2d 110 (Mo. Ct. App. 1982). Therefore, the petitioner in Hagan argued that, in overruling the appellate court’s interpretation of Missouri’s larceny statute, the state supreme court violated Bouie. Hagan, 50 F.3d at 544. The Eighth Circuit disagreed, holding that the change was foreseeable largely because the appellate court’s reasoning was plainly wrong. Id. at 546-47. The larceny statute may have prevented conviction on two separate stealing charges, but not for the distinct robbery and theft charges. Thus, the Double Jeopardy Clause did not bar conviction of the two charges. The court further commented that “[w]e have some doubt whether a state supreme court’s overruling of an intermediate appellate court decision ever can constitute a change in state law for due process purposes.” Id. at 547. 153. 203 F.3d 700 (9th Cir. 2000). 154. Id. at 701. 155. Id. at 702. 156. 515 U.S. 50 (1995). 157. Id. at 65. 2001] JUDGING JUDGING 519 ties agreed that retroactive changes in the length of punishment fall within the protection of the Due Process Clause’s incorporation of ex post facto principles. The Ninth Circuit, however, held that retroactive application of Koray to Newman was permissible, even though it marked a clear change of law within the circuit.158 The court reasoned that “the decision in Koray was reasonably foreseeable given the circuit split on the meaning of [the statute].”159 Cases from other jurisdictions evidently undermined the plaintiff’s entitlement to rely on Ninth Circuit precedent. The Newman court also suggested that a “reversal of a prior incorrect interpretation of a statute . . . [was not] barred by Bouie . . . .”160 The undoing of a precedent, therefore, may be foreseeable.161 Defendants may not be entitled to rely on the construction of a statute adopted by an appellate court because those precedents can always be overturned.162 158. Newman, 203 F.3d at 703. 159. Id. The Ninth Circuit previously had held that defendants may not rely upon any judicial opinion that is subject to rehearing on review or certiorari. See United States v. Ruiz, 935 F.2d 1033, 1038 (9th Cir. 1991) (holding that defendants were not entitled to rely on a prior appellate opinion which was withdrawn by the Ninth Circuit after the plea agreement had been reached); United States v. Kincaid, 898 F.2d 110, 111 (9th Cir. 1990) (holding that defendants could not rely upon the fact that federal sentencing guidelines had been invalidated at the time of sentencing, because the guidelines were later reinstated by the Supreme Court). 160. 203 F.3d at 702. 161. Similarly, in United States v. Rodgers, 466 U.S. 475 (1984), a federal case, the United States Attorney charged the defendant with making a false statement to the FBI and Secret Service in a “matter within the jurisdiction of any department or agency of the United States.” Id. at 476 n.1 (quoting 18 U.S.C. § 1001 (1994)). Defendant allegedly apprised the FBI that his wife was involved in a plot to assassinate the President and that she had been kidnapped. Id. at 476. The defendant moved to dismiss the indictment on the basis of a prior Eighth Circuit decision that construed the language “within the jurisdiction of any department or agency” to refer only to areas in which the agency had the power to make final or binding determinations, such as for monetary awards. Id. at 477-78 (citing Friedman v. United States, 374 F.2d 363 (8th Cir. 1967)). On that authority, the district court and Eighth Circuit in Rodgers dismissed the indictment because the defendant’s comments merely triggered a criminal investigation. United States v. Rodgers, 706 F.2d 854, 855 (8th Cir. 1983). The Supreme Court reversed, holding that the language “within the jurisdiction of any department or agency” was to be read broadly, including any matters pertaining to criminal investigations. Rodgers, 466 U.S. at 479. The defendant argued, however, that the more expansive construction, even if adopted, could not be applied to him retroactively given his reliance on the prior Eighth Circuit precedents. Id. at 484. The Supreme Court disagreed due to “the existence of conflicting cases from other Courts of Appeals [which] made review of that issue by this Court . . . reasonably foreseeable.” Id. For a discussion of Rodgers’ significance, see Trevor W. Morrison, Fair Warning and the Retroactive Judicial Expansion of Federal Criminal Statutes, 74 S. CAL. L. REV. 455, 457-58 (2001). 162. See also Green v. Catoe, 220 F.3d 220 (4th Cir. 2000), abrogated on other grounds by Bell v. Jarvis, 236 F.3d 149, 160 (4th Cir. 2000) (concluding that “if the change of law was reasonably foreseeable, based on indications in prior case law, then the defendant had sufficient notice in the due process sense”). 520 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 The foreseeability inquiry, therefore, has been manipulated substantially. Courts have rationalized that litigants should have expected the new or changed judicial interpretation, even when the judicial interpretation overrules administrative or judicial precedents. Courts have sufficient leeway to permit retroactive application of new statutory interpretations in the criminal context by dint of the foreseeability analysis. Judicial opinions ex post can be justified on the basis of the general evolution of the law.163 Arguably, the foreseeability analysis may further rule of law values in limiting judicial power. The foreseeability inquiry may constitute a sorting mechanism to separate those instances of judicial interpretation that represent conventional common lawmaking or statutory construction from those that are fueled by animus toward the offender. According to traditional analysis, the Ex Post Facto Clause protects against arbitrary governance as well as safeguarding the offender’s reliance interest. Judges may interpret common law doctrines, statutes, or regulations in order to increase the severity of the punishment for one particular odious offender. Perhaps no direct way to police judicial bias exists, but if a construction is foreseeable, then it makes it less likely that animus was a determining factor. On occasion, therefore, the federal courts may be convinced that an impermissible risk of state court bias or arbitrariness exists, as in Bouie itself. In Douglas v. Buder,164 the Supreme Court intervened to stop the apparent arbitrary determination of a Missouri state court judge to reincarcerate an individual who was on probation. The probationer had not immediately reported a traffic citation to his probation officer, which would be required if the citation were the equivalent of an arrest.165 The Supreme Court signaled its disbelief that the Missouri court would ever have determined that a traffic citation constituted an arrest but for the court’s evident animus against the probationer.166 Irrespective of whether the trial court’s interpretation was foreseeable, Bouie allows reviewing courts to rein in lower courts when they appear influenced by vindictive or arbitrary aims. 163. Professor Fisch’s theory of equilibrium is similar. See Jill E. Fisch, Retroactivity and Legal Change: An Equilibrium Approach, 110 HARV. L. REV. 1055 (1997). She argues that most common lawmaking—as opposed to legislation—should be considered foreseeable because the common law reflects an unstable equilibrium that can be moved with the slightest tug in any direction. Id. at 1108-09. Thus, litigants cannot justifiably rely on any particular common law position. The only exception she notes is when the Supreme Court overturns its own precedents. Id. at 1107-08. In that context, adjudication more closely resembles legislation and may disrupt a stable equilibrium upon which litigants are entitled to rely. Id. at 1108. 164. 412 U.S. 430 (1973). 165. Id. at 431. 166. See id. at 431-32. Indeed, the probation officer and prosecutor in the case had argued that the probationer should not be reincarcerated. Id. at 431. 2001] JUDGING JUDGING 521 Thus, on balance, Bouie is best understood as a means of limiting judicial vindictiveness rather than protecting offenders’ reliance interests or ensuring repose. The foreseeability analysis acts as a crude filter to separate judicial reinterpretations animated by retributive impulses from the more familiar common law evolutionary process. Courts seem to determine whether barring retroactive application of a new interpretation is needed to restrain judicial overreaching. The Bouie inquiry is admittedly ad hoc, but serves as a last-line check against prosecutorial and judicial overreaching. B. Contracts Clause After Gelpcke v. City of Dubuque,167 courts enforced the judicial impairments doctrine for the next generation. Judicial decisions that altered prior precedent in a way that impaired contract rights were suspect.168 But by the late nineteenth century, federal courts began retreating from the judicial impairments doctrine. For reasons of politics and judicial ideology, they became far more chary of pinning federal rights on a conclusion that state judicial opinions had “changed” law.169 Federal courts became increasingly loathe to second-guess a state court’s decision as to state law matters.170 For instance, some federal courts strained to find no change on the ground that the prior state court decisions were not controlling. In Keokuk & W.R. Co. v. County Court,171 the question for resolution, although quite convoluted, concerned whether a railroad could be exempt from taxation. Missouri had granted exemptions to certain railroads but then forbade new exemptions several years later. At the time that the plaintiff railroad had formed through consolidation with another, a prior Missouri state court decision172 assumed that companies did not lose their tax exemption through merger or other consolidation.173 That court rejected the argument “that a change in the political form of civil society has the magical effect of dissolving 167. 68 U.S. (1 Wall.) 175 (1863). 168. Federal courts principally deployed the doctrine when lower courts departed from prior precedent as opposed to when they merely rested a decision on an unforeseeable interpretation of a statute. But see Pine Grove Township v. Talcott, 86 U.S. (19 Wall.) 666, 678 (1874) (finding an impairment of contract in absence of prior state court decision—“In cases properly brought before us that end can be accomplished unwarrantably no more by judicial decisions than by legislation”). 169. Thompson, supra note 75, at 1426-38. 170. For a time, federal courts also disregarded state courts’ older precedents if convinced that they were plainly incorrect. See, e.g., Bd. of Comm’rs v. Coler, 113 F. 705, 708 (4th Cir. 1902), aff’d, 190 U.S. 437 (1903); Jones v. Great S. Fireproof Hotel Co., 86 F. 370, 372 (6th Cir. 1898), aff’d, 177 U.S. 449 (1900). 171. 41 F. 305, 310 (E.D. Mo. 1890), aff’d, 152 U.S. 301 (1894). The Supreme Court affirmed on slightly different grounds. 172. See Scotland County v. Mo., Iowa & Neb. Ry. Co., 65 Mo. 123 (1877). 173. Id. at 134-35. 522 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 its moral obligations or impairing contracts previously vested”174 and accordingly upheld the tax exemption for the railroad. In Keokuk, the contract for consolidation of the two railroads included a clause that any immunities previously enjoyed would continue in the new company. The Missouri high court later held, however, that the consolidated railroad was in effect a new corporation. Accordingly, the railroad could not rely on the prior decision. The federal courts agreed, reasoning that: “To say that, because a state court has once decided that a certain corporation is entitled to exemption from taxation, the decision must thereafter be followed, although erroneous, would involve consequences of such a serious [financial] nature that any court ought to hesitate . . . .”175 The second Missouri decision merely clarified the reach of the legislative tax exemption. Accordingly, the federal court rejected plaintiff’s argument that the second state court decision had impaired contractual obligations (and violated a property right). Similarly, federal courts strained to distinguish the holdings in prior state court decisions. The Supreme Court’s decision in Sauer v. New York176 provides a good example. There, a homeowner challenged the city’s construction of an elevated viaduct near his home on several grounds. New York’s courts rejected the challenge, reasoning that a homeowner did not, under New York law, enjoy the right to unimpeded access to his home. In turn, Sauer sued in federal court, arguing that the New York decision changed the preexisting law in New York. He cited a case decided four years before he acquired title to his land—Story v. Elevated Railroad177—which assured landowners abutting public roads to easements of access, light, and air. The United States Supreme Court, however, was unmoved. It distinguished the prior case on the grounds that the city there had blocked access to benefit private corporations such as railroads: “The structure in these cases were held to violate the land owners’ rights, not only because they were elevated and thereby obstructed access, light and air, but also because they were designed for the exclusive and permanent use of private corporations.”178 Accordingly, it held that plaintiff “has not shown that in his case the state court has changed, to his injury, the interpretation of his contract with the city . . . .”179 174. 175. 176. 177. 178. 179. Harlem Id. at 135. Keokuk, 41 F. at 310. 206 U.S. 536 (1907). 90 N.Y. 122 (1882). Sauer, 206 U.S. at 552. Id. at 555. As Justice Holmes commented in dissent in Muhlker v. New York & Railroad, 197 U.S. 544 (1905), federal courts should not reverse state courts if the 2001] JUDGING JUDGING 523 The Justices in dissent accused the majority of reinterpreting precedents to arrive at the conclusion of no changed law.180 The prior decision had mentioned, but not stressed, the fact that a private corporation was involved. For instance, the court in Story had asked what “is the extent of this easement? What rights or privilege are secured thereby? Generally, it may be said, it is to have the street kept open, so that from it access may be had to the lot and light and air furnished across the open way.”181 And, the court stressed that the “public purpose of a street requires of the soil the surface only.”182 The dissenting judges would have held that the new judicial decision breached the state’s obligation to its contract.183 Like the Bouie doctrine today, however, the judicial impairments doctrine presented a flexible means for setting aside state court decisions as in Gelpcke that seemed arbitrary or targeted at outsiders. For instance, in McCullough v. Virginia,184 bondholders challenged a Virginia Supreme Court of Appeals decision that bondholders could not require the state to accept coupons as payment of taxes because the 1871 bond issuance was illegal. The U.S. Supreme Court reversed, stressing that the Virginia high court on several prior occasions had upheld the bonds’ validity: These . . . bonds, amounting to many millions of dollars, have passed into the markets of the world, and have so passed accredited, not merely by the action of the General Assembly of the State of Virginia, but by the repeated decisions of her highest court . . . for substantially a quarter of a century . . . .185 Indeed, the Court stressed that in “reversing its prior rulings” the Virginia court “put[] at naught the repeated decisions of this court as well as its own.”186 As in Bouie, therefore, the judicial impairments doctrine served as a last-ditch means of restraining state court prejudice. Within a generation, however, the Court stopped enforcing the judicial impairments doctrine. In part, the change stemmed from greater acceptance of judicial realism. As Justice Brandeis described recent state courts’ efforts to distinguish prior cases were “not wanting in good sense.” Id. at 574-75. 180. Sauer, 206 U.S. at 556-60. 181. Story, 90 N.Y. at 146. 182. Id. at 161. 183. In addition, the U.S. Supreme Court refused to take some cases during that period even if some change existed on the ground that “mere reversal by a state court of its previous decision [did] not . . . violate any clause of the federal constitution.” Tidal Oil Co. v. Flanagan, 263 U.S. 444, 454-55 (1924); see also Brinkeroff-Faris Trust & Sav. Co. v. Hill, 281 U.S. 673, 681 n.8 (1930). 184. 172 U.S. 102 (1898). 185. Id. at 108. 186. Id. at 122-23. 524 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 in Brinkerhoff-Faris Trust & Savings Co. v. Hill,187 “[t]he process of trial and error, of change of decision in order to conform with changing ideas and conditions,”188 was a fundamental part of all judging.189 He continued that “[s]tate courts, like this Court, may ordinarily overrule their own decisions without offending constitutional guaranties, even though parties may have acted to their prejudice on the faith of the earlier decisions.”190 With Burns Mortgage Co. v. Fried191 in 1934, the Court implicitly overruled the Gelpcke line of cases by holding that, in diversity cases, the Court had no choice but to accept state court interpretations.192 The Supreme Court has not revisited the Gelpcke doctrine since the beginning of the twentieth century— judicial changes do not violate the Contracts Clause. No one explanation for the decline in enforcement of the judicial impairments and Bouie doctrines is likely dispositive. And, the judges’ motivations for the decline may have changed over time. Nonetheless, striking similarities for the decline exist. First, as a matter of federalism, it became increasingly difficult to justify intervention in state law matters. The Supreme Court rarely second-guessed state law interpretation in the hundreds of cases decided each year. Although some oversight might be necessary to protect the federal constitutional rights at stake, each reversal of a state court decision on the ground that the court had changed state law appeared unseemly. Moreover, over time, there may have become less reason to mistrust state court judges. For instance, after the mid-nineteenth century municipal bond crisis, state courts were not as active in attempting to protect local taxpayers at the expense of out-of-state investors. Similarly, after the initial years of the civil rights movement, state judges were not as hostile to the rights of those in the movement, as they had been in Bouie itself. The United States Supreme Court began to trust state courts to enforce federal rights. Second, distinguishing state cases that changed prior law from those that refined it or just applied it to a new setting became increasingly problematic. As judicial realism became more accepted, the Court was less secure in reaching the conclusion that a state court decision had in fact changed state law. Given the complexities 187. 281 U.S. 673 (1930). 188. Id. at 681 n.8; see also Great N. Ry. v. Sunburst Oil & Ref. Co., 287 U.S. 358 (1932). 189. The Supreme Court consistently had declined to find that judicial changes in the law have effected compensable takings. See Thompson, supra note 15. 190. Hill, 281 U.S. at 681 n.8. 191. 292 U.S. 487 (1934) (relying on section 34 of the Judiciary Act of 1789). 192. Id. at 493-94. Despite Erie v. Tompkins Railroad Co., 304 U.S. 64 (1938), the Court still could have exercised the power to review claims directly under the Contracts Clause, as it had on occasion previously. See supra Part II.B. 2001] JUDGING JUDGING 525 of judicial decisionmaking and the difficulty of construing statutes, demarcating a bright line between law application and lawmaking is daunting. Third, the Court may have believed that notice was not as important as previously thought. Under Bouie, cases such as Mummey demonstrate the fictive nature of reliance. Criminal defendants do not scan judicial opinions prior to committing an offense. Although commercial entities may be more familiar with the peculiarities of judicial interpretations prior to entering into a deal, counsel themselves may well predict that certain doctrines or precedents are unstable. Counsel, for instance, could have predicted that the tax exemption in Keokuk was not secure. In other words, judges may have become less committed to maintaining the status quo. Fourth, and relatedly, the Court may have become more reluctant to privilege older as opposed to newer judicial interpretations. The second interpretation may be “better,” and therefore justice may be served by upholding the second interpretation. For instance, in Mummey, precluding courts from applying the new definition of “weapon” would have resulted in a lesser penalty for a perceived wrongdoer. In Metheny, retroactive application of the state court decision prevented parole for a category of offenders that the legislature deemed likely to be recidivist. Similarly, in cases such as New Orleans Waterworks Co. v. Louisiana Sugar Refining Co.,193 the judicial impairment stemmed from judicial efforts to peel back legislatively granted monopolies, a goal that many federal jurists likely favored.194 In Sauer, the Court may have wanted to pave the way for internal improvements such as the pedestrian overpass considered there. Preventing state courts from changing law may freeze law which is normatively unattractive. Finally, the U.S. Supreme Court may have become more concerned about the institutional impact of encouraging claims alleging federal constitutional violations arising from changed state law. With both Bouie and Contracts Clause claims, the Court perceived that the doctrines triggered a substantial increase in litigation. The Bouie doctrine presented a convenient hook upon which state defendants could hang their arguments, and Gelpcke threatened to convert many disagreements over state contract rights into a federal constitutional 193. 125 U.S. 18 (1888) (upholding legislature’s decision to grant permission to refining company to lay water pipes despite prior legislative monopoly). 194. See, e.g., Lehigh Water Co. v. Easton, 121 U.S. 388 (1887) (holding that legislative monopoly did not preclude competition from water works operated by municipality). 526 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 claim.195 Retrenchment may be seen as a natural response designed to preserve institutional resources. In short, the U.S. Supreme Court has backed off from both the Bouie and judicial impairment doctrines for similar reasons. Concerns of federalism, preserving institutional resources, the difficulty of determining when as a principled matter intervention is warranted, and the unattractiveness of compelling state courts to adhere to “law” that may be “wrong” or “backward” all fueled the retreat. C. Enforcement of Article II, Section 1 The Bouie and Contract Clause experiences shed light on the Rehnquist concurrence’s analysis of Florida law. No one doubts that Florida’s legislative scheme was open to differing interpretations. The only question is how far the Florida court strayed from the text. Few benchmarks aid that inquiry. Nonetheless, courts in the Bouie and judicial impairments context have asked four general questions. First, courts have inquired whether the judicial interpretation under scrutiny reversed settled precedents. Indeed, that was the principal question asked in the judicial impairments context. The Court used that factor to stress that “new” law had been fashioned in McCullough. However, this factor by itself is not dispositive as suggested by the Dale and Newman cases. Individuals (or counsel) at times should be able to anticipate judicial change. In any event, the Florida court’s decision in Gore in no way overturned any pertinent precedent. Indeed, the 1999 revision to the election code had never been reviewed judicially prior to the presidential election controversy. Second, to a lesser extent, federal courts have focused on whether state courts have overruled administrative precedents. Legislatures often delegate subsidiary lawmaking authority to administrative entities, and courts will defer to reasonable agency interpretations of law that emerge from the lawmaking or enforcement context.196 Brushing aside administrative interpretations may suggest, therefore, that courts are fashioning new law. The Florida court decision 195. On the increased litigation in contracts cases, see CHARLES FAIRMAN, MR. JUSTICE MILLER AND THE SUPREME COURT 1862-1890, at 401-05 (1939); FELIX FRANKFURTER & JAMES M. LANDIS, THE BUSINESS OF THE SUPREME COURT 60-102 (1928). 196. Under the federal doctrine of Chevron U.S.A., Inc. v. Natural Resources Defense Council, 467 U.S. 837 (1984), the Supreme Court has directed that courts should defer to reasonable constructions of statutory language by agencies. Florida courts may afford less deference than the U.S. Supreme Court to agency construction of statutory language, but it defers nonetheless. See Donato v. AT&T, 767 So. 2d 1146 (Fla. 2000); Tampa Elec. Co. v. Garcia, 767 So. 2d 428 (Fla. 2000); Krivanek v. Take Back Tampa Political Comm., 625 So. 2d 840, 844 (Fla. 1993); cf. James Rossi, HAMSTRINGING STATE AGENCY AUTHORITY TO PROMULGATE RULES: A QUESTIONABLE WAY TO IMPROVE ENVIRONMENTAL REGULATION, 29 ENVTL. L. REP. 10735 (Dec. 1, 1999). 2001] JUDGING JUDGING 527 set aside the Florida Secretary of State’s opinion on defining “lawful” votes,197 and the legislature had delegated the Secretary authority to oversee elections.198 That clash, however, does not significantly support the Rehnquist concurrence’s determination. As an initial matter, the Secretary of State issued the opinion in the midst of the election crisis, not before. The administrative precedent, therefore was in no way settled. In addition, courts retain the right to reject any administrative construction of the law that is deemed unreasonable. As the Metheny example illustrates, courts have not hesitated to find that judicial reversals of administrative precedent are foreseeable. Even if the court erred in overturning the precedent, that error is not by itself tantamount to a conclusion that the court changed the manner that the legislature provided for selecting presidential electors. Courts may pay insufficient heed to administrative rulings without changing the law. Third, courts have investigated whether developments in other jurisdictions made the state court decision under scrutiny foreseeable. Although an interpretation by another state’s highest court obviously is not controlling, reviewing courts have relied on such developments to suggest that the state court decision under consideration is not unexpected. Indeed, as in Rogers and Mummey, one might defend the Florida court’s analysis as to what constitutes a lawful vote by reference to the position adopted by other states.199 State courts had held that the intent of the voter standard governs. From the perspective of the developing law in other jurisdictions, therefore, the Florida Supreme Court decision seems foreseeable.200 Fourth, the question of whether the Florida court’s interpretation derived from the legislature’s text is more open-ended. Judgments can differ. Certainly, the concurrence’s sally that the Florida court rid the certification of independent meaning finds its mark. On the other hand, no specific language in the statute mandates deference. Moreover, the statute is not clear as to whether the intent of the voter standard was to govern. Overall, as compared to cases such as Mummey and Lanier, the Florida court’s interpretation rests more 197. Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d 1273, 1282-83 (Fla. 2000), vacated sub nom. Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000). 198. See FLA. STAT. § 97.012(1) (2000) (designating Secretary of State as “chief election officer“ with the responsibility to “[o]btain and maintain uniformity in the application, operation, and interpretation of election laws”). 199. See the discussion in Bush v. Gore, 531 U.S. 98, 150-54 (2000) (Breyer, J., dissenting); see also Brief for Respondents at 36, Bush v. Gore, 531 U.S. 98 (2000) (No. 00-949). 200. Professor Pildes argues that, in the election context, federal courts should stay their hands unless plaintiffs can show, in addition, detrimental reliance. Pildes, supra note 13, at 722-23. 528 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 comfortably on the legislative language.201 Using the benchmarks set out in the Bouie and Contracts Clause context highlights the weakness of the concurrence’s analysis. In retrospect, the concurrence’s chief error may have stemmed from its assumption that the Article II, Section 1 directive demands stringent enforcement. The question should not be whether the Florida court “distorted the [statutory language] beyond what a fair reading required.” Because the “beyond . . . a fair reading” standard is so open-ended, it invites second-guessing by reviewing courts. The standard converges with one of clear legal error. The Supreme Court’s own interpretation of the Equal Protection Clause, in many minds, might be characterized as distorting the Equal Protection Clause “beyond what a fair reading required.” The Court never applied Bouie or the judicial impairments doctrine to state court decisions that were merely wrong, or even clearly wrong. Rather, intervention should be reserved only for the most adventurous or unanticipated judicial changes in the law.202 Full enforcement of the Article II, Section 1 provision should be rejected. Despite the importance of federal oversight, the concerns prompting a more hands-off stance in the Bouie and Contracts Clause contexts fully apply. Indeed, the reasons against federal judicial second-guessing were even stronger in Bush v. Gore than in the contexts described above. As an initial matter, the institutional arguments are quite powerful. The Rehnquist concurrence would open federal courthouse doors to any claim that state jurists have changed the manner in which electors can be chosen. Any judicial interpretation of appropriate districts set out by the legislature, of appropriate voting methodology, of tabulation procedures, or of a state’s contest process could form the basis of a federal challenge. Particularly when a presidential election may hang in the balance, inviting such challenges risks embroiling the Court in political disputes. Over time, courts will likely become more reticent about inviting challenges to judicial interpretations of selection decisions given the high stakes. Although the Court—and particularly the concurring Justices—seemed eager to take on the challenge of determining whether the Florida Justices changed law in Bush v. Gore, it is doubtful whether subsequent courts will be so willing. 201. See also Michael J. Klarman, Bush v. Gore Through the Lens of Constitutional History, 89 CAL. L. REV. 1721 (2001) (lambasting concurrence’s analysis of Florida law). 202. Even under the concurrence’s standard of “beyond what a fair reading required,” the dissenting Justices were on firm terrain in finding that the Florida Supreme Court’s interpretations were grounded in the statutory language. 2001] JUDGING JUDGING 529 Aside from the institutional concerns, the inquiry of whether a judicial construction of a state’s legislative scheme for selecting electors changed the law is insufficiently bounded. On the one hand, almost every judicial decision of interest either elaborates on a previously uncertain point or clarifies a prior conflict. The common law method presupposes at least interstitial change. Evolution may at times appear more dramatic than at others. In a sense, therefore, many judicial decisions “change” the law, whether through interpretation of statutes or modifications of prior judicial doctrines. On the other hand, judges can always rationalize various opinions as foreseeable. Some decisions are, of course, more anticipated than others, but the decisions following Bouie are ample testament to the elasticity of the inquiry. In Bush v. Gore, for instance, the fact that four of the seven Justices on the Florida Supreme Court agreed on a set of key interpretations of the legislative landscape, and that their view was defended as plausible by four U.S. Supreme Court Justices, demonstrates the difficulty of distinguishing interpretation from lawmaking. Reasonable minds may disagree as to what is a reasonable interpretation. There are few objective benchmarks available to determine which decisions “change” the law as opposed to adapting law to respond to new social, economic, or technological realities, or even an unforeseen set of facts. The general foreseeability inquiry in the Bouie and judicial impairments contexts—based on consistency with judicial precedent, administrative precedent, precedent from other jurisdictions, and the statutory text—does not add sufficient concreteness or predictability. In addition, the concurrence’s approach undermines the respect due state judiciaries. Although this Supreme Court has championed federalism by limiting the powers of Congress,203 the concurrence’s position all but heaps scorn on state jurists. Second-guessing state court decisions to determine whether they go beyond “interpreting” law undermines the respect due to the highest court of the states. The decline in enforcement of the Bouie and judicial impairment doctrines in no small measure arose because of increased confidence and respect in state judges. Indeed, review of state court decisionmaking under Article II— from a federalism perspective—is far more jolting than under Bouie 203. For a sampling of recent decisions, see Bd. of Trustees v. Garrett, 531 U.S. 356 (2000) (holding that Congress could not subject states to suit in federal court for violations of the Americans with Disabilities Act); United States v. Morrison, 529 U.S. 598 (2000) (limiting Congress’s power under Commerce Clause); Kimel v. Fla. Bd. of Regents, 528 U.S. 62 (2000) (limiting Congress’s power to impose suit against states under the Eleventh Amendment); Printz v. United States, 521 U.S. 898 (1997) (deriving principle that prevents congressional commandeering of state officials under the Tenth Amendment); see also City of Boerne v. Flores, 521 U.S. 507 (1997) (limiting Congress’s power to fashion appropriate legislation to enforce equal protection requirements in the Fourteenth Amendment). 530 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 or the Contracts Clause. The consequence of a successful challenge in the latter two contexts is that the interpretation cannot be applied retroactively. The state courts, however, remain free to apply interpretations such as the trespassing rule in Bouie or the authority to issue bonds in Gelpcke prospectively. Retroactive invalidation does not tie the state courts’ hands in future cases, once notice has been assured. The Rehnquist concurrence, however, would have precluded application of the Florida Supreme Court’s interpretation of election law in all future cases. Even if the Florida court afforded notice, the concurrence still would have invalidated the state court holding for abridging the legislature’s authority under Article II to select the manner in which electors are chosen. Viewed in that light, review under Article II is not concerned so much with retroactivity but with the allocation of power between state legislature and judiciary. Review by federal courts, therefore, compromises state judicial authority to a greater degree than in the retroactivity cases.204 The threat to federalism accordingly is more acute in the Article II setting than in other contexts. As an analogy, consider the independent and adequate state grounds doctrine. Under Michigan v. Long,205 the Court will not review a state court decision if an independent state ground exists, even if the decision includes an analysis of a federal constitutional guarantee. The state ground may have been crafted in order to insulate the state court decision on federal law from Supreme Court review. Regardless, the U.S. Court will not intervene. Prior to Michigan v. Long, however, the Court scrutinized the state law ground to determine whether it was “adequate” enough to immunize the decision from federal court review.206 Justifications deemed too makeweight or too ill thought-out were not sufficient to insulate the state courts from federal review. The Court rejected that tradition in Michigan v. Long in part because “[t]he process of examining state law is unsatisfactory because it requires us to interpret state laws with which we are generally unfamiliar.”207 In addition, the Court justified its refusal to second-guess state law grounds on “[r]espect for state courts.”208 Under Michigan v. Long, therefore, the Court will permit state court interpretations of state law with which it would be unlikely to concur. For instance, the Florida Supreme 204. I am indebted to Trevor Morrison on this point. 205. 463 U.S. 1032, 1037-41 (1983). 206. See, e.g., Texas v. Grown, 460 U.S. 730, 732 n.1 (1983); Abie State Bank v. Bryan, 282 U.S. 765 (1931). 207. Michigan v. Long, 463 U.S. at 1039. 208. Id. at 1040. 2001] JUDGING JUDGING 531 Court in State v. Lavazzoli209 disagreed with the U.S. Supreme Court’s view of search and seizure doctrine but pegged its decision on a novel interpretation of Florida’s constitutional requirements.210 Under the independent and adequate grounds doctrine, no federal review could follow. The Rehnquist concurrence does not manifest the same measure of respect for state court interpretation of state law. The concurrence dismissed difficult questions of state law with the back of its hand. On questions of the U.S. Supreme Court’s own power, the Rehnquist Court ironically has become nationalistic.211 Finally, the radical “realism” of the Rehnquist concurrence— equating judicial interpretation with lawmaking—if followed, might well have a destabilizing effect on other areas of the law. Viewing judicial interpretation as changed law can open a wide array of constitutional challenges to judicial decisionmaking. For instance, courts might heed the analysis in the concurrence to permit far more Bouie challenges despite Rogers. As a result, the burden of convicting and sentencing clearly guilty offenders would increase. Similarly, pressure would increase to recognize takings challenges to judicial decisions, an area in which courts have so far steered clear.212 Curiously, Justice Scalia, in James B. Beam Distilling Co. v. Georgia,213 recognized the difficulty of characterizing judicial lawmaking as legislation. There, the Court considered whether to apply new constitutional rules retroactively. Concurring in the judgment, Justice Scalia explained why he believed that courts could not, like legislatures, apply new rules prospectively only: the Article III judicial power “must be deemed to be the judicial power as understood by our common-law tradition. That is the power ‘to say what the law is,’ not the power to change it.”214 Justice Scalia immediately continued, however, that “I am not so naive (nor do I think our forebears were) as to be unaware that judges in a real sense ‘make’ law.”215 But, they make law “as judges make it, which is to say as though they were ‘finding’ it—discerning what the law is, rather than decreeing what it 209. 434 So. 2d 321 (Fla. 1983). 210. Id. at 323-24. 211. See Dickerson v. United States, 530 U.S. 426 (2000) (holding that state (and lower federal) courts must abide by Supreme Court’s constitutional common lawmaking in the Miranda decision); Nat’l Private Truck Council, Inc. v. Okla. Tax Comm’n, 515 U.S. 582 (1995) (depriving state courts of power to enjoin state taxing authority sued under federal law). 212. See Thompson, supra note 15, at 1498-1502 (addressing institutional factors that have led some to caution against recognizing judicial takings). 213. 501 U.S. 529 (1991). 214. Id. at 549 (Scalia, J., concurring) (quoting Marbury v. Madison, 5 U.S. (1 Cranch) 137, 177 (1803)). 215. Id. (Scalia, J., concurring). 532 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 is today changed to, or what it will tomorrow be.”216 Maintaining the fiction that judges do not make law may be important to the judicial system as we know it. To the extent that judges are considered lawmakers, pressure builds to impose the same constraints on judges that currently circumscribe legislative authority. In particular, the limits on retroactive decisionmaking embodied in the Ex Post Facto Clause, Bill of Attainder, and Takings Clauses might apply. Such constraints would, for better or worse, straightjacket judicial decisions considerably. At least in contexts such as criminal law and property rights, in which retroactivity is disfavored, much judicial decisionmaking would be suspect. In short, the concurrence plausibly concluded that judges can review claims alleging that state legislatures have been deprived of their constitutional power to determine the manner in which presidential electors are selected. Nonetheless, concerns for preserving institutional capital, avoiding administrative line-drawing difficulties, maintaining tenets of federalism, and fostering stability in other doctrinal areas all suggest that the constitutional provision should not be aggressively enforced. Underenforcing the Article II, Section 1 directive would not be novel. One relatively familiar example may prove helpful. In Garcia v. San Antonio Metropolitan Transit Authority,217 the Court declined to enforce the Tenth Amendment principally for institutional reasons. In rejecting a challenge to application of the Federal Labor Standards Act (FLSA) to a municipal body, the Court reasoned that a judicially crafted test to distinguish traditional from other state functions was “unsound in principle and unworkable in practice.”218 The Court had previously struggled in fashioning a test to determine which state functions to preserve from federal interference. Despite the Court’s decision not to enforce the Tenth Amendment directly, it signaled that the Tenth Amendment norm continued to merit deference. In legislating, Congress still presumably considered the Tenth Amendment norm to the extent that it considers any constitutional 216. Id. (Scalia, J., concurring). 217. 469 U.S. 528 (1985). 218. Id. at 546. The Court justified its refusal to enforce the Amendment on the ground that political process checks existed to check Congress’s regulation of state and local governmental authorities, namely that state and local governments enjoyed ample voice in Congress’s affairs. See Michael W. McConnell, The Supreme Court, 1996 Term—Comment: Institutions and Interpretation: A Critique of City of Boerne v. Flores, 111 HARV. L. REV. 153, 173 (1997) (arguing that the court should defer to congressional interpretation of constitutional text when judicially appropriate standards are wanting). 2001] JUDGING JUDGING 533 question.219 Indeed, Congress amended the FLSA after Garcia to protect local governments from some of the Act’s provisions.220 A constitutional provision can retain vitality even when not actively enforced by the judiciary.221 In a system marked by separated powers and federalism, the Court logically may allow institutional factors to shape constitutional rights.222 Just as in Garcia, an underenforcement tack in reviewing claims under Article II, Section 1 would reflect a strategy of enforcing constitutional rights indirectly to avoid the political or institutional pitfalls of more direct enforcement. The Supreme Court should only conclude that state courts have changed the manner in which electors are selected in the rare case. The U.S. Court should not intervene unless the state court’s construction is both unsupportable by reference to conventional textual analysis and relevant precedents and, in addition, threatens to gut the legislature’s control of the process to select presidential electors. CONCLUSION The concurrence’s opinion in Bush v. Gore is as startling as it is misconceived. The opinion oozes disrespect for the Justices on the Florida Supreme Court and flies in the face of the Court’s frequent paeans to the fundamental role of federalism in our system. Federalism is something evidently that Congress, not the Court, must live with. To be sure, the concurrence was on firm ground in asserting the power to review the Florida Supreme Court’s construction of state 219. Obviously, members of Congress do not usually pore over constitutional law tomes prior to voting on proposals. Most members are probably inclined to allow the judiciary to resolve any disputed issue. But members of Congress formally have the obligation to uphold the Constitution. Some may solicit the views of others as to the constitutionality of various provisions, and others may in fact reach their independent view of the constitutional question prior to voting. 220. Fair Labor Standards Amendments of 1985, Pub. L. No. 99-150, 99 Stat. 787 (codifed at 29 U.S.C. § 207(o)(2)(A) (1994)). Moreover, the Court has construed legislation narrowly in light of Tenth Amendment concerns to preserve state sovereignty. In Gregory v. Ashcroft, 501 U.S. 452 (1991), the Court determined that Congress had not evinced a clear enough intent to subject state judges to the Age Discrimination in Employment Act. The Court explained that its clear statement approach ensures “that the legislature has in fact faced, and intended to bring into issue, the critical matters involved in the judicial decision.” Id. at 461 (quoting United States v. Bass, 404 U.S. 336, 349 (1971)). 221. The Court has recently enforced the Tenth Amendment more actively. See Printz v. United States, 521 U.S. 898 (1997) (limiting scope of Brady Handgun Violence Prevention Act); New York v. United States, 505 U.S. 144 (1992) (stating that Congress cannot compel states to enforce federal regulatory programs). 222. The Supreme Court’s decision in Buckley v. Valeo, 424 U.S. 1 (1976), can be viewed from a similar perspective. The Court may have upheld the $1,000 contribution limitation on the ground that, despite the First Amendment interests at stake, Congress was institutionally better able to set an appropriate limit. Id. at 24-30. The nondelegation doctrine can be viewed in similar light. See Whitman v. Am. Trucking Ass’n, 531 U.S. 457 (2001). 534 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:493 law. Otherwise, the Article II, Section 1 directive that state legislatures must select the manner in which presidential electors are chosen might become a dead letter. State judges through their interpretations might change the way that electors are selected and thereby rob the legislature of its constitutionally assigned function. But judicial interpretations should only rarely be considered tantamount to lawmaking. The Bouie line of cases and judicial impairments precedents manifest skepticism about the wisdom of secondguessing state court interpretations of state law—the concurrence simply ignored the doctrinal history. Courts backed off from full enforcement for a number of reasons. Drawing a line between interpreting and making law is notoriously slippery. Few benchmarks exist to determine when interpretation stops and lawmaking starts. Moreover, equating judicial interpretation with lawmaking provides incentive for disappointed litigants to seek redress in federal court for their disagreement with the state court interpretation of state law. And, the more we equate judicial interpretation with lawmaking, the more society may be willing to restrain judges. The lessons of history should have alerted the concurrence to the quagmire of robust enforcement of the Article II, Section 1 directive—some constitutional rights are better left underenforced. DISAPPEARING DEMOCRACY: HOW BUSH V. GORE UNDERMINED THE FEDERAL RIGHT TO VOTE FOR PRESIDENTIAL ELECTORS PETER M. SHANE* I. THE FEDERAL RIGHT TO VOTE FOR PRESIDENT .................................................. II. BUSH V. GORE AND VOTE TABULATION AS MASS ADJUDICATION ....................... A. Equal Protection and Due Process in Bush v. Gore .................................... B. Applying Due Process I: The Hypothetical Case of Gore v. Harris............. C. Applying Due Process II: The Real Case of Bush v. Gore ........................... III. REAL DEMOCRACY: BUSH V. GORE AND INSTITUTIONAL RESTRAINT ................. CONCLUSION ........................................................................................................ 537 550 550 553 568 578 584 Imagine, as you may already wish, that it is 2004. Despite flickers of opposition from Senator John McCain and former Nebraska Senator Robert Kerrey, George W. Bush and Al Gore appear to have sown up their respective renominations by mid-spring. In some state with, say, twenty-five electors, the Democratic legislature and the Democratic governor are worried. Polls predict a razor-thin Gore victory in the state, but, if voting machines malfunction or if inclement weather depresses the senior citizen vote, the state could be lost to Bush. No one wants a repeat of Florida’s 2000 travails. The solution? The legislature enacts a bill, eagerly signed by the Governor, providing that the state legislature itself, by a majority vote of each house, shall choose the state’s electors in 2004 for President and Vice President of the United States. Under Bush v. Gore,1 my hypothetical statute is constitutional. A state legislature’s authority to disenfranchise the entire citizenry is the very premise with which the majority in Bush v. Gore commences its legal analysis.2 But this premise is wrong. Its cavalier utterance by the majority exemplifies one of the opinion’s most extraordinary aspects, its obliviousness to the values of democracy. * Distinguished Service Professor of Law and Public Policy and Director of the Institute for the Study of Information Technology and Society, H. John Heinz III School of Law and Public Policy, Carnegie Mellon University; Professor of Law, University of Pittsburgh School of Law. I am grateful to Professor Bernard Hibbitts of the University of Pittsburgh School of Law and Dean Tony Sutin of the Appalachian School of Law, who created a virtual library on Presidential Election Law within the Jurist web site, see http://jurist.law.pitt.edu/election2000.htm, that proved invaluable for real-time research and analysis regarding the 2000 election. This Article benefitted also from comments by Professor Greg Magarian of Villanova Law School and attorneys Reed Hundt and Carolyn Shapiro, as well as from the research assistance of Michael Vasiliadis, University of Pittsburgh School of Law, Class of 2002. Copyright 2001. Peter M. Shane. All rights reserved. 1. 531 U.S. 98 (2000). 2. Id. at 104. 535 536 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 On any account, Bush v. Gore marked an astonishing event in the history of democratic governance. Never before in the history of democratic government has an unelected judicial organ chosen the head of state by preventing the counting of votes. Such an event cuts entirely against the grain of our political history. The past 200 years have witnessed a broadening of the franchise in the United States (and throughout the world) in terms of both eligibility and applicability. The right to vote has become central to our conception of citizenship. It is hard to imagine any modern-day Western theory of governmental legitimacy that does not rest in some essential aspect on “the electoral connection.” Thus, it was startling to witness the Supreme Court’s incautious embrace of a theory of the world’s most important elected office that treats its democratic character as merely discretionary. Bush v. Gore is antidemocratic in more than its ordination of a particular electoral outcome. It is oblivious to the democratic character of our Constitution in every aspect of its analysis. Its very starting point—the asserted authority of the states to disenfranchise voters altogether from participation in the selection of presidential electors—is unpersuasive in the face of the text and history of the Fourteenth Amendment, and Part I below explains why the Constitution ought now be interpreted to protect the rights of individuals to vote for state electors for President and Vice President of the United States. This analysis points to what should have been the foundational premise of Bush v. Gore: namely, the conspicuous trajectory of our constitutional development toward more democracy. And, it explains why the Florida Legislature would have been acting unconstitutionally in December 2000 had it proceeded to authorize its own slate of electors in lieu of those chosen on Election Day. Part II explores how the Fourteenth Amendment should have been deployed in light of democratic values to resolve the questions actually presented by Bush v. Gore. The majority purported to address an equal protection problem in Bush v. Gore, although none of the practices being challenged amounted, under anyone’s account, to a form of explicit or otherwise intentional discrimination against Bush voters—the sort of harm typically addressed through an equal protection rubric.3 By contrast, the importance of procedural due process in elections as an essential bulwark of democracy was utterly overlooked. Vote tabulation is a species of administrative adjudication, and voters should be deemed minimally entitled under the Fourteenth Amendment to voting tabulation systems rationally calculated to ascertain their intent accurately. A due process analysis of the Florida election supports the conclusion that hand recounts in 3. See Washington v. Davis, 426 U.S. 229 (1976). 2001] DISAPPEARING DEMOCRACY 537 challenged counties should have been deemed a constitutional prerequisite to the casting of Florida’s electoral votes. There was no legal or practical impediment to conducting a statewide recount consistent with the Fourteenth Amendment, and, compared to the Court’s assault on due process, the equal protection issue raised by the majority and accepted as nontrivial by Justices Souter and Breyer4 pales as insubstantial. Part III considers whether the issues presented in Bush v. Gore should have been considered political questions. The most compelling demand of democratic principle in Bush v. Gore may well have been that the Court defer to Congress, as an elected federal institution, for the resolution of a wholly political contest. Both the textual and the institutional arguments for deferring to Congress would have been stronger in Bush v. Gore than they were in Nixon v. United States.5 In that earlier case, Chief Justice Rehnquist, writing for himself and for Justices Stevens, Scalia, Kennedy, O’ Connor and Thomas, held it to be a political question whether the Senate was constitutionally required to hold a full evidentiary hearing before the entire Senate in order to convict and remove an impeached judge.6 The abandonment by the Bush v. Gore majority of this jurisprudential commitment, as well as others, helped to create the distressing impression that five Justices were determined to ordain a Bush victory, regardless of the legal merits. It is a measure of the Court’s imprudence in deciding Bush v. Gore both that such a possibility cannot be dismissed out of hand and that the Court could have responsibly deferred to Congress as the more accountable and constitutionally appropriate decisionmaking institution to resolve the Florida controversy. I. THE FEDERAL RIGHT TO VOTE FOR PRESIDENT Bush v. Gore commences its legal analysis by addressing a problem not actually at issue in the case: “The individual citizen has no federal constitutional right to vote for electors for the President of the United States unless and until the state legislature chooses a statewide election as the means to implement its power to appoint members of the Electoral College.”7 This is an odd starting place because, of course, there was no question in Bush v. Gore regarding Floridians’ entitlement to vote. One would have thought the more important premise would have been the fundamental status of the right to vote once conferred by state law, a legal proposition that the Court recognizes just a few sentences later: “When the state legisla4. 5. 6. 7. 531 U.S. at 134 (Souter, J., dissenting); id. at 145 (Breyer, J., dissenting). 506 U.S. 224 (1993). Id. at 226. 531 U.S. at 104. 538 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 ture vests the right to vote for President in its people, the right to vote as the legislature has prescribed is fundamental; and one source of its fundamental nature lies in the equal weight accorded to each vote and the equal dignity owed to each voter.”8 From this utterance, it sounds as if the Court is prepared to focus on democracy after all. But the rhetorical mood reverts immediately to reemphasize the contingent character of national democracy. The Court stresses in its very next sentence: “The State, of course, after granting the franchise in the special context of Article II, can take back the power to appoint electors.”9 The Court thus founds its analysis on the presumptive right of legislators to exclude citizens from presidential elections. Why start the opinion this way? In retrospect, the gambit looks obvious. The majority is about to apply the Equal Protection Clause in a way that will disenfranchise thousands of Florida voters. The majority anticipates that such may be the case.10 Yet, the majority is prepared to treat another issue as more important, namely, what the majority takes to be a threatened procedural error in the addition of hand-counted votes to Florida’s machine-tallied county vote totals. The suggestion implicit in the opening paragraph, foreshadowing the end game to come, is that any disenfranchisement that the Court inflicts should not be perceived as a harm more serious than the procedural infirmity that the Court is prepared to remedy. That is, the majority wants the country to believe that any improper diminishment of a ballot actually counted by Florida’s voting machines amounts to a more serious harm than the wholesale rejection of valid ballots that the voting machines have failed to register. Why? Because “[t]he individual citizen has no federal constitutional right to vote for electors for the President of the United States.”11 The Court’s implicit stance is that voters whose votes are counted but improperly weighted lose something that the Constitution protects, but that disenfranchised voters do not. There is no other sensible reason for starting where the Court starts.12 But this message is wrong. It is wrong, in part, because it could not possibly be the case in law or in reality that a voter whose ballot 8. Id. 9. Id. 10. Id. (“This case has shown that punchcard balloting machines can produce an unfortunate number of ballots which are not punched in a clean, complete way by the voter.”). 11. Id. 12. Actually, if one were yet more cynical about the majority than even I, there would be yet another explanation. The reiteration that state legislatures cannot only disenfranchise the electorate ex ante from the selection of presidential electors but may also take back the franchise once granted could be read as a message to the Florida Legislature: “If Gore does not concede following this opinion, you have the right to intervene.” For reasons described in the text below, I believe this message, if intended, would be wrong as to the Constitution. Its utterance would also embody so profound a corruption of the judicial role that I would prefer to think that the Court did not intend such a message. 2001] DISAPPEARING DEMOCRACY 539 is marginally devalued is worse off than a voter whose ballot is utterly discarded. Even more fundamentally, it is wrong because the Fourteenth Amendment, persuasively read, does guarantee individual citizens the right to vote for presidential electors. The straightforward argument for this position begins with Section 2 of the Fourteenth Amendment, which provides: Representatives shall be apportioned among the several States according to their respective numbers, counting the whole number of persons in each State, excluding Indians not taxed. But when the right to vote at any election for the choice of electors for President and Vice President of the United States, Representatives in Congress, the Executive and Judicial officers of a State, or the members of the Legislature thereof, is denied to any of the male inhabitants of such State, being twenty-one years of age, and citizens of the United States, or in any way abridged, except for participation in rebellion, or other crime, the basis of representation therein shall be reduced in the proportion which the number of such male citizens shall bear to the whole number of male citizens twenty-one years of age in such State.13 It is the italicized reference to “the choice of electors for President and Vice President of the United States” that most obviously ratifies this process as one in which individual citizens must be allowed to participate. Section 2 of the Fourteenth Amendment was drafted with a transparent aim. With the abolition of slavery, each African American living in a formerly Confederate state would now count as five-fifths and not merely three-fifths of a person for purposes of congressional apportionment.14 Congress was not yet ready in 1866 to approve a constitutional amendment guaranteeing suffrage regardless of race. But neither did the Republicans in control of Congress want to award the Confederate states an enlarged population, and thus additional seats in the House,15 if the constituency most likely to support the Republicans—namely, newly freed African Americans— were not going to be allowed to vote. Black suffrage, at least in theory, was to be the price of an enlarged congressional delegation. The original wording of Section 2, as reported to the House of Representatives by the Joint Committee of Fifteen on Reconstruction, had actually been significantly different in a critical respect: 13. U.S. CONST. amend. XIV, § 2 (emphasis added). 14. See id. art. I, § 2, cl. 3. 15. The enlargement of any state’s House delegation would also increase its number of votes in the Electoral College. See id. art. II, § 1, cl. 2 (each state shall have “a Number of Electors, equal to the whole Number of Senators and Representatives to which the State may be entitled in the Congress”). 540 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 Representatives shall be apportioned among the several States which may be included within this Union according to their respective numbers, counting the whole number of persons in each State, excluding Indians not taxed. But whenever in any State the elective franchise shall be denied to any portion of its male citizens not less than twenty-one years of age, or in any way abridged, except for participation in rebellion or other crime, the basis of representation in such State shall be reduced in the proportion which the number of male citizens shall bear to the whole number of such male citizens not less than twenty-one years of age.16 The important difference in this version, as compared to the Amendment as finally transmitted by Congress, is that the second sentence of the original version did not specify the elections to which it would apply. That is, this version would have reduced a state’s right of representation proportionally if, in any election, the state disenfranchised any portion of its male citizenry, twenty-one years of age or older, for reasons other than “participation in rebellion or other crime.” The House approved this version of the Amendment on May 10, 1866.17 In the Senate, however, the question was raised whether the scope of the second sentence was too great. Senator Henderson of Missouri raised the issue: Now, we have in our State an election for school directors, a general election held in every municipal township throughout the State of Missouri, at a certain time. At that election there are qualifications prescribed that we deem absolutely essential to keep up the common-school system in our State. For instance, property holders only vote for school directors. . . . There is an election also for school trustees. Those trustees are elected by the persons who have children to send to school. Now, if it be intended to exclude all persons who cannot vote at those elections from the basis of representation, I apprehend that not only will the negroes of my State be excluded under the proposed amendment, which will lose us a member in Congress, but it will exclude two thirds of the whites of the State of Missouri. I desire to know whether any such construction can be given to this proposition.18 There followed a brief colloquy in which there was agreement that the amendment would not cover school directors and trustees, but during which sponsors suggested that the language would cover “municipal officers.”19 Henderson suggested rewriting the disputed 16. 17. 18. 19. CONG. GLOBE, 39th Cong., 1st Sess. 2286 (1866) (remarks of Rep. Stevens). Id. at 2545. Id. at 3010 (remarks of Sen. Henderson). Id. (remarks of Sens. Fessenden and Henderson). 2001] DISAPPEARING DEMOCRACY 541 sentence to cover elections only for “Governor, judges, or members of either branch of the Legislature.”20 Senator Fessenden objected immediately that this would not cover elections for the House of Representatives.21 Senator Henderson responded that it would. The Constitution provides that “the Electors in each State [for the House of Representatives] shall have the Qualifications requisite for Electors of the most numerous Branch of the State Legislature.”22 Hence, Senator Henderson reasoned, regulating state electoral qualifications for the state legislature would necessarily have the desired impact on House of Representatives elections, as well. Anything more than that, according to Senator Henderson, should be beyond the purview of the new amendment. According to Henderson: The only election that can be held under the Constitution and laws of the United States is for members of Congress. Electors [for President and Vice President], as now provided by the Constitution, are to be elected by the State in any manner it chooses. . . . [U]nless you alter the Constitution on the subject the State Legislatures will yet have the power to regulate that matter entirely as they please, and this amendment will not change it at all. There is, therefore, but one election that can be held under the Constitution and the laws of the United States, and that is the election of members of the lower branch of Congress, because Senators are elected by the legislatures.23 Because Senator Howard of Michigan then raised questions, however, regarding the precise impact of the proposed Henderson amendment, the Senate was adjourned for the day so that Henderson and Howard might work out their differences.24 The next morning, Senator Williams of Oregon proposed that Section 2 be amended to apply to elections “held under the Constitution and laws of the United States, or of any State.”25 Objection was made immediately that this formulation was yet more confusing. Senator Johnson of Maryland suggested as an alternative: [T]hat the phraseology of this amendment, if it is to prevail, shall be so changed as to leave it beyond doubt that all that is meant is to except out of the whole number of inhabitants of the age of twenty-one years or upward, who are citizens of the State, those 20. Id. at 3011 (remarks of Sen. Henderson). 21. Id. (remarks of Sen. Fessenden). 22. U.S. CONST. art. I, § 2, cl. 1. 23. CONG. GLOBE, 39th Cong., 1st Sess. 3011 (1866) (remarks of Sen. Henderson) (emphasis added). 24. Id. 25. Id. at 3026 (remarks of Sen. Williams). 542 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 who are denied the right to vote at any State election, as contradistinguished from any municipal or local election.26 He then proceeded to attack Section 2 for a variety of allegedly perverse results he believed would follow from its application.27 Senator Williams interrupted, however, to concede only the point that Section 2 should specify with complete precision the elections that the amendment was intended to cover. He replaced the suggested phrase from the first version of his amendment to Section 2— “election held under the Constitution and laws of the United States or of any State”—with the phrase: “any election for the choice of electors for President and Vice President of the United States, Representatives in Congress, the executive and judicial officers of a State, or members of the Legislature thereof.”28 The Senate subsequently adopted this language,29 which was accepted also by the House.30 The substitution offered by Senator Williams and ultimately ratified as part of the Fourteenth Amendment embodied a critically important move, and it is intriguing that the point elicited no further discussion on the floor of either the Senate or the House. It must be recalled that, as Senator Henderson had expressly noted, the original Constitution provided three separate modes for the election of federal officials. The House of Representatives was to be chosen by “Electors in each State [who] shall have the Qualifications requisite for Electors of the most numerous Branch of the State Legislature.”31 In contrast, the Senate would be composed of “two Senators from each State, chosen by the Legislature thereof.”32 The President and Vice President would be chosen by electors that “[e]ach State shall appoint, in such Manner as the Legislature thereof may direct.”33 As Senator Henderson insisted, only one of these electoral processes guaranteed some form of franchise directly to individual citizens of the several states, namely, elections for the House. A popular franchise was precluded altogether by the constitutional provisions regarding the Senate and permitted, but not required, with regard to the selection of presidential electors. And yet, in translating his original formulation, “election[s] held under the Constitution and laws of the United States,”34 into a more specific articulation, Senator Williams included among the elections covered by Section 2 “any 26. 27. 28. 29. 30. 31. 32. 33. 34. Id. at 3027 (remarks of Sen. Johnson). Id. at 3026-29 (remarks of Sen. Johnson). Id. at 3029 (remarks of Sen. Williams). Id. at 3040. Id. at 3149. U.S. CONST. art. I, § 2, cl. 1. Id. art. I, § 3, cl. 1. Id. art. II, § 1, cl. 2. CONG. GLOBE, 39th Cong., 1st Sess. 3029 (1866) (remarks of Sen. Williams). 2001] DISAPPEARING DEMOCRACY 543 election for the choice of electors for President and Vice President of the United States, [or] Representatives in Congress.”35 The issue is how to interpret this translation. If what counts in interpreting text are subjective states of mind, then, of course, we cannot know with certainty either Senator Williams’s thought processes or those of the legislators who had the chance to review his handiwork. One possibility is that Senator Williams did not intend Section 2 as implying state compulsion to hold any particular elections. Section 2, that is, may provide only a limited rule of nondiscrimination in elections that are guaranteed by some law other than Section 2 itself, whether that law is another federal constitutional provision or, in most cases, a state law. On this hypothesis, he included elections for presidential and vicepresidential electors only to recognize that many such elections are held under laws of the respective states, and, to the extent they are held, he wanted them held, among men at least, on a racially nondiscriminatory basis. The strongest argument in support of this reading is that Section 2 applies to “any election for . . . Judicial officers of a State”36 but has not been interpreted to require state judicial elections. This might be read likewise to leave elections in connection with the choice of presidential electors optional. The reason would be that, if the reference to “any election for the choice of electors for President and Vice President”37 compelled the use of some popular vote, then referring to “any election for . . . Judicial officers of a State”38 would have to compel the use of a popular vote to choose judges, as well. Both clauses, to this extent, ought to be read in the same way. This argument, however, is hardly conclusive. It might make perfect sense on federalism grounds to read Section 2 as embodying a background guarantee that the federal offices to which it refers are mandatorily subject to popular votes, while selection systems for the state offices to which it refers remain discretionary with the states. Indeed, a devastating problem for the more limited interpretation regarding presidential elections is that, if the use of the popular vote in presidential elections remains discretionary under Section 2, then it would be easy for state legislatures to undermine the Republicans’ aim of making black suffrage the price of reempowering the Southern states. Imagine that Section 2 is read to cover House elections but not to require popular votes for presidential electors. Imagine further that a Southern state subsequently enfranchised all propertied male 35. 36. 37. 38. Id. U.S. CONST. amend. XIV, § 2. Id. Id. 544 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 adult citizens in House elections but barred black men from voting for President. The consequence of allowing all men to vote for House seats would likely be an enlarged congressional delegation for the state in question because the entirety of the state’s black population would now be counted as part of its proper apportionment base. The enlarged House delegation would necessarily increase the size of the state’s presidential elector delegation because the number of a state’s electors is equal to the total of its senatorial delegation (always two) and its House delegation. Disenfranchising black voters in such a state from presidential elections would defeat the objectives of congressional Republicans because white voters in this hypothetical Southern state would then have managed to achieve a larger voice in the Electoral College without permitting blacks to vote for President. This would subvert a significant portion of what Section 2 was intended to achieve. So long as we interpret the Constitution as making popular involvement in presidential elections no more than a state legislative prerogative, Section 2 cannot prevent white Southerners from doing just this very thing. Senator Henderson made precisely this point with regard to an earlier formulation of Section 2.39 All that need happen in a Southern state intent on maintaining white control of presidential elector appointments is for the majority-white legislature to institute or maintain a practice of having itself select slates of presidential electors without any popular vote involvement. In other words, Section 2 can effectively impose black suffrage as the price of increased Southern strength in the Electoral College only if each state is required to include a popular vote mechanism in the process of choosing its presidential electors. Thus, whatever his subjective intentions, Senator Williams’s textual revision makes purposive sense only if the Constitution, as altered by the Fourteenth Amendment, is deemed to signal a background understanding that individual citizens in the several states would now be guaranteed some form of franchise in presidential elections, as well as in House contests.40 39. CONG. GLOBE, 39th Cong., 1st Sess. 3011 (1866) (remarks of Sen. Henderson). 40. My argument thus does not rest on the supposition that, because of their denotational meaning and Section 2’s syntax alone, the words of Section 2 must be interpreted as commanding popular involvement in presidential elections. It is rather the purpose behind the inclusion of those words that commands popular involvement in presidential elections just as, say, it is Article I, Section 2 that provides the requirement that a popular vote be staged in order to select members of the House of Representatives. On the same basis, unless there is something in the Federal Constitution other than the syntax of Section 2 that compels state judicial elections, then the mere linguistic reference to state judicial elections in Section 2 would not be enough to deprive states of discretion to prefer appointed judiciaries. That “something” would have to be either another clause in the Constitution or a demonstration that eliminating appointed state judiciaries is implicit in the history or design of Section 2. 2001] DISAPPEARING DEMOCRACY 545 For these reasons, one need not look further than Section 2 for a constitutional guarantee that individual citizens of the several states shall be entitled to participate in the selection of presidential electors. The near-complete consistency of post-Fourteenth Amendment state practice with this reading also validates it.41 On only two occasions since ratification has a state legislature taken upon itself the authority to appoint an entire slate of presidential electors. Florida did so in 1868, presumably because the process was already in place prior to the ratification of the Fourteenth Amendment on July 21, 1868.42 Colorado’s legislature did the same in 1876, apparently because the grant of statehood that year was deemed to leave too little 41. On the far more frequent appointment of presidential electors by state legislatures prior to the Fourteenth Amendment, see McPherson v. Blacker, 146 U.S. 1, 29-33 (1892): At the first presidential election the appointment of electors was made by the legislatures of Connecticut, Delaware, Georgia, New Jersey, and South Carolina . . . . Fifteen States participated in the second presidential election, in nine of which electors were chosen by the legislatures . . . . Sixteen States took part in the third presidential election, Tennessee having been admitted June 1, 1796. In nine States the electors were appointed by the legislatures . . . . In the fourth presidential election, Virginia, under the advice of Mr. Jefferson, adopted the general ticket, at least ‘until some uniform mode of choosing a President and Vice-President of the United States shall be prescribed by an amendment to the Constitution.’ Laws Va. 1799, 1800, p. 3. Massachusetts passed a resolution providing that the electors of that State should be appointed by joint ballot of the senate and house. . . . Pennsylvania appointed by the legislature, and upon a contest between the senate and house, the latter was forced to yield to the senate in agreeing to an arrangement which resulted in dividing the vote of the electors. 26 Niles’ Reg. 17 . . . . . . . Massachusetts . . . chose electors by joint ballot of the legislature in 1808 and in 1816. . . . The appointment of electors by the legislature, instead of by popular vote, was made use of by North Carolina, Vermont, and New Jersey in 1812. In 1824 the electors were chosen by popular vote, by districts, and by general ticket, in all the States excepting Delaware, Georgia, Louisiana, New York, South Carolina, and Vermont, where they were still chosen by the legislature. After 1832 electors were chosen by general ticket in all the States excepting South Carolina, where the legislature chose them up to and including 1860. . . . And this was the mode adopted by Florida in 1868, (Laws 1868, p. 166) and by Colorado in 1876, as prescribed by § 19 of the schedule to the constitution of the State, which was admitted into the Union, August 1, 1876, Gen. Laws Colo. 1877, pp. 79, 990. McPherson upheld a challenge to Michigan’s appointment of presidential electors based on the popular vote in each congressional district. Although language in the case is strongly supportive of the legislature’s plenary power to employ any mode of appointment whatever, McPherson, 146 U.S. at 10, the case simply did not present the question I am discussing here. The Michigan legislature had not purported to appoint presidential electors through a process independent of a popular vote. 42. See JOHN E. NOWAK & RONALD D. ROTUNDA, CONSTITUTIONAL LAW 1440 (6th. ed. 2000). 546 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 time to put a popular vote for electors into place.43 As Justice Frankfurter wrote in another legal context: Deeply embedded traditional ways of conducting government cannot supplant the Constitution or legislation, but they give meaning to the words of a text or supply them. It is an inadmissibly narrow conception of American constitutional law to confine it to the words of the Constitution and to disregard the gloss which life has written upon them.44 Popular votes for presidential electors are a “deeply embedded traditional way” of conducting presidential elections, and this history should be decisive in resolving any ambiguity in Section 2. There is, however, yet another route to recognizing a Fourteenth Amendment right for voters to participate in the selection of presidential electors. It would start with accepting as a premise that Section 2 has only the more limited meaning suggested above, namely, that Section 2 embodies a limited nondiscrimination rule in certain categories of elections, so long as the popular franchise in such elections is guaranteed by some law other than Section 2 itself. In the case of presidential electors, it is possible to infer such a guarantee outside Section 2—specifically, from the Privileges or Immunities Clause in Section 1. That clause provides: “No State shall make or enforce any law which shall abridge the privileges or immunities of citizens of the United States.”45 Prospects for establishing a robust jurisprudence of individual rights based on “privileges or immunities” were shattered by the Court’s 5-4 decision in the 1872 Slaughterhouse Cases.46 That decision rejected a Fourteenth Amendment challenge by New Orleans butchers to a monopoly granted by the Louisiana Legislature to the Crescent City Live-Stock Landing and Slaughter-House Company to operate a slaughterhouse in New Orleans. The butchers based their challenge on a variety of grounds, including protections assertedly embodied in the Privileges or Immunities Clause against arbitrary laws in violation of inalienable fundamental rights.47 The majority rejected this analysis. According to the majority, the “privileges or immunities” to which the Fourteenth Amendment referred were exclusively “privileges or immunities” of national, rather than state citizenship. These, the Court said, em- 43. Colorado was admitted to the Union on August 1, 1876. McPherson, 146 U.S. at 910. 44. Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579, 610 (1952) (Frankfurter, J., concurring). 45. U.S. CONST. amend. XIV, § 1. 46. 83 U.S. 36 (1872). 47. Id. at 66. 2001] DISAPPEARING DEMOCRACY 547 braced only those rights that owed “their existence to the Federal government, its National character, its Constitution, or its laws.”48 However meager this interpretation of the Privileges or Immunities Clause, it is still sufficient to ground an individual right to participate in the selection of presidential electors.49 Although chosen through processes that states design and administer, presidential electors hold a federal office that is rooted entirely in the national Constitution. This point was not lost in the Fourteenth Amendment debates. During floor discussion immediately prior to the House adoption of the Fourteenth Amendment, Representative Bingham stated, without contradiction: “The franchise of a Federal elective office is as clearly one of the privileges of a citizen of the United States as is the elective franchise for choosing Representatives in Congress or presidential electors. They are both provided for and guaranteed in your Constitution.”50 It is thus entirely sensible, especially in light of Section 2 of the Fourteenth Amendment, to interpret the Privileges or Immunities Clause also as embracing a right of individual citizens to participate in the choice of presidential electors.51 48. Id. at 79. 49. In urging this interpretation, I am plainly giving little or no weight to Minor v. Happersett, 88 U.S. 162 (1874), which considered whether the Privileges or Immunities Clause of the Fourteenth Amendment conferred a right of state suffrage upon women. In reaching its negative conclusion to this inquiry, the Court held that neither the Privileges or Immunities Clause, nor any other clause of the Constitution, creates voters per se. Hapersett is of little consequence, however, to modern voting rights questions. Despite the Court’s repeated insistence that the Constitution does not vest suffrage directly, the Court has also characterized the right to vote as “fundamental.” E.g., Harper v. Va. State Bd. of Elections, 383 U.S. 663, 670 (1966). It is on that basis that the Court has further held that classifications limiting the franchise must be subject to a “close and exacting examination.” Kramer v. Union Free Sch. Dist. No. 15, 395 U.S. 621, 626 (1969). The consequences of requiring such an examination are generally indistinguishable from what would be the consequences of holding that the Constitution does directly confer suffrage rights, which, in turn, may not be subjected to unreasonable burdens. In other words, it would hardly matter if the Privileges or Immunities Clause did not directly vest rights in anyone to help choose presidential electors, so long as the clause is read to bar a state legislature from excluding the entire citizenry from any elector selection process that the legislature decides to authorize. It would be sensible to conclude, that is: (1) that the “privileges or immunities” of citizenship include freedom from arbitrary limitations on those whom the state, at its discretion, enfranchises to choose presidential electors, and (2) that any classification that eliminates from the process all members of the electorate who are not also state legislators is constitutionally impermissible. Those more respectful than I of Happersett might prefer to attach such conclusions to the Equal Protection Clause rather than the Privileges or Immunities Clause. Of course, such a result would be equally congenial to the democratic values of our present-day Constitution. 50. CONG. GLOBE, 39th Cong., 1st Sess. 2542 (1866). 51. It may be questioned whether the history of the Fifteenth Amendment casts doubt on my reading of the Fourteenth. That is because, in deliberations leading to the latter amendment, the Senate initially approved an additional constitutional reform that would have expressly required that presidential electors in each state be chosen by a popular vote in which the franchise would extend to all those qualified in the respective states to vote for members of the House of Representatives. See CONG. GLOBE, 40th Cong., 3d Sess. 1042 (1869). At first blush, it might seem implausible that such a proposal would have surfaced 548 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 There is, of course, a perhaps even more compelling reason to read the Fourteenth Amendment as supportive of democracy: namely, the plain democratic trajectory of constitutional development since 1868. Since the ratification of the Fourteenth Amendment, we have added thirteen other amendments to the U.S. Constitution. Six of these were specifically intended to further our constitutional commitment to the democratic process: extending the vote to persons of all races,52 providing for the direct election of Senators,53 extending the franchise to women,54 permitting District of Columbia voters to choose electors,55 eliminating federal poll taxes,56 and lowering the voting age to in 1869 if the Fourteenth Amendment had already required that state legislatures incorporate a popular vote mechanism into the selection of presidential electors. In fact, however, my reading of the Fourteenth Amendment would not have made this additional proposal superfluous. The 1869 proposal would have gone beyond the Fourteenth Amendment in constraining the role of state legislatures in the appointment of presidential electors. It would have eliminated that role altogether. This move would have left to Congress the decision whether to provide winner-take-all systems in every state or to permit electors to be chosen by congressional district. So marked a diminution in state authority might itself have been a sufficient reason to support or oppose the proposal, regardless of what the Fourteenth Amendment had already accomplished. So democratizing a move was consistent with the breadth of radical Republican proposals for the Fifteenth Amendment, which envisioned an end not only to racial restrictions on the franchise but also an end to restrictions based on property holding, taxpayer status, nativity, and literacy. See ALEXANDER KEYSSAR, THE RIGHT TO VOTE: THE CONTESTED HISTORY OF DEMOCRACY IN THE UNITED STATES 95 (2000) (discussing proposals by Representative Samuel Shellabarger of Ohio and Senator (later Vice President) Henry Wilson of Massachusetts). The fact of the 1869 proposal, therefore, casts little light on what its supporters or opponents understood to be the proper interpretation of either the Privileges or Immunities Clause or Section 2 of the Fourteenth Amendment. Of course, even if we assume that its drafters would not have consciously anticipated that the Fourteenth Amendment created an individual right of participation in choosing presidential electors, that is hardly conclusive as to the most compelling reading of the text. The inference now of such a right would follow a well-established jurisprudence that extends the Fourteenth Amendment to applications unanticipated in 1868. Prominent examples include Brown v. Board of Education, 347 U.S. 483 (1954) (invalidating the government-imposed racial segregation of public schools), and Kramer, 395 U.S. 621 (invalidating restriction of school district franchise to voters who were parents of children enrolled in district’s schools or otherwise owners or lessors of taxable property). Especially because so little discussion occurred to illuminate the drafters’ subjective intentions, interpretation of the Fourteenth Amendment with regard to presidential elections should rest on the most plausible implications of its wording and not on the subjective expectations of the members of the Thirty-ninth Congress. And, as compared to the innovative equal protection cases of the past half century, a decision to interpret the Fourteenth Amendment as assuring the right to vote in presidential elections would rest on textual clues more specific and, to that limited extent, more compelling than the text available to guide a number of such earlier cases. See, for example, Harper v. Virginia State Board of Elections, 383 U.S. 663 (1966), which managed to invalidate state poll taxes under the Constitution despite a constitutional amendment just two years earlier that had eliminated the poll tax explicitly only with regard to federal elections. 52. U.S. CONST. amend. XV. 53. Id. amend. XVII. 54. Id. amend. XIX. 55. Id. amend. XXIII. 56. Id. amend. XXIV. I am grateful to attorney Vasan Kesavan for pointing out to me that the wording of the Twenty-fourth Amendment would seem especially odd if the Con- 2001] DISAPPEARING DEMOCRACY 549 eighteen.57 It is unthinkable, against this history of constitutional development, that a state legislature should still be deemed authorized to usurp the people’s role in choosing presidential electors. What difference does this make? Had the Supreme Court not decided Bush v. Gore, the Florida Legislature stood poised to test these very principles. On December 12, 2000, the Florida House of Representatives named a slate of twenty-five electors for candidate Bush, anticipating that judicial processes for certifying the state’s presidential vote might extend beyond December 18, the date for federal electoral balloting.58 The Florida Senate abandoned the measure only after Gore conceded the election in the wake of the Supreme Court’s decision, thus averting the possibility of constitutional crisis.59 But this near-miss clearly highlights the legal question: What should have been the result had the legislature gone ahead with its slate and Florida Gore voters sought to enjoin balloting by a legislatively chosen set of electors? Such a move would plainly have deprived the voters of Florida of their Fourteenth Amendment right to participate in the selection of presidential electors. As of the time of the Florida House vote, proceedings in the Florida courts had every prospect of reaching a timely conclusion. Federal law provides: “Whenever any State has held an election for the purpose of choosing electors, and has failed to make a choice on the day prescribed by law, the electors may be appointed on a subsequent day in such a manner as the legislature of such State may direct.”60 As of December 12, however, it had not been established that the people of Florida had “failed to make a choice on the day prescribed by law.” Moreover, any legislative authority to direct the appointment of electors on a day subsequent to Election Day would still have to be exercised in a manner consistent with the Fourteenth Amendment requirement for including a process of popular voting in the selection of presidential voters.61 Giving the Fourstitution does not require state legislatures to include a popular vote mechanism in the section of electors. 57. Id. amend. XXVI. 58. Marego Athans, Florida House OKs Bush Elector Slate—Measure is Passed Along Party Lines 79-41, After 5½ Hour Debate, BALT. SUN, Dec. 13, 2000, at 22A, available at 2000 WL 4886822. 59. Florida Legislature Passes Up Naming Slate of Electors, COM. APPEAL (Memphis, Tenn.), Dec. 15, 2000, at A11, available at 2000 WL 27945697. 60. 3 U.S.C. § 2 (1994). 61. It would additionally have been offensive to the Constitution to undo Florida’s legally authorized November 7 balloting retroactively. As the Bush brief argued in the Supreme Court (in a far less persuasive context): “[C]onstitutional principles of due process and fundamental fairness preclude the States from adopting ‘a post-election departure from previous practice’ and applying that post-election rule retroactively to determine the outcome of an election.” Brief For Petitioner at 28, Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000) (No. 00-836) (citing Roe v. Alabama, 43 F.3d 574, 581 (11th Cir. 550 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 teenth Amendment its most compelling reading would have rendered Florida’s attempted legislative appointment of electors unconstitutional. But a proper reading of the Fourteenth Amendment with regard to the popular franchise in presidential elections should have been influential in Bush v. Gore itself. As noted earlier, the question whether Floridians were constitutionally entitled to participate in the appointment of presidential electors was technically not an issue presented in this case.62 Imagine, however, that the majority’s legal analysis had started as follows: “The Fourteenth Amendment has granted Americans, as individual citizens, a federal constitutional right to vote for electors for the President of the United States.” It is implausible that a decision so starting could have ended with the judicially compelled disenfranchisement of thousands of Florida voters. This constitutional premise, diametrically opposed to the Court’s actual starting point, would have required that every ambiguity be resolved and every presumption indulged in favor of a comprehensive count of all Florida ballots. Thus, it is hardly a surprise that the Florida Supreme Court’s first opinion in this episode begins with that state’s fundamental legal commitments to the right to vote and to a determination of electoral outcomes that reflects “the will of the people.”63 It is shameful that the Supreme Court of the United States spurned those commitments. II. BUSH V. GORE AND VOTE TABULATION AS MASS ADJUDICATION A. Equal Protection and Due Process in Bush v. Gore The issue that the Court did have to face in Bush v. Gore was the consistency with federal law of the manual recount of Florida ballots that had been ordered by the Florida Supreme Court in response to Gore’s contest of the certification of Bush electors. Five Justices held that “[t]he recount mechanisms implemented in response to the decisions of the Florida Supreme Court do not satisfy the minimum re1995)), available at http://election2000.stanford.edu/briefforpetitioner.pdf. Nothing in Florida’s law made the November 7 election merely advisory to the state legislature. Second, in purporting to adjudicate the actual outcome of the November 7 election, the Florida Legislature would have been grabbing judicial authority in violation of due process. It was state legislative usurpation of judicial power during the 1780s that motivated the Philadelphia drafters to provide the nation with an independent federal judiciary and to oust Congress from adjudication except with regard to impeachment and judging the qualifications and conduct of its own members. I.N.S. v. Chadha, 462 U.S. 919, 961-962 (1983) (Powell, J., concurring in the judgment). Now that the Fourteenth Amendment imposes due process obligations upon the states, state legislatures should be deemed to have identical limits on their capacity to assert adjudicatory power. 62. See supra text accompanying note 7. 63. Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d 1220, 1228 (Fla. 2000), rev’d sub nom. Bush v. Palm County Canvassing Bd., 531 U.S. 70 (2000). 2001] DISAPPEARING DEMOCRACY 551 quirement for nonarbitrary treatment of voters necessary to secure the fundamental right [to equal protection].”64 Although the majority suggested a variety of respects in which the recount mechanisms fell short of constitutional requirements, the essential infirmity of the ordered recount was ostensibly the absence of a uniform standard to be followed in each county for the discernment of voter intent. Because, based on this infirmity, “it [was] evident that any recount seeking to meet [a] December 12 [deadline] will be unconstitutional,” the Court “reverse[d] the judgment of the Supreme Court of Florida ordering a recount to proceed.”65 I think the Court’s analysis was wrong on its own terms—its remedial holding is especially indefensible. But one of the more unfortunate aspects of the opinion is not just that it is wrong, but that it focuses on the wrong, or at least the less compelling, thing—equal protection—rather than due process.66 This may seem an odd complaint. I do not mean to suggest that, if the majority had simply decided Bush v. Gore under the “right” clause, five Justices would have come out the right way. As a rhetorical matter, however, the focus on equal protection rather than due process tends to obscure the real is64. Bush v. Gore, 531 U.S. 98, 105 (2000). Justice Souter, while dissenting from the Court’s remedial order (and, indeed, from its grant of review), agreed that “no legitimate state interest” justified the imposition of different standards in different counties for measuring voter intent. Id. at 134 (Souter, J., dissenting). Justice Breyer took a more nuanced position. He implied that the availability in Florida of judicial review to examine disputed ballots would, in ordinary times, suffice to permit recounts to proceed under an administrative standard no more specific than the general “intent of the voter” standard imposed by the Florida Supreme Court. Id. at 145 (Breyer, J., dissenting). Because “time was, and is, too short,” however, “to permit the lower courts to iron out significant differences through ordinary judicial review,” and because the Florida Supreme Court’s order had already included tabulations from counties employing different standards, Breyer agreed that, “in these very special circumstances, basic principles of fairness should have counseled the adoption of a uniform standard to address the problem.” Id. at 145-46. “In light of the majority’s disposition” of the case, Breyer did not discuss “whether, or the extent to which, as a remedial matter, the Constitution would place limits upon the content of the uniform standard.” Id. at 146. 65. Id. at 110. 66. The per curiam opinion is actually somewhat careless in identifying the constitutional clause on which it hangs its analysis. At the outset, the majority seems pointed in resting its conclusions on equal protection. It foreshadows its discussion as follows: “The petition presents the following questions: . . . whether the use of standardless manual recounts violates the Equal Protection and Due Process Clauses. With respect to the equal protection question, we find a violation of the Equal Protection Clause.” Id. at 103. This framing of the issues seems to leave due process out of the equation. Yet, after describing the alleged procedural defects of the judicially mandated recount, the Court simply tosses in, without any analysis, due process as an additional ground for its conclusion. “Upon due consideration of the difficulties identified to this point, it is obvious that the recount cannot be conducted in compliance with the requirements of equal protection and due process without substantial additional work.” Id. at 110. Because the reference to “the requirements of . . . due process” occurs without any specification of what they might be, I think it a fair statement that the majority gave no sustained thought to the Florida election from a due process standpoint. 552 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 sues confronting Florida in designing and implementing a fair vote tabulation system. The Equal Protection Clause may seem at first the more natural lens through which to view an election contest. Equal protection challenges generally involve the differential treatment of persons, implicit or explicit, because of groups to which they belong, for example, groups of Bush supporters or Gore supporters. Unlike typical equal protection cases, however, Bush v. Gore does not exhibit any obvious link between a challenged classification or criterion for treatment and any group’s systematic advantage or disadvantage.67 The majority opinion identifies three sources of differentiation in the treatment of voters by different counties: differential standards for evaluating contested ballots,68 differences between recounts limited to undervotes and recounts that reexamined all of the ballots cast,69 and differences between tabulations based on partial recounts versus tabulations based on completed recounts.70 Legally speaking, these may be phenomena worth noting, but it is hard to predict who will be hurt by them. There was no allegation that any amounted to an intentionally invidious discriminatory practice—the kind of practice that the Court normally requires before it elevates the intensity of its constitutional scrutiny of state practices under the Equal Protection Clause.71 Due process, by contrast, focuses on the adequacies of a governmental process for making adjudicatory decisions, whether formal or informal. Adjudicatory decisions are what vote tabulation, most obviously manual vote counting, is all about. That is, when an administrative body counts votes, it is discerning on the basis of a formal piece of evidence—a ballot—the intent of the voter regarding the outcome of a political contest. Analytically, due process provides the more helpful framework for assessing the fairness of an adjudicatory system to all those affected by its outcomes. It also is directly supportive of the democratic principle. There would plainly be little significance to the practice of voting if voters were not assured rational treatment in the counting of their votes. 67. Bush was clearly being hurt by the Florida recount only in the sense that, having been certified the statewide winner by 537 votes on November 26, see BUSH V. GORE: THE COURT CASES AND THE COMMENTARY xiii (E.J. Dionne Jr. & William Kristol eds., 2001), he would clearly have been better off if he could have prevented any further counting at all, regardless of the procedures used. There is no obvious way in which the kinds of disparities mentioned by the Court could be said to favor Democrats or Republicans. From all the Court knew on December 12, Bush would have been threatened as much by the constitutionally most exacting recount as he would by any other. 68. Bush v. Gore, 531 U.S. at 106-07. 69. Id. at 107. 70. Id. at 108-09. 71. See Washington v. Davis, 426 U.S. 229 (1976). 2001] DISAPPEARING DEMOCRACY 553 To make that assessment comprehensively, however, I would like to look at two cases: a hypothetical Gore v. Harris and the actual Bush v. Gore. The point of the first inquiry is both to establish the amenability of the Florida election to a sound due process analysis and to point out that, with regard to the issue of procedural fairness, the per curiam opinion again ignores the proper constitutional baseline from which it should have judged the Bush challenges. That is, Bush v. Gore should have been resolved against a baseline understanding of Gore’s constitutional entitlement to a manual recount in four disputed counties. Even if the Privileges or Immunities Clause and Section 2 of the Fourteenth Amendment conferred no right on Florida voters to participate in the appointment of presidential electors, the voters were still entitled to an adjudication of voter intent consistent with due process. The Supreme Court effectively deprived them of that right. B. Applying Due Process I: The Hypothetical Case of Gore v. Harris To see the implications of due process from the position of Democratic voters, it is helpful to start a chapter or two back from the actual decision in Bush v. Gore and to consider how due process would have looked in the week after Election Day from the Gore perspective. To do so requires us to recall where matters stood legally about a month before the Supreme Court’s final decision. To make that recollection possible, a brief review of Florida law in effect at the time of the election72 is necessary. In Florida, vote counting within each county was the initial responsibility of so-called election boards, which include inspectors and clerks for every precinct who are appointed to their positions by the Supervisors of Elections in each of the respective counties.73 The boards in each county were required to prepare certified tallies, which were delivered, in turn, to the Supervisor of Elections and to the county court judge.74 In addition, each Supervisor of Elections sat on a so-called county canvassing board, along with two other members—the county court judge, who acted as chair of the canvassing board, and the chair of the board of county commissioners.75 It was the primary duty of each county canvassing board to canvass absentee ballots, and then prepare a canvass of the entire election result within each county, “as shown by the returns . . . on file in the office of the supervisor of elections and the office of the county court 72. The Florida Legislature passed wholesale amendments to its election procedures in the 2001 session. See Florida Election Reform Act of 2001, 2001 Fla. Laws ch. 40, at 11773. 73. FLA. STAT. § 102.012 (2000) (amended 2001). 74. Id. § 102.071. 75. Id. § 102.141(1). 554 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 judge.”76 The county canvassing board was then charged with certifying the county’s results to the Secretary of State, who was required ultimately to certify the results of any statewide election.77 Florida law vested county canvassing boards with essential responsibilities in the event of challenged elections. The relevant statute permitted “[a]ny candidate whose name appeared on the ballot” to file within seventy-two hours of an election “a written request with the county canvassing board for a manual recount.”78 It also permitted “[a]ny candidate . . . to protest the returns of the election as being erroneous by filing with the appropriate canvassing board a sworn, written protest . . . prior to the time the canvassing board certifies the results for the office being protested or within 5 days” thereafter.79 The statute provided that a county canvassing board that had been asked for a manual recount “may” authorize such a process, provided that the “[t]he manual recount must include at least three precincts and at least 1 percent of the total votes cast for such candidate or issue.”80 Although Florida law specified no criteria for proceeding with such a preliminary manual recount, it did mandate a particular course of events should such a preliminary recount be directed: If the manual recount indicates an error in the vote tabulation which could affect the outcome of the election, the county canvassing board shall: (a) Correct the error and recount the remaining precincts with the vote tabulation system; (b) Request the Department of State to verify the tabulation software; or (c) Manually recount all ballots.81 In counties using paper ballots that cannot be read properly because of uncorrectable problems with the tabulation devices involved, only the third option—manually recounting all ballots—was available to fulfill what appears to be a statutory duty under this section. Within forty-eight hours of election day, a machine recount required under Florida law for close elections narrowed Bush’s initial 1784-vote lead to 327 votes.82 There was no mechanism available to Gore at that point to trigger a unified statewide recount. He could 76. 77. 78. 79. 80. 81. 82. Id. § 102.141(2) (amended 2001). Id. §§ 102.151, 102.155. Id. § 102.166(4)(a) (amended 2001). Id. § 102.166(1)(2) (amended 2001). Id. § 102.166(4)(d) (amended 2001). Id. § 102.166(5) (amended 2001). BUSH V. GORE: THE COURT CASES AND THE COMMENTARY, supra note 67, at xi. 2001] DISAPPEARING DEMOCRACY 555 have sought manual recounts in every county in Florida, but his advisers believed that such a strategy would be both chaotic and unnecessary and might appear too impolitic—too much the desperate move of a sore loser.83 The decision was made on November 9 to pursue recounts in only four heavily Democratic counties from which numerous complaints of irregularities had emerged—Broward, Miami-Dade, Palm Beach, and Volusia. The Gore team hoped that recounts in these counties would suffice to overcome the tissue-thin Bush lead.84 By this time, it became equally clear that the Bush campaign and Republican election officials in Florida would try to prevent the outcomes of hand recounts from affecting the certification of Florida’s statewide vote. In particular, Katherine Harris, Florida’s Secretary of State and co-chair of the Bush campaign in Florida, announced that she would not waive the apparent November 14 statutory deadline for the submission to her of county vote totals for certification.85 On Saturday, November 11, the Bush campaign sought from the U.S. District Court for the Southern District of Florida an injunction to block any hand recounts.86 I refer to November 14 as the “apparent” deadline because, as the Florida courts would soon discuss, Florida statutes were resolutely ambiguous on the issue of deadlines. On one hand, section 102.111, Florida Statutes, stated in seemingly unequivocal terms: “If the county returns are not received by the Department of State by 5 p.m. of the seventh day following an election, all missing counties shall be ignored, and the results shown by the returns on file shall be certified.”87 This provision was echoed by section 102.112(1), Florida Statutes, which appeared to obligate county canvassing boards to submit county returns within seven days: The county canvassing board or a majority thereof shall file the county returns for the election of a federal or state officer with the Department of State immediately after certification of the election results. Returns must be filed by 5 p.m. on the 7th day following the first primary and general election and by 3 p.m. on the 3rd day following the second primary.88 These two provisions seemed to dictate that the 2000 county returns would have to be returned by November 14, the seventh day following the general election. 83. See Dave Von Drehle et al., In Florida, Drawing the Battle Lines: Big Guns Assembled as Recount Began, WASH. POST, Jan. 29, 2001, at A01. 84. Id. 85. BUSH V. GORE: THE COURT CASES AND THE COMMENTARY, supra note 67, at xi. 86. Id. at xii. 87. FLA. STAT. § 102.111(1) (amended 2001). 88. Id. § 102.112(1) (amended 2001). 556 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 That very same paragraph of section 102.112, however, stated equally explicitly that the decision whether or not to include in statewide totals any county returns that are submitted after the specified deadline was within the discretion of the Department of State: “If the [county] returns are not received by the department by the time specified, such returns may be ignored and the results on file at that time may be certified by the department.”89 The confusion was compounded in the next subsection, which provided: “The department shall fine each board member $200 for each day such returns are late, the fine to be paid only from the board member’s personal funds.”90 Under this provision, it was plainly the policy of the Florida Legislature that county canvassing boards should be encouraged to submit even late returns as early as possible. This would be a puzzling concern if county returns even an hour or a day late were mandatorily to be ignored. Notwithstanding Secretary Harris’s insistence on the November 14 deadline, both Volusia and Palm Beach Counties readied to do the preliminary hand recounts authorized by section 102.166(4), Florida Statutes. Palm Beach County, home of the controversial butterfly ballot, presented the threshold question of how to discern voter intent from a punch ballot that had not been read by the vote tabulation machine. That is, how should a ballot be interpreted if the ballot, with regard to a particular office, did not exhibit a single unambiguous hole from which the perforated rectangle—the now notorious “chad”—had been completely and successfully removed? On Saturday morning, November 11, the Palm Beach County Canvassing Board (PBCCB) agreed on a “sunshine rule,” under which a vote would be tallied if the impression made on an imperfectly removed chad nonetheless allowed light to pass through the ballot in the proper place.91 Based on that standard, at 2 a.m. on Sunday, November 12, the PBCCB voted two to one that a net gain for Gore of nineteen votes from a sample of four precincts warranted a complete manual recount of the entire county.92 Before proceeding further, County Judge Charles Burton, chair of the PBCCB, nonetheless wanted an official opinion from the Florida Department of State Division of Elections on two issues: First, he wanted to know if vote totals based on the recount were to be certified to the Secretary of State after Tuesday, November 14, at 5 p.m.—whether they would be counted in the certification of statewide 89. 90. 91. 92. Id. (emphasis added). Id. § 102.112(2) (amended 2001). Von Drehle et al., supra note 83. Id. 2001] DISAPPEARING DEMOCRACY 557 results.93 Second, Burton wanted to know if the undercount for Gore detected by the preliminary manual recount over the weekend really triggered a mandatory recount under section 102.166(5) as a matter of law.94 That is, he wanted to know whether the apparent undercounting of Gore votes amounted to what section 102.166 calls “an error in voting tabulation that could affect the outcome of an election,” which would obligate the PBCCB to conduct a countywide hand recount.95 On Monday, November 13, L. Clayton Roberts, the director of the Division of Elections of the Florida Department of State, issued a negative response to both questions. Roberts noted the apparent discrepancy between section 102.111, which seemed to make mandatory the exclusion from statewide totals of any county votes certified after the statutory deadline, and section 102.112, which seemed to render their inclusion or exclusion discretionary.96 He nonetheless deemed any discretion conferred by section 102.112 to be irrelevant to the problem presented. In his judgment, whatever discretion section 102.112 allowed to the Department of State was intended only “[for] unforeseen circumstances not specifically contemplated by the legislature. Such unforeseen circumstances might include a natural disaster such [sic] Hurricane Andrew, where compliance with the law would be impossible. But a close election, regardless of the identity of the candidates, is not such a circumstance.”97 In a separate opinion, Roberts denied that an undercount would trigger a mandatory recount where the undercount was a result of “[t]he inability of a voting systems [sic] to read an improperly marked marksense or improperly punched ballot . . . .”98 Instead, “An ‘error in the vote tabulation’ [that would trigger a mandatory recount] means a counting error in which the vote tabulation system fails to count properly marked marksense or properly punched punchcard ballots.”99 The obvious implications of these opinions were that the undercount for Gore that the PBCCB detected did not require a countywide recount and that such a recount might be pointless because it might not be possible to complete one before the statutory deadline of November 93. 00-10 Fla. Op. Div. of Elec. (2000), reprinted in BUSH V. GORE: THE COURT CASES supra note 67, at 9. 94. Id. 95. 00-13 Fla. Op. Div. of Elec. (2000), reprinted in BUSH V. GORE: THE COURT CASES AND THE COMMENTARY, supra note 67, at 12. 96. 00-10 Fla. Op. Div. of Elec. (2000), reprinted in BUSH V. GORE: THE COURT CASES AND THE COMMENTARY, supra note 67, at 9-10. 97. Id. at 10. 98. 00-11 Fla. Op. Div. of Elec. (2000), reprinted in BUSH V. GORE: THE COURT CASES AND THE COMMENTARY, supra note 67, at 11. 99. Id. AND THE COMMENTARY, 558 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 14, 2000, at 5 p.m., which, according to Clayton, the Department had no authority to waive. The PBCCB responded to the Clayton opinion on the mandatory recount issue by seeking an advisory opinion on the same question from Florida Attorney General Robert A. Butterworth, who like Secretary Harris was a campaign co-chair for one of the presidential candidates—in Butterworth’s case, for Gore.100 Butterworth disagreed vehemently with Clayton’s conclusion on this issue. He noted first that the error that, under section 102.166, triggers a mandatory countywide recount, is not an error in “the vote tabulation system,” a phrase used elsewhere in the statute, but “an error in the vote tabulation” or enumeration itself.101 It was Butterworth’s view that, when referring to a tabulation system rather than to a vote count, the Florida Legislature consistently used the terms “vote tabulation system” and “automatic tabulating equipment.”102 Having rejected on plain language grounds the notion that “an error in vote tabulation” meant only an error in the vote tabulation system, Butterworth argued: “[An] error in vote tabulation might be caused by a mechanical malfunction in the operation of the vote counting system, but the error might also result from the failure of a properly functioning mechanical system to discern the choices of the voters as revealed by the ballots.”103 Butterworth buttressed his conclusion by observing that section 102.166 dictates recount procedures that include a process for discerning voter intent from visually inspected ballots.104 Moreover, Clayton’s distinction between failures to count “properly” marked or punched ballots and failures to count ballots that were “improperly marked” could not be sustained because Florida election law does not specify how a ballot is to be punched or marked; rather, it contemplates that the discernability of “voter intent” shall be the sole legal standard that renders a ballot countable.105 Butterworth demonstrated that a substantial line of Florida Supreme Court decisions dating to early in the twentieth century confirmed his reading.106 In short, the PBCCB’s discovery of a substantial number of Gore votes that were plainly intended by the voters, but not read by the mechanical system meant that an “error in vote tabulation” had oc- 100. David Von Drehle et al., A “Queen” Kept Clock Running: Harris and Allies Stalled Recounts, and Then Ran Into Fla. High Court, WASH. POST, Jan. 30, 2001, at A01. 101. 00-65 Fla. Op. Att’y Gen. (Nov. 14, 2000), reprinted in BUSH V. GORE: THE COURT CASES AND THE COMMENTARY, supra note 67, at 14-15. The opinion is also available at http://legal1.firn.edu/ago.nsf/$defaultview. 102. Id. 103. Id. 104. Id. at 16. 105. Id. 106. Id. at 17. 2001] DISAPPEARING DEMOCRACY 559 curred, triggering the Board’s obligation to conduct a comprehensive recount. In the meantime, the Volusia County Canvassing Board had gone to Florida Circuit Court to seek a temporary injunction against Secretary Harris and the Department of State that would require them to consider—even after 5 p.m. on November 14—the certified results from counties that could not complete by that deadline the countywide recounts they were legally required to hold. On November 14, Circuit Judge Terry P. Lewis granted the requested relief, in part.107 According to Judge Lewis, Secretary Harris erred in insisting that only an “Act of God” would permit her legally to consider the inclusion in state totals of county returns that were submitted after the seven-day deadline. In Judge Lewis’s view, Harris’s insistence on early finality ignored the legislature’s countervailing interest in vote count accuracy.108 Florida law appeared to anticipate a number of situations in which a manual recount would be called for even if it could not be completed within seven days. “It is unlikely,” he wrote, “that the Legislature would give the right to protest returns, but make it meaningless because it could not be acted upon . . . .”109 To give effect to the legislature’s interests in both finality and accuracy, as well as the language of the statute, Judge Lewis concluded that counties were required to report existing vote totals by 5 p.m. on the seventh day following an election.110 Counties were also entitled, however, to decide to file late returns, which would be included or not within the state count, as the Secretary of State would be entitled to determine within her discretion: [T]he Secretary of State has the authority to exercise her discretion in reviewing that decision [to submit late returns], considering all attendant facts and circumstances, and decide whether to include or to ignore the late filed returns in certifying the election results and declaring the winner. . . . [T]he Secretary cannot decide ahead of time what late returns should or should not be ignored . . . .111 Judge Lewis determined, however, that he could not direct Secretary Harris as to how to exercise her discretion.112 He could go no further than indicate that “the exercise of discretion, by its nature, contem- 107. McDermott v. Harris, No. CV-00-2700, 2000 WL 1693713 (Fla. 2d Cir. Ct. Nov. 14, 2000), rev’d sub nom. Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d 1220 (Fla. 2000), vacated sub nom. Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000). 108. See id. at *1. 109. Id. at *2. 110. Id. at *3. 111. Id. 112. Id. 560 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 plates a decision based upon a weighing and [a] consideration of all attendant facts and circumstances.”113 Secretary Harris’s response to this decision was audacious. One might have thought that among “all attendant facts and circumstances” worthy of consideration in determining whether to include late filed returns would be the substance of those returns. Yet, the afternoon of Judge Lewis’s order, Secretary Harris responded by instructing all counties to report no later than 2 p.m. the following day a statement of those facts and circumstances that, in the views of the respective counties, would justify her inclusion of their late filed returns.114 She wrote an additional letter the following day, specifying the criteria under which she intended to exercise her discretion: Facts & Circumstances Warranting Waiver of Statutory Deadline 1. Where there is proof of voter fraud that affects the outcome of the election. 2. Where there has been a substantial noncompliance with statutory election procedures, and reasonable doubt exists as to whether the certified results expressed the will of the voters. 3. Where election officials have made a good faith effort to comply with the statutory deadline and are prevented from timely complying with their duties as a result of an act of God, or extenuating circumstances beyond their control, by way of example, an electrical power outage, a malfunction of the transmitting equipment, or a mechanical malfunction of the voting tabulation system. Facts & Circumstances Not Warranting Waiver of Statutory Deadline 1. Where there has been substantial compliance with statutory election procedures and the contested results relate to voter error, and there exists a reasonable expectation that the certified results expressed the will of the voters. 2. Where there exists a ballot that may be confusing because of the alignment and location of the candidates’ names, but is otherwise in substantial compliance with the election laws. 3. Where there is nothing “more than a mere possibility that the outcome of the election would have been effected.”115 Following this advice, four counties—Broward, Miami-Dade, Palm Beach, and Volusia—all filed letters expressing an intention to submit late-filed returns.116 That afternoon, Secretary Harris announced her rejection of each of these requests.117 When Volusia 113. Id. 114. Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d 1220, 1226 (Fla. 2000), vacated sub nom. Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000). 115. Id. at 1226 n.5 (citations omitted). 116. Id. at 1226. 117. Id. at 1227. 2001] DISAPPEARING DEMOCRACY 561 County protested this decision to Judge Lewis, he determined, on Friday, November 17, that Harris had “exercised her reasoned judgment to determine what relevant factors and criteria should be considered, applied them to the facts and circumstances pertinent to the individual counties involved, and made her decision.”118 This, he concluded, fulfilled his prior order. In the meantime, the U.S. District Court for the Southern District of Florida had refused the Bush request to block any manual recounts, a decision he appealed on November 15.119 On November 16 and 17, the Florida Supreme Court denied a request by Harris to block the recounts,120 agreed to resolve the dispute between Harris and Attorney General Butterworth regarding the hand counts’ legal basis,121 and stayed any certification of the election by Secretary Harris while it heard the case.122 On Friday, November 18, the final day for the counting of overseas ballots, new statewide totals were announced, expanding the Bush lead to 930.123 With these events, the ground had been laid for the first of four utterly critical judicial events—a decision on the merits by the Florida Supreme Court about the legality of hand recounts prior to a certification of a statewide winner in the presidential race. After hearing oral arguments on Monday, November 20, the Court issued on November 21 its unanimous ruling that (a) the hand counts could continue, and (b) that Secretary Harris was obliged to include in the state totals any returns submitted by the counties involved by November 26, 2000.124 Now, to get a full perspective on the implications of the Due Process Clause for the Florida vote, let us imagine a course of events that did not occur. Imagine that the Florida Supreme Court had decided differently and upheld Secretary Harris’s opinion that candidates could secure manual recounts only if the tabulation errors alleged were the result of fraud, noncompliance with statutory procedures, or an Act of God. Imagine further that, in the wake of that opinion, Al Gore had sued to demand a manual recount as an imperative of due 118. McDermott v. Harris, No. CV-2700, 2000 WL 1714590, at *1 (Fla. 2d Cir. Ct. Nov. 17, 2000), rev’d sub nom. Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d 1220 (Fla. 2000), vacated sub nom. Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000). 119. BUSH V. GORE: THE COURT CASES AND THE COMMENTARY, supra note 67, at xii. 120. Palm Beach County Canvassing Bd. v. Harris, No. SC00-2346, 2000 WL 1708520 (Fla. Nov 16, 2000). 121. Palm Beach County Canvassing Bd. v. Harris, No. SC00-2346, SC00-2348, SC002349, 2000 WL 1716481 (Fla. Nov 17, 2000). 122. Palm Beach County Canvassing Bd. v. Harris, No. SC00-2346, SC00-2348, SC002349, 2000 WL 1716480 (Fla. Nov. 17, 2000). 123. BUSH V. GORE: THE COURT CASES AND THE COMMENTARY, supra note 67, at xii. 124. Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d 1220, 1240 (Fla. 2000), vacated sub nom. Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000). 562 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 process in Broward, Miami-Dade, Palm Beach, and Volusia Counties. This is my hypothetical case of Gore v. Harris.125 The law to be applied in my hypothetical case is the now wellroutinized due process doctrine that the Court crystallized in the 1970s into an oft-cited, even if academically criticized, two-step analysis. The first question presented in procedural due process cases is whether the individual interest at stake qualifies as “liberty” or “property” entitling the interest holder to due process protection.126 The second is whether, in light of the competing individual and governmental interests at stake, additional decisionmaking procedures promise a sufficient likelihood of increased accuracy as to warrant their mandatory imposition.127 Gore would certainly have had no trouble establishing that the Due Process Clause protects the right to vote as a form of “liberty” or “property.” At one time, this might have seemed a difficult question because an obvious implication of the debates leading to the Fourteenth Amendment and of the subsequent adoption of the Fifteenth Amendment is that Congress, in 1866, did not anticipate that Section 1 of the Fourteenth Amendment would protect voting rights. Such was Justice Harlan’s position in his dissents in Reynolds v. Sims128 and Carrington v. Rash.129 Beginning with the reapportionment cases, however, the Court has taken a consistent position that Section 1 does protect voting rights, resting chiefly on the “fundamental rights” strand of equal 125. In so naming the case, I have elided—as did the Florida Courts—the question of a candidate’s standing to vindicate the rights of voters. In my hypothetical case, as in the genuine litigation, there was no question that, in their respective suits, candidates Bush and Gore had each alleged sufficiently concrete injury to meet the federal constitutional requirements for standing. See Peter M. Shane, Returning Separation-of-Powers Analysis to Its Normative Roots: The Constitutionality of Qui Tam Actions and Other Private Suits to Enforce Civil Fines, 30 ENVTL. L. REP. 11,081, 11,085-89 (2000) (reviewing the Supreme Court’s standing decisions with regard to the injury requirement). The issue that a federal court might have stopped to consider is whether it would nonetheless have been appropriate to invoke the prudential rule against so-called third-party standing, which ordinarily bars even injured parties from seeking federal judicial intervention where the cause of their alleged injury is a violation not of their own rights but of another person’s. See Singleton v. Wulff, 428 U.S. 106 (1976). In this case, at least three factors argue persuasively for permitting the candidates to litigate their supporters’ rights: the close interrelationship of the candidate’s and voters’ interests, the certainty that the candidates would be vigorous proponents of the voters’ rights, and the possibility that direct voter suits might be deterred by uncertainty among the voters as to which of them were specifically affected by the state tabulation practices in dispute. Cf. Touchston v. McDermott, 234 F.3d 1133, 1151 (11th Cir. 2000) (Tjoflat, J., dissenting). 126. Bd. of Regents v. Roth, 408 U.S. 564, 569 (1972). 127. Mathews v. Eldridge, 424 U.S. 319, 334-35 (1976). 128. 377 U.S. 533, 589 (1964) (Harlan, J., dissenting). 129. 380 U.S. 89, 97 (1965) (Harlan, J., dissenting). 2001] DISAPPEARING DEMOCRACY 563 protection analysis.130 It is hardly a leap from acknowledging the status of voting rights as fundamental for equal protection purposes to recognizing voting as a protected form of liberty under the Due Process Clause. The question is made even easier, however, because of the Court’s current approach to the identification of protectible property interests under the Fourteenth Amendment. Since the early 1970s, it has been the Court’s consistent position that an individual’s benefits or interests that are created by state law are protected “property” if they are subject to “rules or understandings that secure [such] benefits and that support claims of entitlement to those benefits.”131 That is, there must be “rules or mutually explicit understandings that support [a] claim of entitlement to the benefit and that [the beneficiary] may invoke at a hearing” involving the benefit in question.132 Well-known examples of such protected interests include many government jobs,133 public assistance payments,134 government disability insurance,135 public education,136 and state licenses.137 The right to vote in Florida—and the implicit concomitant right of having one’s vote counted—is supported by explicit rules under Florida law. Section 97.041, Florida Statutes, prescribes the state’s qualifications to register and vote in categorical and nondiscretionary terms.138 There can be no doubt that Section 97.041 creates a protectible property interest for Fourteenth Amendment purposes. 130. See, e.g., Kramer v. Union Free Sch. Dist. No. 15, 395 U.S. 621, 626 (1969); Harper v. Va. State Bd. of Elections, 383 U.S. 663, 670 (1966). 131. Roth, 408 U.S. at 577. 132. Perry v. Sindermann, 408 U.S. 593, 601 (1972). 133. Cleveland Bd. of Educ. v. Loudermill, 470 U.S. 352 (1985). 134. Goldberg v. Kelly, 397 U.S. 254 (1970). 135. Mathews v. Eldridge, 424 U.S. 319 (1976). 136. Goss v. Lopez, 419 U.S. 565 (1975). 137. Barry v. Barchi, 443 U.S. 55 (1979). 138. Section 97.041 reads: (1)(a) A person may become a registered voter only if that person: 1. Is at least 18 years of age; 2. Is a citizen of the United States; 3. Is a legal resident of the State of Florida; 4. Is a legal resident of the county in which that person seeks to be registered; and 5. Registers pursuant to the Florida Election Code. (b) A person who is otherwise qualified may preregister on or after that person’s 17th birthday and may vote in any election occurring on or after that person’s 18th birthday. (2) The following persons, who might be otherwise qualified, are not entitled to register or vote: (a) A person who has been adjudicated mentally incapacitated with respect to voting in this or any other state and who has not had his or her right to vote restored pursuant to law. (b) A person who has been convicted of any felony by any court of record and who has not had his or her right to vote restored pursuant to law. (3) A person who is not registered may not vote. FLA. STAT. § 97.041 (2000). 564 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 The task a Gore suit would thus have presented, as the Supreme Court has put it, would be “identif[ying] . . . the specific dictates of due process”139 with regard to having one’s vote counted. The hypothetical suit by Gore voters demanding a manual recount in selected counties would be based on the following claim regarding the adjudicatory process of vote tabulation. Florida has in place an array of mechanical and electromechanical systems for discerning voter intent. As interpreted by Secretary Harris, these mechanical and electromechanical systems provide the exclusive means for discerning voter intent absent fraud, noncompliance with statutory procedures, or an Act of God. Due process, according to Gore, would demand a manual review of the ballots in an additional category of circumstances: namely, whenever a candidate could establish a prima facie likelihood that some failure had occurred, for whatever reason, for a county’s automatic tabulation system to count legally valid ballots of sufficient number to have swayed the electoral outcome. The Court’s due process decisions have articulated a balancing rubric to assess what is reasonable in just this sort of procedural due process contest. As first crystallized in Mathews v. Eldridge: [I]dentification of the specific dictates of due process generally requires consideration of three distinct factors: First, the private interest that will be affected by the official action; second, the risk of an erroneous deprivation of such interest through the procedures used, and the probable value, if any, of additional or substitute procedural safeguards; and finally, the Government’s interest, including the function involved and the fiscal and administrative burdens that the additional or substitute procedural requirement would entail.140 In other words, a due process challenge of the kind I am hypothesizing requires a court to compare the decisionmaking procedures sought by a plaintiff with the presumably more summary procedures offered already by the state authority being challenged. The court is to award the additional procedures sought (a) only if existing procedures run the risk of erroneous results and (b) only if there is sufficient value to the additional procedures, while (c) taking into account the relevant—and often competing—interests of the individual plaintiff and of the government.141 Despite widespread scholarly theoretical dissatisfaction with this particular three-part test,142 its application to my hypothetical Gore 139. Mathews, 424 U.S. at 335. 140. Id. 141. See id. 142. The seminal critique is Jerry L. Mashaw, The Supreme Court’s Due Process Calculus for Administrative Adjudication in Mathews v. Eldridge: Three Factors in Search of a Theory of Value, 44 U. CHI. L. REV. 28 (1976). 2001] DISAPPEARING DEMOCRACY 565 suit is straightforward. In the wake of Florida’s electoral experience, there is no serious doubt that the state’s mechanical and electromechanical systems run the risk of substantial error. Punch-card voting, in particular, appears to pose the danger of significantly inaccurate tabulation. Florida law essentially concedes this risk because it provides expressly for manual recounting as a check on the error of mechanical and electromechanical systems. On this point, there is truly no contest. That conclusion, however, simply moves us to the next step of the inquiry. What would be the value of additional procedures, in light of the competing interests involved of Florida’s voters and of the state government? As a general proposition, it can hardly be gainsaid that a manual inspection is valuable in addressing the erroneous tabulation of ballots by mechanical and electromechanical systems. Again, this proposition is a major premise of the Florida law that applies to election protests—and is consistent with the laws of at least twentyseven other states that expressly contemplate the availability of manual counting.143 Especially in a state such as Florida, in which 143. Laws in the following twenty-eight states contemplate manual recounts expressly or by necessary implication: Alaska, ALASKA STAT. § 17-15-7(1) (Michie 2000) (court to hold hearing “to determine the procedures to be used for the examination” of ballots when “person or candidate seek[s] to examine the ballots”); California, CAL. ELEC. CODE § 15270 (West 2001) (providing for manual vote counts); Colorado, COLO. REV. STAT. § 1-10.5-102(3) (2000) (canvass board to select “a test number of ballots on which to conduct a machine count and hand count”; recount to be “conducted in the same manner as the original ballot count” if machine and hand counts prove “identical”); Florida, FLA. STAT. § 102.166 (2001) (authorizing manual recounts); Georgia, GA. CODE ANN. § 21-2-495(a) (2000) (providing for recounts where paper ballots or vote recorders have been used); Indiana, IND. CODE ANN. § 3-12-11-17.5 (Michie 2000) (permitting petitions for manual recounts); Kansas, KAN. STAT. ANN. § 25-4412(c) (2000) (ballots to be manually counted if ballot card damaged or defective and cannot be counted properly by automatic tabulating equipment); Maine, ME. REV. STAT. ANN. tit. 21-A, § 737-A (West 1999) (permitting candidate’s representatives or counsel actually to conduct the recount); Massachusetts, MASS. ANN. LAWS ch. 54, § 135 (Law Co-op. 2001) (providing for manual recount procedure); Michigan, MICH. STAT. ANN. § 6.1871 (Michie 2000) (authorizing manual recounts); Minnesota, MINN. STAT. § 204C.21 (2000) (procedure for manual counts); Missouri, MO. REV. STAT. § 115.585 (2000) (providing for manual inspection of votes cast on paper ballots); Montana, MONT. CODE ANN. § 1316-412 (2000) (providing for manual inspection of votes cast on paper ballots); Nebraska, NEB. REV. STAT. § 32-1114 (2000) (procedure to open and inspect ballots); Nevada, NEV. REV. STAT. § 293.423 (2001) (providing for opening and recount of ballots); New Hampshire, N.H. REV. STAT. ANN. § 660:5 (2000) (providing for manual recounts); New Jersey, N.J. STAT. ANN. § 19:53A-8 (West 2001) (providing for manual recounts when a count by tabulating equipment “becomes impracticable”); North Dakota, N.D. CENT. CODE § 16.116-01 (2000) (county auditor to review all ballots for recount); Ohio, OHIO REV. CODE ANN. § 3515.04 (Anderson 2001) (procedure for manual recounts); Oklahoma, OKLA. STAT. tit. 26, § 8-114 (2000) (procedure for manual recounts); Pennsylvania, PA. STAT. ANN. tit. 25, § 3261 (West 2001) (providing for opening of ballot boxes and recounting of votes by persons designated by specified courts or judges); Rhode Island, R.I. GEN. LAWS § 17-19-37.1 (2001) (authorizing candidate requests for manual recounts); Texas, TEX. ELEC. CODE ANN. § 214.049 (Vernon 2000) (procedure for manual recounts); Vermont, VT. STAT. ANN. tit. 17, § 2602 (2001) (providing detailed manual recount procedures); Virginia, VA. CODE ANN. § 24.2-802 (Michie 2001) (authorizing manual inspection of all ballots); Washington, WASH. 566 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 any ballot is lawful that expresses the intent of the voter in a discernible way, a manual count is the only means of discovering intentions left unread because of either machine or voter error. Whether or not the burden of such additional procedures is warranted, however, depends upon the Supreme Court’s mandatory interest balancing. With regard to the demand for recounts I have hypothesized, that balance overwhelmingly favors Gore. Insofar as Gore’s rights were inextricably intertwined with the rights of those voters who supported him, Gore would have been invoking an individual interest that the Supreme Court has characterized repeatedly as being of the highest order. As early as 1886, the Supreme Court described voting “as a fundamental political right, because preservative of all rights.”144 The Court’s modern equal protection voting jurisprudence is founded on this premise. What, then, would be the state’s countervailing interests? A state’s usual interests in avoiding increased procedural formality are cost-savings, time-savings, and achieving finality in administrative decisionmaking. In this case, the financial interest, although genuine, is hardly forbidding. The Court has never permitted costsavings, by itself, to be a deciding factor in a case involving substantial individual rights. The time factor, of course, is potentially compelling in an electoral context. With regard to the appointment of presidential electors, states surely have a critical interest in resolving disputes early enough to permit actual balloting in the Electoral College. Bush v. Gore was careless in its analyses of key deadlines, but, for purposes of my hypothetical, its mistakes are irrelevant. There is no real reason to doubt that manual counts in disputed counties could have occurred in timely fashion had they commenced promptly after election day. In assessing the state’s interest in promptness, moreover, as well as its closely related interest in finality, it is worth recalling that, in any event, the state could not have certified final victors in the November 7 poll until November 17 because of the time legally required for the counting of overseas ballots.145 Had hand counts proceeded expeditiously, it is probable they REV. CODE § 29.64.015 (2001) (providing for manual recount unless candidates agree on alternative procedure); West Virginia, W. VA. CODE § 3-6-9 (2001) (manual inspection procedures); Wisconsin, WIS. STAT. § 9.01(1)(b) (2000) (providing for manual inspection of ballots). These do not include states in which manual recounting would appear to be within a state official’s broadly granted discretion, for example, Oregon, OR. REV. STAT. § 258.150 (1999) (Secretary of State to determine most appropriate recount method), or in which manual recounting is authorized only upon a demonstration that automatic vote count technologies had failed, for example, Wyoming, WYO. STAT. ANN. § 22-11-107 (Michie 2001) (recount made by alternate method if automatic tabulating equipment fails). 144. Yick Wo v. Hopkins, 118 U.S. 356, 370 (1886). 145. See FLA. ADMIN. CODE ANN. r. 1S-2.013(7). With respect to the presidential preference primary and the general election, any absentee ballot cast for a federal office by an overseas elector which is post- 2001] DISAPPEARING DEMOCRACY 567 could have been completed by November 17. In short, the state’s interests in avoiding a manual count appear to be notably less singular and fundamental than the interest of each Florida voter in having his or her intentions accurately ascertained. But even this does not tell the entire story. Any accounting for the state’s interests must take notice also of those state interests that actually favor a manual recount. It should hardly have required our current fiasco to remind us that any government stands to pay a dire price in lost legitimacy should its citizenry begin to lose faith that electoral results actually reflect the expressed intentions of the voting public. That risk would plainly loom largest in just that category of cases in which Gore would be insisting on the constitutionally mandatory status of manual recounts, that is, cases in which a candidate or voter had established a prima facie likelihood that a county’s automatic tabulation system had failed to count legally valid ballots of sufficient number to have potentially swayed the electoral outcome. These insights counsel powerfully for a state’s adoption of procedures that provide citizens a reasonable guarantee that the electoral connection between the people and their officials is a genuine one. Where, as in this matter, a state’s own interests, fully accounted for, militate in favor of additional procedural protections, the Supreme Court has deemed that fact relevant to the implementation of its due process calculus.146 Based on this analysis, the proper resolution to the hypothetical case of Gore v. Harris is plain: had the Florida Supreme Court upheld Secretary Harris’s view of the law and had Vice President Gore demanded the recounts he sought under the Due Process Clause, he should have won his case. He had shown the challenged tabulation to be significantly in error. There was no doubt that a manual count had substantial potential to remedy the problems presented. A manual count could have proceeded without compromising Florida’s inpostmarked or signed and dated no later than the date of the federal election shall be counted if received no later than 10 days from the date of the federal election as long as such absentee ballot is otherwise proper. Id. 146. See Goldberg v. Kelly, 397 U.S. 254, 265 (1970): [W]elfare guards against the societal malaise that may flow from a widespread sense of unjustified frustration and insecurity. Public assistance, then, is not mere charity, but a means to ‘promote the general Welfare, and secure the Blessings of Liberty to ourselves and our Posterity.’ The same governmental interests that counsel the provision of welfare, counsel as well its uninterrupted provision to those eligible to receive it; pre-termination evidentiary hearings are indispensable to that end. Cf. Goss v. Lopez, 419 U.S. 565, 583 (1975) (“In holding as we do, we do not believe that we have imposed procedures on school disciplinarians which are inappropriate in a classroom setting. Instead we have imposed requirements which are, if anything, less than a fair-minded school principal would impose upon himself in order to avoid unfair suspensions.”). 568 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 terest in a timely, final certification. The financial interest of the state in summary procedures would not have justified the risk that obviously existed of disenfranchising a significant number of those Floridians who cast intelligible ballots on November 7. Moreover, Florida’s interest in maintaining the confidence of the electorate in the accuracy and fairness of the state’s electoral system would actually have supported a Gore suit. Of course, Al Gore did not seek to require manual counts in the four disputed counties under the Due Process Clause. Florida law promised him the very result he sought. The Florida Legislature— without constitutional prodding by the federal courts—had already designed an electoral procedure that effectively tracks the requirements of due process. Florida statutes provided genuine opportunities for the administrative correction of tabulation error, including the possibility of manual recounts where necessary. They made such recounts mandatory upon a prima facie showing from sample precincts that “indicates an error in the vote tabulation which could affect the outcome of the election.”147 The administrative process is subject to independent judicial review. Because the Florida Supreme Court upheld Gore’s view of the law, no other federal litigation may have seemed necessary. But reviewing the hypothetical case of Gore v. Harris provides a helpful demonstration of three things. The first is the amenability of the Florida electoral controversy to conventional due process analysis. The second is the importance of procedural due process in the tabulation of votes as a bulwark of democratic government. The third is that Gore was entitled to mandatory recounts in at least four counties not only because of Florida statutes, but also because of the Fourteenth Amendment. The Supreme Court effectively undercut democracy and deprived Gore voters of their constitutional rights. C. Applying Due Process II: The Real Case of Bush v. Gore That brings us to the issue actually addressed by the per curiam opinion in Bush v. Gore: the threatened lack of uniformity in different counties’ implementation of their manual recounts. A complex series of events had transpired since the Florida Supreme Court opinion rejecting Secretary Harris’s view of Florida’s election protest law. On November 26, Harris certified Bush the winner in Florida, including in her totals the results of recounts in Broward and Volusia Counties. She refused to include the Palm Beach recount totals, which had reached her office about two hours after the court-imposed deadline. There was also no comprehensive recount from Miami- 147. FLA. STAT. § 102.166(5) (2000) (amended 2001). 2001] DISAPPEARING DEMOCRACY 569 Dade County, where the county canvassing board, through a torturous process, had finally decided that a countywide recount was appropriate, but that the count could not be completed by the courtimposed date. Harris refused to make adjustments based on the partial recount of sample precincts.148 Following Harris’s certification, Gore filed a contest of the election in Leon County, which, under Florida law, is the proper venue for election contests involving more than a single county.149 The case was tried before Leon County Circuit Judge N. Sanders Sauls while the U.S. Supreme Court had under advisement Bush’s challenge to the Florida Supreme Court decision that had allowed the precertification recounts to continue. On December 4, both the Supreme Court and Judge Sauls acted. The Supreme Court refused to address Bush’s supposed federal issues on the merits, but nonetheless vacated and remanded the Florida Supreme Court’s November 21 opinion for clarification with regard to two questions of dubious legal relevance.150 Judge Sauls rejected Gore’s contest on the ground that he had failed to prove a “reasonable probability” that the election would have turned out differently if not for the problems he identified in counting ballots.151 On December 6, the Eleventh Circuit rejected Bush’s request for an injunction against manual recounts.152 The next day, the Florida Supreme Court heard oral arguments in the Gore appeal from Judge Sauls’s ruling. On December 8, the Florida Supreme Court decided, four to three, not only to grant Gore the relief he requested, but also to direct the circuit court to order the relevant officials “in all counties that have not conducted a manual recount or tabulation of the undervotes in this election to do so forthwith . . . .”153 Judge Sauls recused himself from implementing the court’s mandate154 and was replaced by Leon County Circuit Court Judge Terry Lewis. Despite Judge Lewis’s immediate steps to implement the mandate, the U.S. Supreme Court stayed the recount by a five to four vote on December 9.155 After oral arguments on De- 148. Gore v. Harris, 772 So. 2d 1243, 1248 (Fla. 2000), rev’d sub nom. Bush v. Gore, 531 U.S. 98 (2000); Tim Collie & Mark Hollis, Bush Claims Win, Gore Fights Count, SUN-SENTINEL (Ft. Lauderdale, Fla.), Nov. 27, 2000, at 1A, available at 2000 WL 28992947. 149. FLA. STAT. § 102.1685 (2000). 150. Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000). 151. Gore v. Harris, No. 00-2808, 2000 WL 1770257 (Fla. 2d Cir. Ct. Dec. 4, 2000), rev’d 772 So. 2d 1243 (Fla. 2000), rev’d sub nom. Bush v. Gore, 531 U.S. 98 (2000), reprinted in BUSH V. GORE: THE COURT CASES AND THE COMMENTARY, supra note 67, at 53, 55. 152. Siegel v. LePore, 234 F.3d 1163 (11th Cir. 2000); Touchston v. McDermott, 234 F.3d 1133 (11th Cir. 2000), cert. denied 531 U.S. 1061 (2001). 153. Gore v. Harris, 772 So. 2d at 1262. 154. Gore v. Harris, No. 00-2808, 2000 WL 1801773 (Fla. 2d Cir. Ct. Dec. 8, 2000). 155. Bush v. Gore, 531 U.S. 1046 (2000). 570 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 cember 10, the Supreme Court decided Bush v. Gore on December 12.156 What the Bush suit effectively asked the Supreme Court to do was to judge the fairness of the system of mass adjudication that the Florida Supreme Court had put into place in order to determine the resolution of the Gore election contest. That process, in essence, was to consist of seven elements: 1. All counties that had not conducted a manual recount or tabulation of the undervotes since Election Day were to conduct such tabulations in the respective counties.157 2. Miami-Dade would tabulate by hand approximately 9,000 ballots, which its counting machine had registered as non-votes, but which not been manually reviewed as part of its earlier partial recount.158 3. Ballots would be counted (and thus treated as legal) throughout the state if they revealed a “clear intent of the voter.”159 4. The recounting would be done by as many two person teams in each county as would be required to complete the respective recounts in timely fashion.160 5. If questions arose with regard to any ballot, they would be reviewed in each county by two circuit judges who, if they could not agree, would refer the matter for determination to Judge Lewis, who would preside over the statewide recount.161 6. Both the Gore and Bush campaigns would be entitled to have observers watching every counting team. Objections to the treatment of any ballot could be registered in writing by either observer and filed with the Clerk of the Leon County Circuit Court for Judge Lewis’s review.162 7. A new statewide total would be certified to include the results from all recounts that occurred prior to November 26, the results from Palm Beach, the results from Miami-Dade’s pre-November 26 partial recount, results from the recounts of the remaining MiamiDade undervotes plus recounted results in all remaining counties, and any adjustment Judge Lewis might order based on his review of any objections forwarded to him.163 156. 531 U.S. 98 (2000). 157. Gore v. Harris, 772 So. 2d at 1252-55. Statewide totals would be adjusted to include the additional votes in Palm Beach County for Gore that had been submitted within hours of the court’s earlier deadline, plus 168 votes from Miami-Dade County that had resulted from that county’s earlier partial recount. Id. at 1260. 158. Id. at 1258-59. 159. Id. at 1257. 160. See Gore v. Harris, No. 00-2808, at 2 (Fla. 2d Cir. Ct. Dec. 9, 2000) (Florida Recount General Instructions), available at http://election2000.stanford.edu/CV-00-280861.pdf. 161. Id. at 5. 162. Id. at 2-3. 163. Gore v. Harris, 772 So. 2d at 1262. 2001] DISAPPEARING DEMOCRACY 571 The majority per curiam opinion, in turn, identified the five following aspects of the foregoing process as equal protection “problems”: 1. “[T]he absence of specific standards to ensure . . . equal application” of the “clear intent of the voter” standard;164 2. The inclusion in the final statewide total recount results from Broward, Palm Beach, and Miami-Dade that were based on a comprehensive review of all ballots and not limited, as would be the recounts in other counties, to a review of so-called “undervotes”;165 3. The possibility that the final statewide totals would be allowed to include only partial recount results from some counties, but complete totals from others;166 4. The failure of the Florida courts to specify more precisely who would recount the ballots, while relying for dispute resolution on ad hoc teams of state judges who lacked training in handling and interpreting ballots;167 and 5. Limiting the objection process to written objections to be adjudicated after the recount, rather than during it.168 In his dissent, Justice Souter agreed that the first problem resulted in “wholly arbitrary” differences in different counties’ treatment of ballots, and thus violated equal protection’s fundamental requirement of rationality.169 Justice Breyer stated that “principles of fairness should have counseled the adoption of a uniform standard” because of the election’s “very special circumstances.”170 The “very special circumstances” Breyer identifies are these: [T]he use of different standards could favor one or the other of the candidates since time was, and is, too short to permit the lower courts to iron out significant differences through ordinary judicial review, and . . . the relevant distinction [between strict and lenient standards] was embodied in the order of the State’s highest court . . . .171 Because the majority, however, had called the recount off, he declined to “decide whether, or the extent to which, as a remedial matter, the Constitution would place limits upon the content of the uniform standard.”172 164. 165. 166. 167. 168. 169. 170. 171. 172. Bush v. Gore, 531 U.S. 98, 105-06 (2000). Id. at 107-08. Id. at 108. Id. at 109. Id. Id. at 134 (Souter, J., dissenting). Id. at 146 (Breyer, J., dissenting). Id. at 145-46 (Breyer, J., dissenting). Id. at 146 (Breyer, J., dissenting). 572 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 Of the five infirmities identified by the majority, it is difficult even to articulate how the reliance on unnamed volunteers and untrained judges or the limitation of the candidates’ protest to written objections to be decided by a single judge could amount to an equal protection problem. Truly, if these features compromised the fairness of the recount at all, they would be matters of due process, not equal treatment.173 That leaves as articulable equal protection problems the disparate treatment of ballots in different counties because of how different counties might implement the “clear intent of the voter” standard, the comprehensiveness of recounting in some counties as compared to those that would re-inspect only “undervotes,” and the possible inclusion in the statewide total of some counties’ recounts that were only partial, rather than complete. Except for the fact that these are, of course, differences, it is hard to see how they implicate “equal protection.” The per curiam opinion states: “The recount mechanisms implemented in response to the decisions of the Florida Supreme Court do not satisfy the minimum requirement for non-arbitrary treatment of voters necessary to secure the fundamental right.”174 The majority does no more to explain this conclusion than simply to point to differences in the indicia of voter preference deemed to suffice as a matter of demonstrating voter intent in Miami-Dade, Palm Beach, and Broward Counties. At most, there is a possible implication that the different practices identified—in the case of partial versus total recounts, a hypothetically different practice—might result in a different weighting of the votes in different counties. The mere fact of differences among counties cannot plausibly be enough, however, to establish an equal protection violation. Different practices by different counties in implementing the law probably exist for every state administrative program in the country that states delegate to counties for implementation. There are undoubtedly discernible differences in county-administered adjudicatory programs in every state as varied as driver licensing and the certification of foster homes. With regard to the Florida election itself, there is no doubt that the biggest difference in the likelihood of registering voter intent in one county versus another was the variation in the voting machinery employed. It is a safe bet that differences existed also among different counties in the degree of assistance afforded to confused vot173. The majority’s implicit thought regarding equality might be as follows: Judge Lewis’s remedial scheme would have left voters in counties with untrained but adept judges with some unspecified unjust advantage in comparison with voters in those counties whose untrained judges were also bad at their job. If this distinction were of constitutional magnitude, the Court would conceivably have found a reason to find the entire civil and criminal justice system of the United States unconstitutional. 174. Bush v. Gore, 531 U.S. at 105. 2001] DISAPPEARING DEMOCRACY 573 ers, the inspectors’ willingness to provide new ballots to voters who inadvertently defaced theirs, and in the waiting times to cast votes— all of which would surely produce differences in voter treatment that could have significant electoral impact.175 If the majority perceived it to be a constitutional problem that different vote count procedures could not be justified by any nonarbitrary rationale, then the Court was doubly misled. It was misled because none of the other differences I have articulated in election practice could be justified by any less arbitrary a rationale, and, in fact, all of them are justified on the same rational basis: namely, the state’s preference for empowering locally accountable officials with important local functions. Nor is it at all clear how to apply the equality demands of the Court’s reapportionment jurisprudence. The race for presidential electors is statewide; every vote counts the same toward the final result, namely, 1/N, where N is the total number of legal ballots cast. The only intercounty difference on which the Court could be focusing is a marginal difference in the likelihood between counties that a particular voter’s ballot will be deemed to convey a “clear indication of voter intent.” The differences resulting, however, from the alleged infirmities identified by the Court are overwhelmed by the statistical differences imposed by the differences in voting machinery. And, in any event—even if this sort of difference is of the same constitutional seriousness as a departure from “one person, one vote”—the fact is that the Court does not demand perfect adherence to “one person, one vote” even in districted elections. The Court, for example, once approved an apportionment plan for the Virginia House of Representatives in which there existed among districts a maximum percentage variation from ideal equality of 16.4%.176 It approved a similar plan for Wyoming in which the average deviation from equality was 16% and the maximum deviation 89%.177 The rationales behind these plans were hardly more compelling than the rationality of delegating 175. Cf. Marcia Coyle, Gauging ‘Bush v. Gore’ Fallout, NAT’L L.J., Dec. 25, 2000-Jan. 1, 2001, at A4: In Iowa, where [University of Iowa Law] Prof. [Randall] Bezanson lives, optical scanners are used to tabulate votes, he said. Some counties program the scanners to reject “overvotes,” or double voting in a race. When the ballot is rejected, the machine spits it out and the voter, who is still present, is given a new ballot to correct the error. But other counties, he said, program the scanners simply to reject the overvotes with no chance for correction. “Is the inequality between those counties that follow a different set of programming instructions a violation of the equal protection clause because there’s a systematic difference in the kinds of votes that are counted?” asked the professor. Looking at the Bush-Gore per curiam, he added, “It’s very hard to explain why, as a matter of principle, that isn’t every bit as unconstitutional as the different recounting standards applied in different counties in Florida.” 176. Mahan v. State Bd. of Elections, 410 U.S. 315, 319 (1973). 177. Brown v. Thomson, 462 U.S. 835, 839 (1983). 574 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 to local officials in Florida the discretion to decide how best to determine the “clear intent of the voter.” The resulting departures from equality were dramatically greater than anything that could have resulted from the problems asserted to exist by the majority in Bush v. Gore with regard to the Florida statewide recount. But the unpersuasiveness of the Court’s analysis should not end the inquiry into electoral fairness. The fairness of any recount process in a presidential election is plainly critical to its legitimacy. As it happens, however, employing due process doctrine as the better framework for analysis would have shown that Florida law, not the Supreme Court majority opinion, embodied the sounder view of what fairness should have entailed. Let me start with the two supposed equal protection problems that plausibly implicate the accuracy, as opposed to the consistency, of the recount. These are the failure of the Florida courts to rely exclusively on officials trained in handling and interpreting ballots, and Judge Lewis’s determination to limit the objection process to written objections to be adjudicated after the recount rather than during it. Applying to these problems the Mathews v. Eldridge due process calculus, it is obvious that neither is a problem of constitutional magnitude. It is speculative at best that some measure of training in handling and interpreting ballots would lead to more reliable results than depending on volunteers under constant vigilance by representatives of the two parties, circuit judges whose good faith and professionalism could surely be presumed, and the watchful eye of Judge Lewis. As for the process of relying on written objections, rather than contemporaneous oral objections, it is hard to see why this would have any impact at all on the accuracy of ultimate determinations. In any event, neither practice would seem to threaten any significant prospect of disenfranchising Florida voters. And, the state’s rationale for these practices is a strong one, namely, finishing the electoral contest in time for Florida’s electors to vote. The due process objections to the other alleged infirmities are of a different order. They do not actually address the putative accuracy of the court-ordered recount. They are better viewed as substantive due process objections to the alleged irrationality of the practices at issue. Two of these objections can be dealt with all but summarily. First, the Court objected that “[t]he Florida Supreme Court’s decision thus gives no assurance that the recounts included in a final certification must be complete.”178 It might equally be said, however, that nothing in the Florida Supreme Court’s decision suggested that anything else would be the case. The Court’s mandatory inclusion in 178. Bush v. Gore, 531 U.S. at 108. 2001] DISAPPEARING DEMOCRACY 575 the statewide certified totals of the precertification partial recount in Miami-Dade was simply consistent with its decision to include the results from all the precertification recounts. It was emphatically not the case, however, that the Miami-Dade partial recount was the end of the matter for Miami-Dade County. With regard to the remaining undervotes in Miami-Dade, Miami-Dade was to complete its recount in the same manner as every other county: by reviewing all the undervotes. There would, in other words, be a complete and not a partial total in Miami-Dade. A second objection was that the Broward, Palm Beach, and partial Miami-Dade totals were included based on a comprehensive review of the ballots in question, while the recounts under Judge Lewis’s supervision would be limited to undervotes in all other counties not recounted prior to certification.179 There was, however, an obvious rationale for this difference. The counties that completed comprehensive recounts prior to certification were directed to do so by the statutory provisions governing precertification election protests.180 The Florida Supreme Court recognized, however, that counties subject to the court’s December 8 order could not, as part of the postcertification contest process, produce comprehensive recounts quickly enough to take whatever advantage the state might wish with regard to the safe harbor provision in federal law for the certification of state electors—a provision that required the Florida Supreme Court to complete the adjudicatory process by December 12. Given the importance that the United States Supreme Court—however misguidedly—urged the Florida Supreme Court to attach to the federal safe harbor provision,181 it would be implausible to dismiss the Florida court’s attentiveness to that statute as unconstitutionally irrational. That leaves only the variability within and among counties in the implementation of the “clear intent of the vote” standard as a possible source of objection. Examined in context, however, this variability is hardly a defect of constitutional magnitude. As the majority notes, all relevant officials were aware of a single general standard governing the recount procedure; that is the “clear intent of the voter.” The majority conceded: “This is unobjectionable as an abstract proposition and a starting principle,” but added, “[t]he problem inheres in the absence of specific standards to ensure its equal application.”182 This thought was perhaps more helpfully summarized by Justice Breyer: 179. 180. 181. 182. Id. at 107-08. FLA. STAT. § 102.166(5) (2000) (amended 2001). Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70, 77 (2000). Bush v. Gore, 531 U.S. at 106. 576 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 The majority concludes that the Equal Protection Clause requires that a manual recount be governed not only by the uniform general standard of the “clear intent of the voter,” but also by uniform subsidiary standards (for example, a uniform determination whether indented, but not perforated, “undervotes” should count).183 The problem, that is, lay not with the absence of a uniform standard, but with the absence of sufficiently precise and uniform subsidiary standards. If that were so, of course, the fault would lie not with the Florida Supreme Court, but with the Florida Legislature because it specified no standard other than “the clear intent of the voter” as pertaining under Florida law to the evaluation of uncertain ballots. Had the Florida Supreme Court promulgated “uniform subsidiary standards” to define the legislative standard more precisely, it would surely have been accused of violating due process by changing the rules of vote counting after the election. Earlier judicial opinions gave no grounds for divining any subsidiary standards, and the court’s caution could only have been greater because the Supreme Court, in Bush v. Palm Beach County Canvassing Board, had intimated the rather unlikely proposition that any judicial elaboration upon the state’s statutory law might even violate Article II.184 But there are three more fundamental objections to the Supreme Court’s disparagement of Florida’s lack of subsidiary standards. The first is that there would likely be so many ways in which an inspector, reasonably and in good faith, could believe she had discerned a clear voter intent or the absence of it that any precise subsidiary standard would itself involve some degree of arbitrariness. Complicating the process through the imposition of arbitrary sub-rules may not have seemed necessary to achieve reasonably accurate results. It would not have been a surprising experience for officials involved in the recounting efforts that, after reviewing some substantial number of ballots, they would find themselves with shared and wellunderstood common sense norms sufficient to produce relatively consistent results for each counting team and county canvassing board.185 It is not clear that subsidiary standards would have improved accuracy and fairness to any real extent—at least if “the clear intent of the voter” were to remain the governing primary standard. That point brings up the second critical objection to the Supreme Court’s analysis of standards. Because “clear intent of the voter” was 183. Id. at 145 (Breyer, J., dissenting) (emphasis added). 184. 531 U.S. at 76-77. 185. Cf. Tom Collins, Absent a Clear Statute, Broward Judge Says He Turned to Common Sense to Consider Chads, MIAMI DAILY BUS. REV., Feb. 2, 2001, at C1, available at Westlaw MIAMIDBR database. 2001] DISAPPEARING DEMOCRACY 577 the legislatively prescribed standard, a good argument exists under Florida law that any subsidiary standard could operate as no more than a burden-shifting device. In other words, subsidiary standards consistent with the law might have provided, for example, that dimpled chads, in and of themselves, should not automatically be deemed conclusive as to voter intent. It might well have violated Florida law, however, to deny either party the possibility, even under subsidiary standards, of arguing that a dimpled chad on a particular ballot did meet the Florida statutory standard of “clear intent of the voter.” To deny that possibility would have been to amend Florida’s statutory standard. The third, and perhaps most important point, is that Judge Lewis sat at the anticipated end of the recounting process to iron out any unjustifiable inconsistencies. That is, the availability of a single adjudicator to determine whether one ballot was treated too easily as evidencing clear intent or another was excluded too stringently offered a structural assurance of consistency across counties. So long as the state’s attempt at insuring consistency is reasonably meaningful, as this was, the demands of due process are satisfied. None of this is to apologize for the ways in which Florida’s system departed from ideal administrative justice. The state would surely have been better served had all relevant decisionmakers simply ignored the federal safe-harbor provision and pursued comprehensive statewide recounts with yet more time for judicial review as the instrument of achieving tolerable consistency. But the level of inconsistency alleged with regard to Florida’s vote tabulation is of no different order than we live with regularly under virtually every system of mass adjudication. In a famous study of social security adjudication in the 1970s, Jerry Mashaw discovered variations among federal administrative law judges in granting social security disability awards at rates ranging from under ten to over eighty percent of their respective caseloads. Moreover, and quite startlingly, he was not able to find through statistical analysis any systematic differences in the cases or caseloads confronting the Administrative Law Judges that accounted for these different rates of award. In other words, the only good predictor of a litigant’s likelihood of recovering a social security award in a social security hearing was the identity of the Administrative Law Judge.186 This may have been a deplorable feature of the social security system, but there was and there is no well-founded theory of due 186. JERRY L. MASHAW ET AL., SOCIAL SECURITY HEARINGS AND APPEALS: A STUDY OF THE SOCIAL SECURITY ADMINISTRATION HEARING SYSTEM 21 (1978); JERRY L. MASHAW, RICHARD A. MERRILL & PETER M. SHANE, ADMINISTRATIVE LAW: THE AMERICAN PUBLIC LAW SYSTEM 411-12 (4th ed. 1998). 578 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 process law under which it could have been judged unconstitutional. The same statement is equally true of Florida’s aborted recount. In short, despite the Court’s identification of real and potential variations among counties and their vote counters in their treatment of ballots, the majority did not provide an analytic framework under either equal protection or due process that justified terminating the recount on grounds of unconstitutionality. None of the differences in vote counting process identified by the majority as an equal protection problem represented an intentional effort to disenfranchise any identifiable group. All were the by-product of a feature of state governance long treated as reasonable, namely, state deference to county administrative discretion in the implementation of statewide administrative mandates. Of course, even in such circumstances, if some fundamental question of unfairness to individual voters were to arise, then the ground would exist for judicial intervention in the name of due process. But the effect of the Supreme Court’s intervention was not to remedy unfairness to individual voters; it was to prevent the State of Florida from undertaking a good faith effort to ascertain the intentions of as many of its voters as possible. The Bush v. Gore majority ignored a lesson made abundantly clear from the hypothetical case discussed above of Gore v. Harris: the importance of procedural due process in the tabulation of every voter’s ballot as a bulwark of democratic government. In that spirit, the one thing due process should have guaranteed in Florida was a manual recount in those counties where either candidate could make out a prima facie showing that, for whatever reason, some failure had occurred for for a county’s automatic tabulation system to count legally valid ballots of sufficient number to have swayed the electoral outcome. Gore had effectively made such a showing in four counties. Florida’s judicial system was well on its way to fulfilling the demands of due process. It is the Supreme Court of the United States that was the primary agent of unfairness in this episode. III. REAL DEMOCRACY: BUSH V. GORE AND INSTITUTIONAL RESTRAINT Upon his elevation to the Supreme Court, Justice Lewis Powell decided to refrain from voting in presidential elections lest it compromise, however indirectly, the appearance or reality of his obligation to remain impartial.187 By contrast with that display of heroic self-restraint, the current Supreme Court’s conservative majority has created an unsettling appearance of wanting to vote for President not only on Election Day, but to do so as often as necessary thereafter. 187. Telephone Interview with John C. Jeffries, Jr., Dean, University of Virginia School of Law, Justice Powell’s biographer and former clerk (July 2, 2001). 2001] DISAPPEARING DEMOCRACY 579 The arrogance of five Justices in bringing a presidential election to a halt by a one-vote judicial majority will likely haunt history’s appraisal of every other bit of this right-wing faction’s jurisprudential corpus—and it should. If any value should be paramount in constitutional review of a presidential election, it is the value of democracy. Democracy’s nonappearance among the concerns of the Bush v. Gore majority is stunning, especially in light of the Rehnquist Court’s ordinary solicitousness toward protecting political processes against judicial intervention, especially at the state and local level. There is ample evidence of the majority’s tendentiousness.188 The seriousness with which the Court’s earlier opinion in Bush v. Palm Beach County Canvassing Board treated Bush’s near-frivolous Article II argument,189 Justice Scalia’s reasons in support of the decision to stay the judicially ordered recount in Florida,190 and the utter im188. An especially disturbing feature of the lack of restraint in Bush v. Gore is the apparently close alignment of various Justices with conservative electoral politics. One muchreported and especially distressing incident in this vein was recounted in the press as follows: [A]t an election night party on Nov. 7, surrounded for the most part by friends and familiar acquaintances, [Justice Sandra Day O’Connor] let her guard drop for a moment when she heard the first critical returns shortly before 8 p.m. Sitting in her hostess’ den, staring at a small black-and-white television set, she visibly started when CBS anchor Dan Rather called Florida for Al Gore. ‘This is terrible,’ she exclaimed. She explained to another partygoer that Gore’s reported victory in Florida meant that the election was ‘over,’ since Gore had already carried two other swing states, Michigan and Illinois. Moments later, with an air of obvious disgust, she rose to get a plate of food, leaving it to her husband to explain her somewhat uncharacteristic outburst. John O’Connor said his wife was upset because they wanted to retire to Arizona, and a Gore win meant they’d have to wait another four years. Dave Zweifel, Ruling for Bush Fits Justice’s Plans, THE CAPITAL TIMES (Madison, Wis.), Dec. 27, 2000, available at 2000 WL 24299725. That a Supreme Court Justice would attend a social gathering for watching the presidential returns alone suggests an insensitivity to the appearance of impartiality that ought govern the members of the Court. 189. Bush v. Palm Beach County Canvassing Bd., 531 U.S. at 76-77. There is no evidence, in originally granting legislatures the authority to determine how presidential electors would be appointed, that the Framers intended to restrict the authority of states to delimit the powers of their respective legislatures through state constitutions or judicial review. See 5 ELLIOTT’S DEBATES ON THE FEDERAL CONSTITUTION 338 (1996). 190. Bush v. Gore, 531 U.S. 1046, 1046 (2000) (Scalia, J., concurring). Compare the Eleventh Circuit’s analysis three days earlier: At this time, Plaintiffs cannot demonstrate a threat of continuing irreparable harm. At the moment, the candidate Plaintiffs (Governor Bush and Secretary Cheney) are suffering no serious harm, let alone irreparable harm, because they have been certified as the winners of Florida’s electoral votes notwithstanding the inclusion of manually recounted ballots. Moreover, even if manual recounts were to resume pursuant to a state court order, it is wholly speculative as to whether the results of those recounts may eventually place Vice President Gore ahead. . . . Nor are the voter Plaintiffs (all of whom allege that they voted for Governor Bush and Secretary Cheney) suffering serious harm or facing imminent injury. 580 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 plausibility of the majority’s remedial opinion in Bush v. Gore all suggest a Court not merely wrong, but reckless.191 Also, and perhaps most telling, is the dissonance between the per curiam opinion and its authors’ previously articulated jurisprudential commitments. This particular majority’s disregard for federalism and its sudden embrace of equal protection analysis at its most fastidious is only rendered more suspect by the self-declaration that the Justices may be writing a ticket for one ride only: “Our consideration is limited to the present circumstances, for the problem of equal protection in election processes generally presents many complexities.”192 Among the jurisprudential commitments abandoned is these Justices’ prior commitment to the political question doctrine. Bush v. Gore can be usefully compared, for example, to Nixon v. United States,193 in which the Court confronted a challenge by an impeached No voter Plaintiff claims that in this election he was prevented from registering to vote, prevented from voting or prevented from voting for the candidate of his choice. Nor does any voter claim that his vote was rejected or not counted. . . . Even assuming Plaintiffs can assert some kind of injury, they have not shown the kind of serious and immediate injury that demands the extraordinary relief of a preliminary injunction. Additionally, any alleged voter injury, unrelated to the outcome of the election certified by the Florida Secretary of State, can be adequately remedied later. And although these Plaintiffs assert that Florida’s existing manual recount scheme must be invalidated for now and in the future, no one suggests that another election implicating those procedures is underway or imminent. Plaintiffs’ other allegations of irreparable injuries to justify a preliminary injunction are unconvincing. The candidate Plaintiffs contend that if the manual recounts are allowed to proceed, simply rejecting the results of those recounts after the conclusion of this case will not repair the damage to the legitimacy of the Bush Presidency caused by “broadcasting” the flawed results of a recount that put Vice President Gore ahead. But . . . we reject the contention that merely counting ballots gives rise to cognizable injury. Siegel v. LePore, 234 F.3d 1163, 1177 (11th Cir. 2000) (second emphasis added) (citations omitted). 191. In the interest of finality, . . . the majority effectively orders the disenfranchisement of an unknown number of voters whose ballots reveal their intent— and are therefore legal votes under state law—but were for some reason rejected by ballot-counting machines. It does so on the basis of the deadlines set forth in Title 3 of the United States Code. But, as I have already noted, those provisions merely provide rules of decision for Congress to follow when selecting among conflicting slates of electors. They do not prohibit a State from counting what the majority concedes to be legal votes until a bona fide winner is determined. Indeed, in 1960, Hawaii appointed two slates of electors and Congress chose to count the one appointed on January 4, 1961, well after the Title 3 deadlines. Thus, nothing prevents the majority, even if it properly found an equal protection violation, from ordering relief appropriate to remedy that violation without depriving Florida voters of their right to have their votes counted. As the majority notes, “[a] desire for speed is not a general excuse for ignoring equal protection guarantees.” Bush v. Gore, 531 U.S. 98, 127 (2000) (Stevens, J., dissenting) (citations omitted). 192. Id. at 109. 193. 506 U.S. 224 (1993). 2001] DISAPPEARING DEMOCRACY 581 and convicted federal judge to the procedures by which the Senate had removed him from office. Rather than taking evidence in plenary session, the Senate delegated that function to a committee.194 The Senate as a whole met only to review the Committee’s report and to hear such arguments as Judge Nixon was prepared to offer on his own behalf.195 Nixon argued to the Supreme Court that this procedure denied him his constitutional right to be tried by the Senate.196 The Supreme Court unanimously determined that it would not intervene in the matter. The majority, speaking through Chief Justice Rehnquist, held that the constitutional sufficiency of Senate procedures for adjudicating impeachment controversies was a matter to be resolved exclusively by the Senate itself. The constitutional vesting in the Senate of “the sole Power to try all Impeachments”197 was deemed “a textually demonstrable commitment of the issue to a coordinate political department.”198 Moreover, the majority said, the word “try” in Article I was too general to engender “judicially manageable standards” for what would amount to a constitutionally sufficient trial.199 It is unmistakably dramatic testimony to the Court’s determination to respect Congress’s impeachment authorities that the Court was prepared to hold that, for purposes of enforcing Article I, Article III judges could not determine in a sufficiently rigorous way what ought to count as a trial.200 Nixon v. United States rests on sound institutional judgment. Impeachment is the sole constitutionally designated process for achieving judicial accountability for wrongdoing. It would have appeared an unseemly conflict of interest for the Court to have reserved to the judiciary the power to oversee that very process. But the argument for eschewing involvement in the 2000 presidential election is surely even more compelling. For sitting Supreme Court Justices to adjudicate which person shall be entitled to name their successors does unmistakable violence to constitutional checks and balances. Article II and the Twelfth Amendment are readily interpretable as embodying a textually demonstrable commitment to Congress of the power to resolve all issues related to the proper tabulation of electoral 194. See id. at 227. 195. Id. at 227-28. 196. Id. at 228. 197. U.S. CONST. art. I, § 3, cl. 6. 198. Nixon, 506 U.S. at 228. 199. Id. at 230. 200. Nixon was not the first occasion on which now-Chief Justice Rehnquist has been prepared to find a “political question” in the face of constitutional text that would seem susceptible to fairly conventional judicial interpretation. See Goldwater v. Carter, 444 U.S. 996, 1002-06 (1979) (Rehnquist, J., concurring) (arguing that the issue whether the Constitution authorizes Presidents to abrogate treaties unilaterally presents a political question). 582 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 votes. Indeed, Congress has enacted a detailed statutory scheme to make just that process possible.201 In making this point, I do not wish to suggest for a moment that Congress, like the Court, would not have felt the pull of the most blatant partisan politics. The Clinton impeachment episode is a cautionary tale with regard to relying on Congress to abide by longstanding constitutional norms. The national experience with congressional involvement in resolving the Hayes-Tilden presidential contest in 1876 is not heartening.202 But there are at least three reasons why deference to Congress’s authority in this matter would have been far preferable to the Supreme Court’s adventure of December 12, 2000. The first is that, partisan pressure or not, Congress in 2001 would likely have had to deport itself with a degree of openness or “transparency” that did not exist in 1876 for Congress and which does not exist in 2001 for the Supreme Court. To that extent, hope for Congress to adopt a sound approach to the resolution of the Florida controversy, if need be, would have had some rational basis. Second, Congress—unlike the Supreme Court—if it discerned infirmities in Florida’s counting process, could have provided a resolution consistent with the fundamental objective of ascertaining Florida’s actual vote. That is, Congress could have provided by statute for a statewide recount under appropriate standards and procedures to which Congress could then have bound itself. This may have required compromises—Gore would probably have had to give up on “dimpled” chads; Bush would have had to concede the legality of intelligible ballots rendered unreadable to machines due to voter error. But the result would have provided at least a reasonable and publicly acceptable answer to the question, “For whom did Floridians vote?” But the third reason is the most compelling. Even if Congress messed up, even if it cut deals behind closed doors, and even if it failed to deal reasonably with Florida’s difficulties in achieving an accurate count, a simple fact remains: if the people of the United States were unhappy with Congress on any such account, those members of Congress deemed responsible could have been voted out of office. One could hardly imagine a set of decisions in which democratic accountability is more important than those involved in the legitimate political resolution of an election contest. By contrast, despite the ineffable odor of partisanship that hangs over the Court’s opinion in Bush v. Gore, there is no politically appropriate response 201. 3 U.S.C. §§ 1-15 (1994). 202. See ERIC FONER, RECONSTRUCTION: AMERICA’S UNFINISHED REVOLUTION, 18631877, at 575-87 (1988); EDWARD STANWOOD, A HISTORY OF THE PRESIDENCY 357-93 (1898). 2001] DISAPPEARING DEMOCRACY 583 to be levied against the responsible Justices themselves. The Court that decided Nixon v. United States should surely have known better. The only counterargument I can think of is that the 2000 presidential election might have seemed to present a profoundly significant moment for the articulation of constitutional law. Had the Court used its authority to articulate the federal right to vote for presidential electors or had the Court engaged in an analysis of Florida’s vote count procedures that was grounded seriously in its well-established due process jurisprudence, then at least the Court might have claimed to be advancing the causes of fairness and of democracy in a manner true to the trajectory of our constitutional commitments to those values. Instead, it betrayed both. I have written on other occasions that judicial activism can be constructive.203 For that to happen, however, two conditions must prevail. First, there must be a genuine defect in the political process that renders judicial intervention necessary to preserve constitutional values. Second, there must be available to the judiciary a means of persuasively translating its commitment to constitutional values into sound law. That is, judicial creativity is constructive only if amenable to rendition as professionally credible constitutional doctrine based on acceptable forms of good faith legal argument. The doctrine has to be able to do the work of advancing its animating values as an articulation of adequately neutral principle to be persuasively applicable to foreseeable future cases.204 The activism of the Bush v. Gore majority plainly fails the second test. The Court’s prior equal protection jurisprudence does not sustain the stringency of its scrutiny of intrastate procedural variations in the administration of elections. And, it will be interesting, to put it mildly, to see if the Supreme Court is willing to follow the holding of Bush v. Gore in any other case. But, more to the point, the 2000 election did not trigger the first justification for judicial activism. It did not suggest the existence of any defect in our national political process that required creative judicial intervention. Justice Scalia’s line, “Count first, and rule upon legality afterwards, is not a recipe for producing election results that have the public acceptance democratic stability requires,”205 puts the matter exactly backward. Counting is precisely what democracy re203. See, e.g., Peter M. Shane, Federalism’s “Old Deal”: What’s Right and Wrong with Conservative Judicial Activism, 45 VILL. L. REV. 201, 228-29 (2000). 204. Contrary to the conclusions of Professor Pushaw, see Robert J. Pushaw, Jr., The Presidential Election Dispute, the Political Question Doctrine, and the Fourteenth Amendment: A Reply to Professors Krent and Shane, 29 FLA. ST. U. L. REV. 603 (2001), these criteria provide a fully principled basis for applauding the activism of Baker v. Carr, 369 U.S. 186 (1962), while deploring the Court’s decision in Bush v. Gore. 205. Bush v. Gore, 531 U.S. 1046, 1046 (2000) (Scalia, J., concurring). 584 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:535 quires, and the Court’s implicit doubts about the capacity of Congress to act responsibly in assessing the counting process expresses an astonishing distrust of political institutions to resolve essentially political disputes. Given the shallowness of the majority’s analysis, and the Court’s obliviousness to the genuine stakes in the controversy before it, it is possible to say at least this much about what would have been Congress’s resolution of the Florida controversy— Congress could have done no worse than the Court. In ignoring this possibility—as in ignoring the imperatives of due process for Florida voters and the impact of the Fourteenth Amendment on the right of citizens to participate in choosing presidential electors—the Bush v. Gore majority betrayed democracy. CONCLUSION In resolving Bush v. Gore, the Supreme Court had two democratic paths open to it. It could have deferred to Congress under the political question doctrine, holding that the detailed textual provisions of the Constitution concerning presidential elections commit to Congress the final resolution of the question whether any state has properly administered its appointment of presidential electors. Such a decision would have furthered the cause of democracy by placing in the hands of elected officials the determination of our most important political contest. Alternatively, the Court could have adjudicated the constitutional issues before it with an eye to insuring that, to the maximum extent possible, the votes of all Florida voters actually counted in the presidential election. That would have required no more than the application of the due process clause to uphold the Florida Supreme Court’s judgment below—or, even more modestly, the simple rejection on grounds of insubstantiality of the Bush challenges to the Florida hand count process. Unlike the majority’s cut-from-whole-cloth equal protection analysis, either of these approaches would have required nothing other than the conventional application of precedent. Nixon v. United States would have amply justified deference to Congress. The entire line of procedural due process jurisprudence since the 1970s would have supported affirmation of the Florida Supreme Court’s approach to vote counting. But, whatever the majority’s concerns, adhering to conventional approaches to constitutional application was not among them. Without even a glance at the Fourteenth Amendment, the majority commenced its legal analysis with an assertion and reiteration that the citizens of each state have no constitutional entitlement to participate in the appointment of presidential electors. The majority’s premise takes no serious account of the text of the Fourteenth 2001] DISAPPEARING DEMOCRACY 585 Amendment, its history, nearly universal practice since 1868, and the conspicuous deepening of our constitutional commitment to democracy in the ensuing 132 years. The failure even to address these issues mocks the majority’s supposed commitments to textualism and originalism in other contexts. History will record that, in resolving a dispute over the world’s most important elected office, the Supreme Court penned an opinion in which our national commitment to democracy—indeed, the very word, “democracy”—does not appear. Those who hope that a decent presidential performance by the Court’s designated victor will minimize the harm to the Court’s reputation and legitimacy are whistling in the dark. The precedent of such judicial usurpation poses grave long-term peril to both democracy and the rule of law that is not susceptible to ready eradication in the short run—surely not by a lawabiding nation’s prudent resignation to the formality, at least, of government by the winner. Only future courts’ repudiation of Bush v. Gore can begin to erase the stain. UNDULY PARTIAL: THE SUPREME COURT AND THE FOURTEENTH AMENDMENT IN BUSH V. GORE PAMELA S. KARLAN* I. THE FUZZINESS OF SECTION 2 ............................................................................. II. SUBSTANTIVE DUE PROCESS AND THE RIGHT TO VOTE ...................................... III. DUE PROCESS AND EQUAL PROTECTION AS RHETORICAL STRATEGIES .............. CONCLUSION ........................................................................................................ 589 593 599 601 An implicit moral of Peter Shane’s insightful Disappearing Democracy1 is that Laurence Tribe made a key strategic error in the first sentence of his oral argument in Bush v. Palm Beach County Canvassing Board by directing the Supreme Court’s attention away from the Due Process Clause.2 Rather than simply dismissing Bush’s due process arguments,3 Tribe should have argued that both substantive and procedural due process in fact required the Florida Supreme Court to protect the right of every Florida voter to have his or her vote counted. The central message of Disappearing Democracy is that the due process problems with stopping the Florida recounts were far more serious than any equal protection problem with letting them continue. Disappearing Democracy offers two major arguments. The first argument focuses on the Reduction-of-Representation Clause of Section 2 of the Fourteenth Amendment. That clause requires a reduction in a state’s population base for apportionment purposes when a state denies or abridges “the right to vote at any election for the choice of electors for President and Vice President of the United States, Representatives in Congress, the Executive and Judicial officers of a State, or the members of the Legislature thereof.”4 Shane * Kenneth and Harle Montgomery Professor of Public Interest Law, Stanford Law School. I presented some of the ideas in this Comment at Florida State University College of Law’s live symposium on March 23, 2001; others reflect several months’ additional thought about Bush v. Gore, the Supreme Court, and the right to vote. Many of these ideas came from my collaborative work with Sam Issacharoff and Rick Pildes, both on our comprehensive casebook, THE LAW OF DEMOCRACY: LEGAL STRUCTURE OF THE POLITICAL PROCESS (2d ed. 2001) [hereinafter THE LAW OF DEMOCRACY], and on our shorter casebook, WHEN ELECTIONS GO BAD: THE LAW OF DEMOCRACY AND THE PRESIDENTIAL ELECTION OF 2000 (rev. ed. 2001) [hereinafter WHEN ELECTIONS GO BAD]. I also appreciate several suggestions from Viola Canales. 1. Peter M. Shane, Disappearing Democracy: How Bush v. Gore Undermined the Federal Right to Vote for Presidential Electors, 29 FLA. ST. U. L. REV. 535 (2001). 2. Tr. of Oral Argument at 44-45, Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000) (No. 00-836), available at http://election2000.stanford.edu/00-836.pdf (“I think I would want to note at the outset that the alleged due process violation which keeps puffing up and then disappearing . . . is really not before the Court.”). 3. For a brief discussion of this issue, see infra text accompanying notes 76-78. 4. U.S. CONST. amend. XIV, § 2. More precisely, Section 2 requires that a state’s population basis be reduced when the right to vote: 587 588 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:587 argues that implicit in this clause is a “background understanding that individual citizens in the several states would now be guaranteed some form of franchise in presidential elections, as well as in House contests.”5 This conclusion is reinforced, Shane suggests, both by the Privileges or Immunities Clause and by subsequent doctrinal and constitutional developments.6 The second argument focuses on the Due Process Clause.7 Shane points out that voting is a species of liberty (or property) interest.8 Vote tabulation is a “governmental process for making adjudicatory decisions” 9: whether a citizen cast a valid ballot; for whom each valid ballot was cast; and ultimately who has won an election. Shane then shows how fairly straightforward procedural due process analysis leads to the conclusion that inaccuracies in machine-only tabulation of ballots required Florida to have some manual reexamination process.10 The recount ordered by the Florida Supreme Court was necessary to bring Florida’s otherwise inadequate adjudicatory process into compliance with the Due Process Clause. In this Comment, I make three points. The first two focus on Shane’s analysis. I am skeptical that Section 2 necessarily transformed the process by which electors are selected into “one in which individual citizens must be allowed to participate” through popular elections.11 To my mind, substantive due process is a stronger vehicle for safeguarding that right. In a related vein, without resolving, or at least analyzing, the nature of the liberty interest at stake in voting rights cases, it is impossible fully to flesh out the procedural due process issue. The final point is more speculative, and focuses on a due process claim that Shane does not discuss: George W. Bush’s allegation that the Florida Supreme Court’s decisions changed the ex- is denied to any of the male inhabitants of such State, being twenty-one years of age, and citizens of the United States, or in any way abridged, except for participation in rebellion, or other crime [by the] proportion which the number of such male citizens shall bear to the whole number of male citizens twenty-one years of age in such State. Id. Some aspects of Section 2—such as its protection only of men’s right to vote and its age reference—have presumably been superseded by later constitutional amendments. See id. amend. XIX (forbidding discrimination in voting on account of sex); id. amend. XXVI (giving 18 year-olds the right to vote). Other aspects, such as its implicit approval of felon disenfranchisement statutes, see Richardson v. Ramirez, 418 U.S. 24 (1974), remain both vital and controversial. See THE LAW OF DEMOCRACY, supra note *, at 21-40 (discussing the question of felon disenfranchisement). 5. Shane, supra note 1, at 544. 6. See id. at 546-47. 7. U.S. CONST. amend. XIV, § 2. 8. Shane, supra note 1, at 562. 9. Id. at 552. 10. Id. at 562-68. 11. Id. at 539. 2001] UNDULY PARTIAL 589 isting law unfairly.12 Even the three Justices who pressed the Article II theory—that the Florida Supreme Court had infringed the legislature’s federally inviolable prerogative to determine the manner in which a state’s electors are appointed13—ignored a squarely pertinent line of due process cases. I suggest that the Court had several reasons, none of them admirable, for relying on the Equal Protection Clause instead. In particular, reliance on the Due Process Clause would have made clear what the Court’s invocation of equal protection worked to obscure: that the only litigants whose interests the Court’s decision actually served were George W. Bush and Dick Cheney, and not any of the voters in Florida. I. THE FUZZINESS OF SECTION 2 Shane’s argument regarding the constitutional consequences of the Reduction-of-Representation Clause can be described as quasioriginalist. The central concern that animated the drafters was that the readmitted Southern states would have their black population included fully in a state’s apportionment base, thereby increasing their relative number of House seats (and concomitantly, electoral votes), but would disenfranchise black citizens, thereby enhancing the political power of white, unreconstructed Democrats (and diminishing the power of the Northern Republicans who then controlled Congress). Shane argues that “[s]o long as we interpret the Constitution as making popular involvement in presidential elections a state legislative prerogative,”14 Southern states could accomplish precisely their goal of using their black inhabitants as inert ballast for enhancing white political strength (the very effect of the now-repealed “three-fifths” clause): All that need happen in a Southern state intent on maintaining white control of presidential elector appointments is for the majority-white legislature to institute or maintain a practice of having itself select slates of presidential electors without any popular vote involvement.15 I think Shane’s argument focuses too narrowly on the language about the right to vote in elections for presidential and vicepresidential electors. In light of the full list of covered elections, I think it is more sensible to read Section 2 conditionally. If the state employs an election, then it cannot deny or abridge the right to vote in that election without incurring a reduction in its apportionment base. 12. 13. 14. 15. Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70, 73 (2000). Bush v. Gore, 531 U.S. 98, 112-15 (2000) (Rehnquist, C.J., concurring). Shane, supra note 1, at 544. Id. 590 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:587 The Reduction-of-Representation Clause specifies denials or abridgements of the right to vote in five distinct sets of elections as triggering a reduction of representation.16 In addition to the federal offices on which Shane focuses, the Clause also covers elections for state legislators, the state “Executive,” and state judicial officers.17 But many states, both in the 1860s and today, do not select their judges through popular elections. Thus, they consistently and categorically deny the right to vote for state judicial officers to every inhabitant.18 Moreover, “[t]here is no provision of the United States Constitution or any of its amendments which either expressly or impliedly dictates the method a State must use to select its Governor.”19 Several states have constitutional provisions that authorize the state legislature to choose the governor if no candidate receives an outright majority of the votes cast in a popular election.20 These provisions, too, might be described as denying or abridging a right to vote. But it would be absurd to conclude that Section 2 was intended to strip states with appointed judiciaries or legislatively selected governors of all their representation in Congress and all but two of their electoral votes. Put somewhat differently, Section 2 is relative, not absolute. State law defines whether there is to be an election, but once a state decides to fill an office through popular balloting, it faces a strong federal incentive to treat citizens equally. Moreover, the danger that concerns Shane is relatively remote. The likelihood of a state’s abolishing popular election altogether was quite low even in 1868 (and essentially nonexistent today). As Shane notes, by the time of the Fourteenth Amendment, all states used popular election to select their electors.21 To disenfranchise black voters in presidential elections without losing congressional seats, a state would also have to disenfranchise white voters who had long enjoyed (and exercised—voter turnout in the nineteenth century was far higher than today) the right to vote. Making that change would be politically risky.22 As Justice Holmes observed, “[a] thing which 16. See U.S. CONST. amend. XIV, § 2. 17. Id. 18. Of course, the Constitution requires every state to use elections to fill House seats. See id. art. I, § 2. 19. Fortson v. Morris, 385 U.S. 231, 234 (1966). 20. Morris noted then-existing constitutional provisions in Georgia, Mississippi, and Vermont and also referred to constitutional or statutory provisions in thirty-eight other states that would turn to legislative selection if the popular election produced a tie. Id. at 234-35. 21. Shane, supra note 1, at 545. 22. That does not mean it would be impossible. During Redemption, Southern states passed disenfranchising provisions that aimed at black voters but also swept large numbers of white voters off the rolls. But even in those cases, the states did not strip most white voters of the right to vote. Pamela S. Karlan, Loss and Redemption: Voting Rights at the Turn of a Century, 50 VAND. L. REV. 291, 294 (1997). 2001] UNDULY PARTIAL 591 you have enjoyed and used as your own for a long time, whether property or an opinion, takes root in your being and cannot be torn away without your resenting the act and trying to defend yourself, however you came by it.”23 The legislators who voted for such a change would likely face the angry consequences at their next election. And of course, it is not entirely clear that the state legislatures would either have, or perceive themselves as having, the right to make such a drastic change through ordinary legislation.24 There is no reason to suppose that nineteenth-century state legislatures would view themselves as receiving powers from the federal government that were denied them by their own state constitutions. Moreover, I do not read the Bush v. Gore per curiam’s remark that “[t]he State, of course, after granting the franchise in the special context of Article II, can take back the power to appoint electors”25 as necessarily going that far. A state legislature that acts beyond its authority under the state constitution arguably no longer is the state. And if a constitutional amendment were required to abolish popular balloting for presidential electors—as would be the case in many states—it would be hard to imagine the plebiscitary ratification process in which a majority of the electorate would vote to disenfranchise itself. Finally, there is little reason to think that creating an affirmative right to vote in elections to choose electors adds anything very significant to the operation of Section 2, assuming that Section 2 works at all.26 Section 2 protects the right to vote in state legislative elections.27 If that right were fully respected—and of course it was not28— then it often would not matter to the outcome whether a state used popular election or legislative selection to pick its presidential electors. In a state with a white majority and racially polarized voting, the outcome of a winner-take-all popular vote will be that the whitepreferred candidate gets all the state’s electoral votes. That will be 23. Oliver Wendell Holmes, Jr., The Path of the Law, 10 HARV. L. REV. 457, 477 (1897). 24. For extensive discussion of this issue, see James A. Gardner, The Regulatory Role of State Constitutional Structural Constraints in Presidential Elections, 29 FLA. ST. U. L. REV. 625 (2001). 25. Bush v. Gore, 531 U.S. 98, 104 (2000). 26. Despite its sweeping language, Section 2 turned out to be toothless because neither Congress nor the courts ever showed themselves willing to pull the trigger, despite roughly a century of black disenfranchisement in the South. See, e.g., Lampkin v. Connor, 360 F.2d 505 (D.C. Cir. 1966) (dismissing a Section 2 claim challenging the number of seats given to Southern states); George David Zuckerman, A Consideration of the History and Present Status of Section 2 of the Fourteenth Amendment, 30 FORDHAM L. REV. 93 (1961) (recounting the lack of post-ratification enforcement of Section 2 by Congress). 27. See U.S. CONST. amend XIV, § 2. 28. For a recent discussion of the various devices Southern states used to disenfranchise black voters during the nineteenth century, see J. MORGAN KOUSSER, COLORBLIND INJUSTICE: MINORITY VOTING RIGHTS AND THE UNDOING OF THE SECOND RECONSTRUCTION 16-38 (1999). 592 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:587 true whether black voters are permitted to participate or not. The number of black voters in Mississippi has skyrocketed since the passage of the Voting Rights Act of 1965, which effectively reenfranchised African Americans. And yet, in the last nine presidential elections, despite overwhelming black support for the Democratic candidate, the Republicans have carried all the state’s electoral votes eight times.29 The same is true for South Carolina, the state with the second-highest percentage of black residents.30 And if a state legislature were fairly drawn to reflect a state’s racial composition,31 and there were significant racial polarization, then legislative selection would produce the same sweep for the white-preferred candidate. A majority of the legislators would represent majority-white districts and would vote for the white-preferred candidate. By contrast, in a state where the white community was split between the two parties, and the black community was politically cohesive, then the black community might be the swing vote.32 In a popular election, the black-preferred candidate would win if black votes plus the votes of white faction A were greater than the votes of white faction B.33 And if the state legislature were fairly districted (again, a counterfactual hypothesis), there is a reasonable probability that legislative selection would produce the same result. Moreover, a targeted disenfranchisement strategy might be counterproductive to a state wishing to preserve white political supremacy. If blacks could fully participate in every aspect of a state’s political system other than how it selected its electors, unsuccessful white factions would continue to have an incentive to build political coalitions with the black community to gain control over the state government. Over time, this might either result in the reintroduction of popular balloting or result in legislative selection throwing the state’s electoral votes to the candidate supported by the biracial coalition. Disenfranchisement across the board was the only stable 29. See NATIONAL ARCHIVES RECORDS ADMINISTRATION, U.S. ELECTORAL COLLEGE: 2000 ELECTION RESULTS, at http://www.nara.gov/fedreg/elctcoll/index.html (last updated Apr. 5, 2001) (providing state by state electoral votes). 30. During Reconstruction, both Mississippi and South Carolina were majority-black states. As long as each had a majority-black electorate, Republican candidates won all the states’ electoral votes. See KOUSSER, supra note 28, at 20-22, 29. 31. This, too, is a deeply counterfactual assumption. Many of the most heavily black states engaged in white-driven racial gerrymanders throughout Reconstruction. See id. at 28-31. 32. I explore the conditions under which blacks can be the swing vote in Karlan, supra note 22, at 295. 33. Note that black participation is outcome-determinative only if two things are true: (1) the black vote is larger than the difference between the two white factions (otherwise, the gap between the white factions will determine the outcome) and (2) the black community forms a coalition with the smaller of the two white factions (otherwise, the result does not change whether blacks participate or not). 2001] UNDULY PARTIAL 593 strategy. Southern states pursued it ferociously and Congress and the courts proved themselves unwilling to pull the Section 2 trigger even then.34 II. SUBSTANTIVE DUE PROCESS AND THE RIGHT TO VOTE The Supreme Court has also foreclosed Shane’s other candidate for an express constitutional guarantee of the right to vote in presidential election—the Privileges or Immunities Clause of the Fourteenth Amendment.35 In October 1872, Virginia Minor, “a native born, free, white citizen of the United States, and of the State of Missouri, over the age of twenty-one years, wishing to vote for electors for President and Vice-President of the United States,” sought to register.36 The registrar refused her application because she was a woman.37 She sued, squarely pleading that voting was a privilege or immunity of United States citizenship protected by the Fourteenth Amendment.38 In Minor v. Happersett, the Supreme Court unanimously rejected that argument. It noted that women had always been considered citizens of the United States “the same as men.”39 But, the Court concluded, all citizens were not necessarily voters. At the time of the framing, no state had universal citizen suffrage. States commonly restricted the franchise to adult, male property owners who had resided in the state for a substantial period of time,40 despite the fact that children, women, and poor people were undeniably citizens. And at the time the Fourteenth Amendment was adopted, every state— including the readmitted Southern states whose constitutions were reviewed by the Reconstruction Congress—continued to restrict the franchise to adult males.41 Thus, the common understanding of citizenship, the Court suggested, was not coextensive with voting. If the Privileges or Immunities Clause did not protect Virginia Minor’s ability to vote in the 1872 presidential election, it is unclear why it would protect anyone else’s right to vote, let alone anyone’s right to have an election in which to vote. 34. See supra note 26. 35. U.S. CONST. amend. XIV, § 1 (“No State shall make or enforce any law which shall abridge the privileges or immunities of citizens of the United States”). See Shane, supra note 1, at 546-47. 36. Minor v. Happersett, 88 U.S. 162, 163 (1874). 37. Id. at 163-64. 38. Id. at 164. Happersett, the registrar who refused to enroll Mrs. Minor, did not even bother to retain a lawyer to represent him in the Supreme Court. See id. 39. Id. at 169. 40. See id. at 172-73 (describing the original states’ restrictions on the franchise). 41. Id. at 176-77. 594 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:587 In contrast to Shane, the Minor Court thought that Section 2 of the Fourteenth Amendment actually undercut the idea of voting as a privilege or immunity of citizenship: [I]f suffrage was necessarily one of the absolute rights of citizenship, why confine the operation of the [reduction of representation provision] to male inhabitants? Women and children are, as we have seen, “persons.” They are counted in the enumeration upon which the apportionment is to be made, but if they were necessarily voters because of their citizenship unless clearly excluded, why inflict the penalty for the exclusion of males alone? Clearly, no such form of words would have been selected to express the idea here indicated if suffrage was the absolute right of all citizens.42 Minor v. Happersett ended with a Court “unanimously of the opinion that the Constitution of the United States does not confer the right of suffrage upon any one.”43 And yet, as Shane notes, “the plain democratic trajectory of constitutional development since 1868”44 rejects that consensus. That is why the per curiam’s assertions in Bush v. Gore are so jarring: they are completely out of step with the Court’s general jurisprudence. Indeed, the per curiam’s citation of Article II as giving state legislatures plenary power over how presidential electors are selected45 cannot be taken at face value. Even under the most robust interpretation, a state’s Article II powers are constrained by later constitutional amendments. After the Nineteenth Amendment, the states cannot strip women and women alone of the right to vote in presidential elections.46 After the Twenty-sixth Amendment, they cannot decide to exclude the elderly from voting in presidential elections.47 And the Equal Protection Clause constrains the states in a variety of ways: they cannot establish a popular balloting process that weighs the votes of some voters more heavily than others (Bush v. Gore itself suggests this equal protection constraint)48 or restrict the popular franchise to longstanding state residents. Nor, under the suspectclassification strand of equal protection doctrine, could a state legislature adopt a process for picking electors because of its adverse effects upon a suspect or quasi-suspect class.49 Under contemporary doctrine, the Equal Protection Clause would forbid a state from abol- 42. 43. 44. 45. 46. 47. 48. 49. Id. at 174-75. Id. at 178. Shane, supra note 1, at 548. Bush v. Gore, 531 U.S. 98, 104 (2000). See U.S. CONST. amend. XIX, cl. 1. See id. amend. XXVI, § 1. Bush v. Gore, 531 U.S. at 104. See Personnel Adm’r of Mass. v. Feeney, 442 U.S. 256, 279 (1979). 2001] UNDULY PARTIAL 595 ishing popular elections if the change were motivated by the desire to “maintai[n] white control of presidential elector appointments.”50 Moreover, the enforcement clauses of the Reconstruction and later voting-related amendments give Congress power under appropriate circumstances to override a state’s sovereign prerogatives.51 To take just one salient example, section 5 of the Voting Rights Act forbids certain covered jurisdictions from adopting any change with respect to voting if that change would have a retrogressive effect on minority voting rights.52 The Supreme Court has explicitly held that abolishing elections and replacing them with appointive systems is covered by section 5: “In [this case], an important county officer in certain counties was made appointive instead of elective. The power of a citizen’s vote is affected by this amendment; after the change, he is prohibited from electing an officer formerly subject to the approval of the voters.”53 Thus, if a covered jurisdiction (like Mississippi or Arizona) were to abolish popular election, under section 5 it would first have to prove that the change had neither the purpose nor the effect of diminishing the voting power of African-American, Hispanic, or Native-American citizens.54 The second part of Disappearing Democracy, which focuses on procedural due process, is both perceptive and persuasive. Shane surely is right that voting is a species of liberty interest.55 But the implications of his own analysis, as well as the democratic trajectory of constitutional development, raise important questions about the contours of the right to vote. And they suggest a substantive, as well as a procedural, due process component to the right to vote. To begin with, Shane draws too sharp a distinction between equal protection- and due process-based analyses of claims about voting. As I have explained elsewhere,56 the Court’s pronouncements about voting rights are often “double-barrelled,” reflecting judicial skepticism about a state’s restrictions of the franchise under both the suspectclassification and the fundamental-rights strands of strict scrutiny. The paradigmatic example of such a case is Harper v. State Board of 50. Shane, supra note 1, at 544. 51. See U.S CONST. amend. XIV, § 5; id. amend. XV, § 2; id. amend. XIX, cl. 2; id. amend. XXIII, § 2; id. amend. XXIV, § 2; id. amend. XXVI, § 2. 52. 42 U.S.C. § 1973c (1994). For a full treatment of section 5, which is a very complex statute, see THE LAW OF DEMOCRACY, supra note *, at 546-671. 53. Allen v. State Bd. of Elections, 393 U.S. 544, 569-70 (1969). 54. There are some jurisdictions—Florida is one of them, see 28 C.F.R. pt. 51 app. (2000)—which are only partially covered. The question whether preclearance would apply to them is somewhat knottier, see Lopez v. Monterey County, 525 U.S. 266, 271-77 (1999) (addressing this issue in the context of changes to how California, also a partially covered state, elects municipal judges), but my own view is that it would. 55. See Shane, supra note 1, at 562-63. 56. See Pamela S. Karlan, Just Politics?: Five Not So Easy Pieces of the 1995 Term, 34 HOUS. L. REV. 289, 297-99 (1997). 596 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:587 Elections,57 which struck down poll taxes as a violation of the Equal Protection Clause. The Court offered two now distinct reasons for its conclusion. First, it described the right to vote as “a fundamental matter in a free and democratic society”; thus, “any alleged infringement . . . must be carefully and meticulously scrutinized.”58 Second, it found that “[l]ines drawn on the basis of wealth or property, like those of race, are traditionally disfavored”; thus, “the requirement of fee paying causes an ‘invidious’ discrimination.”59 In hindsight, the first line of cases lead to the current black letter law that the right to vote is a fundamental right, while the second line is an evolutionary dead end. But before it petered out, the suspect-classification argument may have contributed to the Court’s adoption of a fundamental rights perspective—the importance of protecting the right to vote was driven home by the invidiousness of the distinction that kept some citizens from the polls.60 While the Court pigeonholes fundamental rights cases as raising equal protection claims, it could just as easily have denominated them as substantive due process claims. A number of scholars have suggested that the Court’s entire equal protection jurisprudence in the area of voting rights may be little more than a Warren Court recasting of substantive due process concerns in more palatable doctrinal language.61 The Court has certainly termed the right to vote 57. 383 U.S. 663 (1966). 58. Id. at 667 (quoting Reynolds v. Sims, 377 U.S. 533, 561-62 (1964)). 59. Id. at 668 (citation omitted). 60. I also think Shane may have conflated two different areas of equal protection doctrine related to voting. In addition to applying a conventional kind of equal protection analysis, the Court has developed doctrines of equal protection in the political arena that have no counterpart elsewhere. The most salient example is the “one-person, one-vote” standard. Because the one-person, one-vote cases treat the right to vote as fundamental, avoidable population deviations trigger heightened scrutiny. See, e.g., Reynolds v. Sims, 377 U.S. 533 (1964). In a unique formulation, the state must show that the disparities are “necessary to achieve some legitimate goal.” Karcher v. Daggett, 462 U.S. 725, 731 (1983) (emphasis added). This test blends the permissible-ends language from rationality review cases with the appropriate-means language from strict scrutiny. Moreover, there is no requirement in one-person, one-vote cases that the plaintiffs show that the population deviations were the product of intentional discrimination against a group of voters. Cf. James U. Blacksher & Larry T. Menefee, From Reynolds v. Sims to City of Mobile v. Bolden: Have the White Suburbs Commandeered the Fifteenth Amendment?, 34 HASTINGS L.J. 1 (1982) (noting the irony that malapportionment claims brought by white suburbanites face a lower standard of proof than racial vote dilution claims brought by black voters). Thus, I disagree with Shane as to the relevance of the fact that Bush v. Gore involved “no allegation that any [of the differential treatments of ballots] amounted to an intentionally invidious discriminatory practice—the kind of practice that the Court normally requires before it elevates the intensity of its constitutional scrutiny of state practices under the Equal Protection Clause.” Shane, supra note 1, at 552. The redistricting cases on which the Bush v. Gore Court relied did not require any proof of intentional invidious discrimination. That “unconstitutional classification” equal protection claims require proof of intent is arguably beside the point. 61. See, e.g., Ira C. Lupu, Untangling the Standards of the Fourteenth Amendment, 77 MICH. L. REV. 981, 1068 (1979). For the best extended discussion of how the Court has 2001] UNDULY PARTIAL 597 “fundamental,”62 and has held that infringements on the right are subject to strict scrutiny.63 The now constitutionally recognized fundamentality of the right to vote is relevant to the question raised in Disappearing Democracy in two respects. First, as Shane recognizes, the importance of the right to vote informs the operation of the three-part procedural due process calculus of Mathews v. Eldridge 64 by putting a heavy thumb on the side of requiring more reliable procedures.65 A critical flaw in the Supreme Court’s decision is that it in fact fails to vindicate any identifiable voter’s interest in having her ballot counted.66 Second, the substantive strand of the Due Process Clause goes beyond requiring fair procedures to “barring certain government actions regardless of the fairness of the procedures used to implement them.”67 If the right to vote is now understood as a fundamental aspect of the liberty the Due Process Clause protects—and the Court has recognized that analysis of liberty interests is deeply informed by tradition, as reflected in longstanding federal and state practices68— then a Court sensitive to our traditions of ordered liberty should find a substantive liberty interest in voting to elect the President. That interest, as it has evolved and solidified, outweighs an Article II interest in replacing popular election with some other method of selectcommingled and confused various doctrinal strands, see James A. Gardner, Liberty, Community and the Constitutional Structure of Political Influence: A Reconsideration of the Right to Vote, 145 U. PA. L. REV. 893 (1997). 62. See, e.g., Dunn v. Blumstein, 405 U.S. 330, 336 (1972); Reynolds, 377 U.S. at 562; Yick Wo v. Hopkins, 118 U.S. 356, 370 (1886). 63. See Dunn, 405 U.S. at 337. 64. 424 U.S. 319, 332-49 (1976). 65. See Shane, supra note 1, at 564-66. 66. I disagree with Shane on one subsidiary point, though. He sees Al Gore as an appropriate plaintiff in a due process lawsuit and “elide[s] . . . the question of a candidate’s standing.” Shane, supra note 1, at 562 n.125. I think, for reasons I discuss elsewhere, that using the candidates as the vehicles for assessing what ultimately are the voters’ claims may divert judicial attention from the proper remedial questions. I argue that George W. Bush, for example, was precisely the wrong person to vindicate voters’ equal protection interests because his own remedial desire was to see the manual recounts stopped altogether (since he was ahead), regardless of how many votes remained uncounted. See Pamela S. Karlan, Nothing Personal: The Evolution of the Newest Equal Protection from Shaw v. Reno to Bush v. Gore, 79 N.C. L. REV. 1345 (2001). 67. Daniels v. Williams, 474 U.S. 327, 331 (1986). 68. The most elegant expression of this insight appears in Justice Harlan’s dissent in Poe v. Ullman, 367 U.S. 497, 542-43 (1961): Due process has not been reduced to any formula; its content cannot be determined by reference to any code. The best that can be said is that through the course of this Court’s decisions it has represented the balance which our Nation, built upon postulates of respect for the liberty of the individual, has struck between that liberty and the demands of organized society . . . , having regard to what history teaches are the traditions from which it developed as well as the traditions from which it broke. That tradition is a living thing. See also, e.g., Planned Parenthood v. Casey, 505 U.S. 833, 850 (1992) (opinion of O’Connor, Kennedy, and Souter, JJ.) (adopting Justice Harlan’s formulation). 598 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:587 ing electors that over the last two centuries has fallen into desuetude. The Fourteenth Amendment has simply evolved beyond the point at which a state can strip citizens of their right to participate in choosing the President. The Court’s voting cases do, however, raise a complicating factor for procedural due process analysis. Put simply, the Court has never precisely defined the “right to vote” to which strict scrutiny applies. Strict scrutiny, if applied across the board, would invalidate many practices that courts have consistently upheld, from somewhat restrictive registration requirements69 and absentee-ballot laws to bans on write-in voting70 and other sorts of limitations on ballot access.71 Instead, what the Court has done, in effect, is to recognize two sorts of “right to vote”: a “core” right to which strict scrutiny attaches and a less fundamental “right to vote” to which a shifting-scale sort of intermediate scrutiny applies. In cases that fall in this second category, the Court asks whether a sufficiently weighty state interest is served by a reasonable and nondiscriminatory restriction on voters or voting.72 The case law, it turns out, does not give real guidance on a critical question in Bush v. Gore for any form of due process analysis: what are the rights of voters who cast ballots that do not comply with state law? Many of the Justices were skeptical that a voter who failed to punch the chad out completely had cast a legal vote in the first place.73 If the votes that machine counts failed to pick up were not legally cast as a matter of Florida law, then a state’s decision to use a tabulation process that fails to capture them arguably causes no due process problem, since no state-created liberty interest is implicated. My own view, which Shane shares, is that, as a matter of Florida law, the “clear intent of the voter” standard meant that many of the ballots that were out of strict compliance with Florida law were nonetheless legal votes. Thus, for me, the due process analysis proceeds as Shane suggests. But it is important at least to notice the countervailing position: if ballots with pregnant or dimpled chads or 69. See, e.g., Marston v. Lewis, 410 U.S. 679, 679-81 (1973) (upholding Arizona’s decision to cut off registration fifty days before state and local elections, despite the fact that for federal elections only a thirty-day cutoff is permitted, see 42 U.S.C. § 1973aa-1 (1994)). 70. See Burdick v. Takushi, 504 U.S. 428 (1992) (upholding Hawaii’s refusal to count write-in votes). 71. For a fuller treatment of these cases, see THE LAW OF DEMOCRACY, supra note *, at 362-73, 418-27. 72. See Timmons v. New Party, 520 U.S. 351, 357-59 (1997) (describing the standard). 73. See Bush v. Gore, 531 U.S. 98, 119 (2000) (Rehnquist, C.J., joined by Scalia & Thomas, JJ., concurring). At oral argument, Justice O’Connor also expressed this concern: “[W]hy isn’t the standard [for deciding if a ballot contains a legal vote] the one that voters are instructed to follow, for goodness sakes? I mean, it couldn’t be clearer. I mean, why don’t we go to that standard?” Tr. of Oral Argument at 58, Bush v. Gore, 531 U.S. 98 (2000) (No. 00-949), available at http://election2000.stanford.edu/949trans.pdf. 2001] UNDULY PARTIAL 599 other forms of noncompliance do not contain legally cast votes, then a recount process that includes them might infringe upon the voting rights of those citizens who did comply with the state’s requirements.74 This sentiment may well have played some role in the Court’s decision. That is, the Court might have decided the equal protection claim the way that it did only because it thought there was little risk of there being a meaningful number of uncounted but legal votes or because it thought that the Florida process was so out of control that it was likely to produce a less fair final accounting. I think those assumptions are wrong. I suspect that the Court’s willingness to indulge in either might be a manifestation of what Paul Brest memorably termed “selective sympathy and indifference.”75 That is, the five Justices in the majority might have been affected by the fact that the voters whose votes were not being included were Democrats who simply couldn’t follow the rules, for goodness sakes!, and were being bailed out by a Democrat-dominated state supreme court. Without some substantive conception of the right to vote, then, pure procedural due process theory can only take us part of the way toward explaining why the Supreme Court’s equal protection decision is so bankrupt. III. DUE PROCESS AND EQUAL PROTECTION AS RHETORICAL STRATEGIES There was another due process claim lurking in Bush v. Gore. As Shane notes, Governor Bush also argued that the manual recounts ordered by the Florida Supreme Court violated the Due Process Clause.76 That claim was never really addressed, at least as a matter of due process. But it was embraced de facto by the Chief Justice and Justices Scalia and Thomas, whose Article II-based concurrence offered a presidential election-specific version of the argument that changing electoral rules in midstream is unconstitutionally unfair to candidates and voters.77 I don’t propose to analyze Bush’s due process claim here. Rick Pildes’s Article contains a discussion that reflects our collaboratively 74. See Roe v. Alabama, 43 F.3d 574, 581 (11th Cir. 1995) (counting absentee ballots that did not strictly comply with various requirements of Alabama law would “dilute the votes of those voters who met the requirements of [the absentee ballot law] as well as those voters who actually went to the polls on election day”). 75. Paul Brest, Foreword: In Defense of the Antidiscrimination Principle, 90 HARV. L. REV. 1, 7 (1976). 76. Shane, supra note 1, at 570. 77. See Bush v. Gore, 531 U.S. at 112-15. 600 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:587 developed perspective.78 Faced with an unprecedented situation and an ambiguous and contradictory statute, it would be difficult for any interpretation of state constitutional law to rise to the level of a due process violation. Rather, I want to consider why, with all these Due Process Clause claims in the air, the Court employed the Equal Protection Clause instead. Disappearing Democracy suggests the beginning of the answer: any reference to due process could have made it uncomfortably clear that the Court was “apply[ing] the Equal Protection Clause in a way that [would] disenfranchise thousands of Florida voters.”79 Only by ignoring the substantive content of the right to vote and focusing instead solely on the comparative treatment of ballots could the Court reach its desired outcome. But the problems with the per curiam go beyond that. In the end, the decision to stop the recount had virtually nothing to do with equal protection. It vindicated no identifiable voter’s interests.80 The form of equality it created was empty: it treated all voters whose ballots had not already been tabulated the same, by denying any of them the ability to have his ballot counted.81 And its remedy perpetuated other forms of inequality that were far more severe: between voters whose ballots were counted by the machine count and voters whose ballots were not, and even between voters in counties that performed timely manual recounts (like Volusia and Broward) and voters in other counties. Even if there had been an equal protection problem with aspects of the procedure ordered by the Florida Supreme Court, that would not have justified the remedy the U.S. Supreme Court ordered. To stop the recount, the per curiam essentially smuggled in through the back door the Article II rationale advanced explicitly by Chief Justice Rehnquist’s concurrence: any constitutionally acceptable recount would require disregarding the Florida Legislature’s presumed interest in obtaining the safe-harbor benefits of the Electoral Count Act.82 So why did the per curiam insist on relying on the Equal Protection Clause? I think the decision was political, in the broad sense of the word. The Court was trying to wrap its decision in the mantle of its most popularly and jurisprudentially successful intervention into the po78. See Richard H. Pildes, Judging “New Law” in Election Disputes, 29 FLA. ST. U. L. REV. 691 (2001). See generally WHEN ELECTIONS GO BAD, supra note *, at 5-19. 79. Shane, supra note 1, at 538. 80. For a full discussion of this point, see Karlan, supra note 66; see also Pamela S. Karlan, Equal Protection: Bush v. Gore and the Making of a Precedent, in THE UNFINISHED ELECTION OF 2000, at 159, 185-93 (Jack N. Rakove ed., 2001). 81. See Bush v. Gore, 531 U.S. at 110-11. 82. Id. at 111-15 (Rehnquist, C.J., concurring). 2001] UNDULY PARTIAL 601 litical process: the one-person, one-vote cases. This is a familiar strategy. Consider Planned Parenthood v. Casey,83 the case in which the Court reaffirmed the central right to reproductive autonomy recognized in Roe v. Wade.84 The joint opinion written by Justices O’Connor, Kennedy, and Souter invoked another iconic Equal Protection Clause case, Brown v. Board of Education.85 It too treated the responsibility of articulating binding principles of constitutional law as an unsought responsibility. And it saw a special dimension “present whenever the Court’s interpretation of the Constitution calls the contending sides of a national controversy to end their national division by accepting a common mandate rooted in the Constitution.”86 It identified only two such occasions “in our lifetime, . . . the decisions of Brown and Roe.”87 Perhaps the Supreme Court saw Bush v. Gore as a third such occasion. Once again, the Court was asking the nation to end its close division by accepting a common mandate rooted in the Constitution and accepting a judicial resolution.88 And as between the Equal Protection Clause—source of some of the Supreme Court’s finest moments—and the other contenders, it was no contest. If the Supreme Court was going to stop the recount, it had to use a constitutional provision with a pedigree. The Equal Protection Clause provided exactly that. Moreover, it allowed the Court to invoke the specter of unfair treatment of voters, whereas the other available constitutional contenders protected either the prerogative of state legislatures (Article II, Section 1) or, even worse, the interests of candidate George W. Bush (the Due Process Clause as it was actually raised in Bush v. Gore). CONCLUSION Shane concludes his Article with a discussion of the political question doctrine. He argues that “Article II and the Twelfth Amendment are readily interpretable as embodying a textually demonstrable commitment to Congress of the power to resolve all issues related to the proper tabulation of electoral votes.”89 While I agree with him that the Constitution confides the question of how to count electoral votes to Congress, I do not think that necessarily makes Bush v. Gore nonjusticiable. Even if the question of which slate of electors to accept is ultimately for Congress to decide, that cannot dispose of the 83. 84. 85. 86. 87. 88. 89. 505 U.S. 833 (1992). 410 U.S. 113 (1973). 347 U.S. 483 (1954). Casey, 505 U.S. at 867. Id. Bush v. Gore, 531 U.S. 98, 110 (2000). Shane, supra note 1, at 581-82. 602 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:587 antecedent question of whether a state has conducted its election process in compliance with the Equal Protection and Due Process Clauses. I imagine that Shane would be disturbed if, faced with the due process claim he identifies—the hypothetical case of Gore v. Harris—the Supreme Court were to say to excluded voters, “go complain to Congress; we are unwilling to step in even if a state chooses to use unreliable methods of counting votes.” Disappearing Democracy is ultimately an Article about why it matters through which lens we examine a problem. As Shane notes, it isn’t just a question of vocabulary.90 It actually affects how courts see the facts and resolve the issues. But just as how we denominate a claim can matter, so too can what we call an institution. We are not the only country recently to have faced the question whether we could hold our presidential inauguration as scheduled. Iran also had that problem. Its inauguration was postponed because the Iranian Constitution requires that the ceremony be carried out in the presence of “all the members” of the Guardians Council, and the reformist parliament was at loggerheads with the conservative judiciary over voting to fill the vacant seats.91 The Iranian Constitution provides for something our Constitution doesn’t: an “Expediency Council.”92 Faced with the problem, the Expediency Council simply rammed through a settlement: the seats would be filled by a plurality vote, rather than the previously required majority vote, and the inauguration could go forward. Our Constitution makes no mention of an Expediency Council. Both in describing the right to vote in presidential elections and in reviewing how Florida sought to protect that right, our Supreme Court acted more like an Expediency Council than a principled judicial body. Ultimately, it denied all of us due process of law. 90. See id. at 551-52. 91. For an account of this controversy, see Neil MacFarquhar, Iran Parliament Clears Way for Khatami’s 2nd Inauguration, N.Y. TIMES, Aug. 8, 2001, at A3. 92. Id. THE PRESIDENTIAL ELECTION DISPUTE, THE POLITICAL QUESTION DOCTRINE, AND THE FOURTEENTH AMENDMENT: A REPLY TO PROFESSORS KRENT AND SHANE ROBERT J. PUSHAW, JR.* I. BUSH V. GORE: BAKER V. CARR WITH A CONSERVATIVE TWIST ......................... A. The Prelude to Baker.................................................................................... B. The Baker Breakthrough ............................................................................. C. Baker Consolidated ...................................................................................... D. Bush v. Gore ................................................................................................. II. THE PRESIDENTIAL ELECTION DISPUTE: “POLITICAL” OR “LEGAL”?.................. A. The “Law” Governing Political Questions ................................................... B. Did Bush v. Gore Present Nonjusticiable Issues? ....................................... 1. The Justices’ Cryptic Opinions .............................................................. 2. Justiciability With Deferential Review ................................................. 3. Application of the Proper Standard of Review in Bush v. Gore .......... III. FOURTEENTH AMENDMENT ISSUES ..................................................................... A. Equal Protection ........................................................................................... B. Due Process and Related Concerns .............................................................. CONCLUSION ........................................................................................................ 605 606 607 608 610 611 611 612 612 613 616 619 619 623 623 Like millions of Americans, my wife Trish and I remained glued to our televisions throughout the 2000 presidential election crisis in Florida. Because Trish is not an attorney, she frequently asked for my legal opinion on the various judicial proceedings. For instance, she believed that the United States Supreme Court was so “peeved at the shenanigans” that it would intervene.1 I assured her, based upon my extensive study of justiciability,2 that the Justices would decline jurisdiction.3 I also confidently predicted that, in the unlikely event * Earl F. Nelson Professor, University of Missouri School of Law; Visiting Professor, Pepperdine University School of Law. J.D., Yale University, 1988. Thanks to Tracey George, Grant Nelson, and the symposium participants for their comments on this paper. 1. Trish grew up in East Africa and Ireland, and she often uses phrases that sound like they belong in an English drawing-room comedy. 2. See, e.g., Robert J. Pushaw, Jr., Why the Supreme Court Never Gets Any “Dear John” Letters: Advisory Opinions in Historical Perspective, 87 GEO. L.J. 473 (1998) [hereinafter Pushaw, Advisory]; Robert J. Pushaw, Jr., Congressional Power Over Federal Court Jurisdiction: A Defense of the Neo-Federalist Interpretation of Article III, 1997 BYU L. REV. 847, 864-73 [hereinafter Pushaw, Congressional Power]; Robert J. Pushaw, Jr., Justiciability and Separation of Powers: A Neo-Federalist Approach, 81 CORNELL L. REV. 393 (1996) [hereinafter Pushaw, Justiciability]; Robert J. Pushaw, Jr., Article III’s Case/Controversy Distinction and the Dual Functions of Federal Courts, 69 NOTRE DAME L. REV. 447 (1994) [hereinafter Pushaw, Case/Controversy]. 3. First, Bush did not seem to have third-party standing to allege that Florida’s nonuniform standards for counting votes violated the Equal Protection rights of state voters. Second, the case did not appear to be ripe because the Constitution and federal statutes provided that presidential election disputes must be resolved, at least initially, in the states and then Congress. Third, success in those political institutions would have mooted 603 604 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:603 the Court reached the merits, it could not possibly hold that the Equal Protection Clause required states to administer their elections without discrepancies, because such inconsistencies are one price we pay for our federal system.4 As usual, I was wrong, and my wife was right. I was focusing on legal abstractions, whereas she was attuned to political realities. A rather obvious point dawned on me: How could judicial decisions on elections, the very focus of politics, not be politicized? And yet there are established, seemingly neutral principles of constitutional “law” that govern judicial review of elections. Peter Shane and Harold Krent are surprised (as I initially was) by the Court’s failure to apply that law in Bush v. Gore.5 Upon further reflection, however, I concluded that the Rehnquist Court was following the law—specifically, the landmark precedent in this area, Baker v. Carr.6 As in Baker, the majority in Bush (1) determined that the political process would not remedy a perceived electoral crisis in a Southern state; (2) ignored principles of justiciability and federalism that generally prevented Article III courts from interfering with state electoral matters; and (3) made up new “standards” under the Equal Protection Clause and held that the state had violated them.7 In each case, dissenting Justices asserted that the majority’s entanglement in the political thicket would erode respect for the Court as the impartial guardian of the rule of law.8 Understanding this connection between Bush v. Gore and Baker v. Carr helps to explain my reaction to the two main arguments presented by my fellow panelists. Bush’s claim. Fourth, the flexible political question doctrine enabled the Court to interpret the Constitution as granting Congress final authority to decide the matter. In short, the Court could have declined to review the case on several justiciability grounds. Robert J. Pushaw, Jr., Bush v. Gore: Looking at Baker v. Carr in a Conservative Mirror, 18 CONST. COMM. (forthcoming 2001) [hereinafter Pushaw, Conservative Mirror] (describing the potential application of the justiciability doctrines in Bush). 4. See Bush v. Gore, 531 U.S. 98, 123 (2000) (Stevens, J., dissenting) (arguing that the Equal Protection Clause did not authorize the Court to question the states’ power either to fix standards for determining whether a vote has been legally cast or to allow counties to design their own balloting systems). 5. See Peter M. Shane, Disappearing Democracy: How Bush v. Gore Undermined the Federal Right to Vote for Presidential Electors, 29 FLA. ST. U. L. REV. 535, 535, 579, 584 (2001) (characterizing the majority’s opinion as “extraordinary,” “stunning,” and “astonishing”); Harold J. Krent, Judging Judging: The Problem of Second-Guessing State Judges’ Interpretation of State Law in Bush v. Gore, 29 FLA. ST. U. L. REV. 493, 493, 497 (2001) (deeming the Court’s decision “surprising,” “astonishing,” and “startling”). 6. 369 U.S. 186 (1962). I have compared these two cases in exhaustive detail in Pushaw, Conservative Mirror, supra note 3. 7. See infra notes 20-50 and accompanying text (fleshing out this analogy between Baker and Bush). 8. See infra notes 31, 49 and accompanying text (summarizing and comparing these dissenting opinions). 2001] REPLY TO KRENT AND SHANE 605 First, I agree with Professor Krent that the political question doctrine did not preclude the Court from deciding the presidential election dispute.9 Baker turned this doctrine into a prudential case-bycase judgment call, and ever since the Court has almost always exercised this unbridled discretion in favor of intervening in heavily politicized disputes—including several that involved state decisions affecting federal elections.10 Professor Shane is undoubtedly right that the Court could have avoided decision on political question grounds,11 but that is always true, because Baker leaves this determination entirely to the Justices’ discretion. Although I am sympathetic to Shane’s position that the Court should have declined to decide the case, that is really an argument for a pre-Baker regime of judicial restraint, not against the Court’s application of the current political question doctrine in Bush. Second, Professor Shane is correct that, contrary to the opinion of seven Justices in Bush, the Equal Protection Clause does not grant candidates a right to uniform standards in counting votes.12 But neither does that Clause require legislative representation to be based solely on population, as Baker held and Shane assumes.13 In any event, Baker provides a “precedent” for the Court to ignore that Clause’s text, drafting and ratification history, implementation by Congress and the states, and cases interpreting it. Therefore, although I believe that both Baker and Bush were wrongly decided, the latter is more defensible simply because it departed less radically from the Court’s existing practice. I will begin by outlining my comparison of these two cases. Against this backdrop, I will evaluate the main themes of Professors Krent and Shane. I. BUSH V. GORE: BAKER V. CARR WITH A CONSERVATIVE TWIST The Court employed the same decisionmaking process in Bush as it did in Baker, albeit to reach a conservative rather than liberal result. 9. See Krent, supra note 5, at 511. 10. See infra notes 23-26, 32-44, 52-59 and accompanying text (discussing the relevant passages from Baker and subsequent cases). 11. See Shane, supra note 5, at 580-82. 12. See id. at 552. 13. Baker v. Carr, 369 U.S. 186, 226 (1962); see also Shane, supra note 5, at 548 (citing the apportionment cases approvingly as part of the “democratic trajectory” of constitutional law since 1868). 606 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:603 A. The Prelude to Baker Baker reversed, sub silentio, an unbroken line of precedent dating back to 1849.14 In Luther v. Borden,15 the Court accepted the judgment of Congress and the President that Rhode Island satisfied Article IV, Section 4, which provides that “the United States shall guarantee to every State in this Union a republican form of Government.”16 Although Luther did not hold that all claims under the Guarantee Clause raised political questions,17 the Court created such an absolute prohibition in 1912.18 Most pertinently, beginning in 1916, it rejected nearly all constitutional attacks on state legislative apportionment schemes based upon either nonjusticiability or equitable discretion to decline jurisdiction because of federalism concerns.19 14. For an extensive analysis of these decisions, see Pushaw, Conservative Mirror, supra note 3. 15. 48 U.S. (7 How.) 1 (1849). 16. Id. at 42-45. 17. For instance, the Court acknowledged that a state could declare martial law temporarily to meet threats to its very existence (as Rhode Island had done), and thus declined “to inquire to what extent, nor under what circumstances, that power may be exercised by a State” before a Guarantee Clause violation would occur. Id. at 45. By negative implication, the Court could hear complaints that a state did allegedly violate that Clause, such as by declaring permanent martial law. Indeed, numerous cases after Luther did exactly that. See New York v. United States, 505 U.S. 144, 184-85 (1992) (citing these decisions and acknowledging that the Court had, over the previous eighty years, misconstrued Luther as barring all claims under the Republican Form of Government Clause). At the time Baker was decided, however, all of the Justices assumed that such allegations raised political questions. 18. Pac. States Tel. & Tel. Co. v. Oregon, 223 U.S. 118, 142-51 (1912) (holding nonjusticiable a complaint that a state government had become unrepublican by enacting a law through the initiative rather than the legislative process). Subsequent cases that deemed Guarantee Clause cases to involve political questions include Highland Farms Dairy v. Agnew, 300 U.S. 608, 612 (1937); Ohio ex rel. Bryant v. Akron Metropolitan Park District, 281 U.S. 74, 79 (1930); and Mountain Timber Co. v. Washington, 243 U.S. 219, 234-35 (1917). 19. The seminal case is Ohio ex rel. Davis v. Hildebrant, 241 U.S. 565 (1916) (refusing to decide the political question of whether the invalidation of a state reapportionment statute by referendum violated the Republican Form of Government Clause). In Colegrove v. Green, 328 U.S. 549 (1946), the Court dismissed a claim that a state law had violated the Constitution by creating congressional districts that did not account for population shifts to metropolitan areas. Justice Frankfurter and two colleagues found a political question. Id. at 552-56. Justice Rutledge assumed justiciability, but invoked equitable discretion to refrain from intervening because of the sensitive political issues involved in drawing congressional districts. Id. at 564-66 (Rutledge, J., concurring). Three dissenters maintained that the statute violated the Equal Protection Clause. Id. at 566-74 (Black, J., dissenting). Two Justices did not participate. The Court swiftly applied Colegrove in a pair of brief per curiam opinions and declined to entertain constitutional challenges to state electoral laws. See Colegrove v. Barrett, 330 U.S. 804 (1947); Turman v. Duckworth, 329 U.S. 829 (1946). In MacDougall v. Green, 335 U.S. 281 (1948), however, a five-man majority held, on the merits, that the Equal Protection Clause did not deprive states of their authority to give rural areas disproportionate influence in the nomination process. Id. at 283-84. Nonetheless, two years later the Court reverted back to the Colegrove approach by dismissing, on “political question” and “equitable discretion” grounds, a Fourteenth Amendment claim 2001] REPLY TO KRENT AND SHANE 607 B. The Baker Breakthrough Baker v. Carr 20 addressed a critical electoral problem that could not readily be solved by the ordinary political process.21 Like most states, Tennessee apportioned legislative districts to account for a host of demographic, geographical, political, economic, and historical interests, often with the special goal of preserving the electoral power of conservative agricultural communities against the burgeoning (and more liberal) metropolitan areas.22 The Court refashioned the political question doctrine as based exclusively upon separation of powers (not federalism)23 and as necessitating a “case-by-case inquiry”24 weighing several factors, most importantly whether there was “a textually demonstrable constitutional commitment of the issue to a coordinate political department.”25 The Court concluded that the plaintiffs’ claim, although nonjusticiable under the Guarantee Clause, could be reviewed under the Equal Protection Clause.26 Instead of setting forth specific legal principles to govern cases involving apportionment controversies, however, the Court merely declared: that Georgia’s voting scheme discriminated against urban dwellers. See South v. Peters, 339 U.S. 276, 277 (1950). Thereafter, the Court applied a similar analysis in refusing to hear all constitutional challenges to state apportionment laws in a series of terse per curiam orders that usually rested on a bare citation to Colegrove. See, e.g., Matthews v. Handley, 361 U.S. 127 (1959); Radford v. Gary, 352 U.S. 991 (1957); Kidd v. McCanless, 352 U.S. 920 (1956); Anderson v. Jordan, 343 U.S. 912 (1952); Remmey v. Smith, 342 U.S. 916 (1952); Tedesco v. Bd. of Supervisors, 339 U.S. 940 (1950). Thus, although in Colegrove only three of seven Justices found the constitutional complaint to be a political question, the case came to stand for the proposition that federal judges should not decide such claims. See Baker v. Carr, 369 U.S. 186, 252 (1962) (Clark, J., concurring) (arguing that the Court had read Colegrove as based upon nonjusticiability and a rejection of any possible Fourteenth Amendment issue). Indeed, the only time the Court exercised jurisdiction was to protect blacks from discriminatory state electoral laws, which violated the Fifteenth (not the Fourteenth) Amendment. See, e.g., Gomillion v. Lightfoot, 364 U.S. 339 (1960). 20. 369 U.S. 186 (1962). 21. For a more thorough explanation of the various opinions in Baker and its progeny, see Pushaw, Conservative Mirror, supra note 3. 22. Baker, 369 U.S. at 187-95 (stressing that Tennessee had failed to amend its 1901 apportionment law, despite the huge subsequent population shift to urban areas); id. at 268-69, 301-24 (Frankfurter, J., dissenting) (emphasizing the multiplicity of interests involved in apportioning legislatures); see also Gus Tyler, Court Versus Legislature, 27 LAW & CONTEMP. PROBS. 390, 395-98 (1962) (arguing that rural overrepresentation maintained the political power of conservatives against more liberal urban and suburban citizens, whose numbers had exploded). 23. Baker, 369 U.S. at 210, 226. The majority attempted to distinguish numerous cases holding that, to preserve federalism, the Court would not interfere with state apportionment schemes. Id. at 231-37. 24. Id. at 210. 25. Id. at 211, 217. 26. Id. at 208-37. 608 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:603 Judicial standards under the Equal Protection Clause are well developed and familiar, and it has been open to courts since the enactment of the Fourteenth Amendment to determine, if on the particular facts they must, that a discrimination reflects no policy, but simply arbitrary and capricious action.27 In separate dissents, Justices Frankfurter and Harlan made three related points.28 First, the majority had repudiated uniform precedent treating apportionment as a political question under the Republican Form of Government Clause, which had reflected the Court’s former respect for state government autonomy in balancing the myriad policy considerations involved.29 Second, neither the Equal Protection Clause nor any other constitutional provision indicated that representation must be based on population alone.30 Third, the majority’s radical intrusion into internal state affairs would compromise public confidence in the Court’s independence and authority, which rested largely upon its detachment from wholly political disputes.31 C. Baker Consolidated The Court quickly translated Baker’s jurisdictional holding into substantive law. Most importantly, Reynolds v. Sims32 established that the Equal Protection Clause requires state legislatures to be apportioned according to equality of population “as nearly as is practicable.”33 In the earlier case of Gray v. Sanders,34 the Court had applied this “one person, one vote” rule to invalidate a state districting framework for elections to statewide executive offices and the federal Senate.35 In Wesberry v. Sanders,36 the Justices found the same prin27. Id. at 226. 28. Id. at 266-330 (Frankfurter, J., dissenting); id. at 330-49 (Harlan, J., dissenting). 29. Id. at 266-69, 277-324 (Frankfurter, J., dissenting); id. at 332-34, 338-39 (Harlan, J., dissenting). 30. Id. at 300-23 (Frankfurter, J., dissenting); id. at 332-36, 338-39 (Harlan, J., dissenting). 31. Id. at 267-69, 324 (Frankfurter, J., dissenting); id. at 339-40 (Harlan, J., dissenting). They also deplored the Court’s refusal to mention any Equal Protection principles and its suggestion that the mere prospect of reapportionment by federal judges under amorphous Equal Protection standards might spur state legislatures to take action without the need for further judicial supervision. See id. at 267-69, 327-28 (Frankfurter, J., dissenting); id. at 339 (Harlan, J., dissenting). Baker, then, is an odd form of advisory opinion. See Pushaw, Advisory, supra note 2, at 477-91 (discussing the historical arguments for and against such opinions). The Court advised state legislatures that they were unconstitutional, failed to tell them exactly why, and threatened them with dire consequences if they did not reapportion. 32. 377 U.S. 533 (1964). 33. Id. at 559. The Court formulated that standard in the course of a lengthy analysis. See id. at 554-87. 34. 372 U.S. 368 (1963). 35. Id. at 370-80. 36. 376 U.S. 1 (1964). 2001] REPLY TO KRENT AND SHANE 609 ciple lurking in Article I, Section 2, Clause 1, which addresses districting for the House of Representatives.37 Finally, the Court extended the principle of these apportionment decisions—that all voters must have an equal opportunity to participate in elections—to voter qualifications (for example, by striking down wealth-based classifications such as poll taxes).38 In all of these cases, Justice Harlan dissented because these holdings (1) lacked any foundation in the Constitution’s text, history, or precedent, which recognized the states’ independence in apportioning legislative districts and setting voter qualifications, and (2) abandoned the Court’s appropriately limited role in our federal system.39 Whatever its constitutional provenance, the “one person, one vote” maxim resonated with Americans, who had become predominantly urban and suburban and thus disliked state laws that magnified rural electoral power.40 Despite some criticism of Baker immediately after its release,41 scholars swiftly embraced it.42 Moreover, the Court has never wavered from its position that electoral disputes are justiciable and that the “one person, one vote” standard should be applied to resolve them.43 Today both Justices and law professors tend to un37. Id. at 7-18. 38. See, e.g., Harper v. Va. Bd. of Elections, 383 U.S. 663, 665-70 (1966). 39. Reynolds v. Sims, 377 U.S. 533, 615-24 (1964) (Harlan, J., dissenting); Gray, 372 U.S. at 382-90 (Harlan, J., dissenting); Wesberry, 376 U.S. at 20-50 (Harlan, J., dissenting); Harper, 383 U.S. at 680-86 (Harlan, J., dissenting). 40. See, e.g., Jo Desha Lucas, Legislative Apportionment and Representative Government: The Meaning of Baker v. Carr, 61 MICH. L. REV. 711, 804 (1963) (conceding Baker’s popularity but decrying its lawlessness). Although the Court in Baker and its progeny espoused democratic ideals, in one sense these cases are the antithesis of democracy: Seven men sitting at the apex of the most unrepresentative branch of government imposed their personal theory of democracy upon every state in the union. 41. See, e.g., Alexander M. Bickel, The Durability of Colegrove v. Green, 72 YALE L.J. 39, 43-45 (1962) (urging the Court to abstain because of the impossibility of developing principled legal standards to determine the validity of apportionment laws); Robert G. McCloskey, The Supreme Court, 1961 Term—Foreword: The Reapportionment Case, 76 HARV. L. REV. 54 (1962). 42. See, e.g., Charles L. Black, Inequities in Districting for Congress: Baker v. Carr and Colegrove v. Green, 72 YALE L.J. 13 (1962) (applauding the Court for remedying injustices in apportionment by recognizing a new fundamental right to equal treatment in voting). Within a few years the legal academy generally had come around to the view of Professor Black and many others. See Pushaw, Conservative Mirror, supra note 3 (summarizing the scholarly reaction to the reapportionment cases). Moreover, Professor Ely bolstered the intellectual credibility of Baker and its progeny by arguing that judicial review should be exercised only in situations where representative government cannot be trusted because of breakdowns in the political process, as occurred when entrenched state legislators refused to reapportion. See JOHN HART ELY, DEMOCRACY AND DISTRUST: A THEORY OF JUDICIAL REVIEW 73-183 (1980). 43. Indeed, the Court has reiterated these principles so often that they no longer generate much dissent. See, e.g., Dep’t of Commerce v. Montana, 503 U.S. 442, 456-59 (1992) (unanimously allowing a challenge to Congress’s choice of a method for apportioning congressional districts among states under Article I, Section 2, but concluding that the statute passed constitutional muster). 610 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:603 questioningly accept Baker as a landmark Warren Court decision that established principles of constitutional “law,” not a mere liberal political coup.44 D. Bush v. Gore Instead of rehashing Bush in Gore-y detail, I will highlight five critical aspects of its similarity to Baker. First, a majority of the Justices saw a national electoral emergency—the 2000 presidential election impasse in Florida—that seemed to defy a clean political solution.45 Second, the Court flicked away justiciability concerns, even though the Constitution appears to entrust the resolution of presidential election disputes to the states and Congress, at least in the first instance.46 Third, the majority concluded that principles of federalism did not require strong deference to state officials (including judges) in interpreting and implementing their election laws.47 Fourth, all but two Justices discovered an unprecedented Equal Protection Clause “right” to uniform standards in counting ballots.48 Fifth, the dissenting Justices accused the majority of abandoning judicial restraint by failing to dismiss on political question grounds and by ignoring precedent concerning judicial federalism and Equal Protection.49 Despite all these similarities to Baker, law professors (including Shane and Krent) have overwhelmingly expressed surprise and out- 44. See, e.g., ERWIN CHEMERINSKY, INTERPRETING THE CONSTITUTION 127, 141 (1987); CASS R. SUNSTEIN, THE PARTIAL CONSTITUTION 93, 98, 260 (1993); see also J.M. Balkin & Sanford Levinson, The Canons of Constitutional Law, 111 HARV. L. REV. 964, 997-1002 (1998) (noting that liberal scholars have canonized Baker and other “beloved” Warren Court precedents and have not tolerated any suggestion that these cases might be incorrect). A few scholars have challenged this orthodoxy. See, e.g., Samuel Issacharoff, Political Judgments, 68 U. CHI. L. REV. 637, 639-40 (2001) (arguing that the Baker Court failed to explain coherently its Equal Protection holding, its abandonment of the political question doctrine, and its notion that federal judges could avoid getting sullied in adjudicating electoral disputes). 45. See Bush v. Gore, 531 U.S. 98, 100-03 (2000) (detailing the relevant facts and procedural history). The Court, of course, said that it was intervening to uphold the law, not to break a political deadlock. See id. at 111. 46. See id. at 111. 47. See id. at 103-11; see also id. at 111-22 (Rehnquist, C.J., concurring) (charging the Florida Supreme Court with amending, not merely construing, its legislature’s directives regarding the appointment of electors). 48. See id. at 102-08 (per curiam opinion joined by Rehnquist, C.J., and O’Connor, Scalia, Kennedy, and Thomas, JJ.). Justices Souter and Breyer agreed with the Equal Protection holding, but they would have dismissed the case as nonjusticiable. Id. at 133-34 (Souter, J., dissenting); id. at 143-46 (Breyer, J., dissenting). 49. See id. at 123-29 (Stevens, J., dissenting); id. at 135-44 (Ginsburg, J., dissenting). Of course, I recognize certain distinctions between Baker and Bush, but I find them to be largely irrelevant for reasons outlined in Pushaw, Conservative Mirror, supra note 3. 2001] REPLY TO KRENT AND SHANE 611 rage at Bush.50 This reaction reminds me of the scene in Casablanca where Louis proclaims that he is “shocked” to discover that people are gambling in Rick’s Cafe and then pockets a payoff. The most plausible explanation for this response is that Bush reached a politically conservative result that most of the legal academy finds distasteful, whereas Baker produced a politically liberal result that they favor. But the process of decisionmaking in both cases was nearly identical. With this background in mind, I will analyze the two major issues explored by my co-panelists. II. THE PRESIDENTIAL ELECTION DISPUTE: “POLITICAL” OR “LEGAL”? Although Bush potentially implicated every category of justiciability,51 I will follow the lead of Professors Shane and Krent by focusing on one—the political question doctrine. A. The “Law” Governing Political Questions In Baker v. Carr,52 the Court declared: Prominent on the surface of any case held to involve a political question is found a textually demonstrable constitutional commitment of the issue to a coordinate political department; or a lack of judicially discoverable and manageable standards for resolving it; or the impossibility of deciding without an initial policy determination of a kind clearly for nonjudicial discretion; or the impossibility of a court’s undertaking independent resolution without expressing lack of respect due coordinate branches of government; or an unusual need for unquestioning adherence to a political decision already made; or the potentiality of embarrassment from multifarious pronouncements by various departments on one question.53 Unfortunately, these six criteria provide little concrete guidance in distinguishing “political” from “justiciable” questions. Most significantly, Articles I and II contain “a textually demonstrable commitment” of all legislative and executive powers to Congress and the President, but that surely does not immunize their exercise from judicial review. Rather, the pivotal consideration must be whether the Constitution leaves an issue to a political branch for final decision. That determination, however, cannot be made by consulting the Constitution’s text, which nowhere explicitly mentions judicial re50. To take the most publicized example, well over 500 legal academics signed an advertisement in The New York Times on January 13, 2001, asserting that the Court “used its power to act as political partisans, not judges of a court of law. . . . By taking power from the voters, the Supreme Court has tarnished its own legitimacy.” 51. See supra note 3 and accompanying text. 52. 369 U.S. 186 (1962). The following discussion of Baker incorporates the conclusions I presented in Pushaw, Justiciability, supra note 2, at 498-501. 53. Baker, 369 U.S. at 217. 612 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:603 view—much less exceptions to this power. The other five Baker factors are also infinitely elastic.54 In its “case-by-case inquiry”55 weighing these six criteria in sixteen situations, the Court routinely has found even the most politically laden matters to be justiciable56—including every constitutional challenge to electoral decisions.57 Only two questions have been deemed “political”: military training and procedures (Gilligan v. Morgan)58 and impeachment (Walter Nixon v. United States).59 With rare exceptions, then, application of the Baker test results in judicial intervention. Moreover, these decisions are merely the tip of the iceberg, because often the Court summarily asserts jurisdiction over a case that arguably raises political questions without even mentioning Baker. B. Did Bush v. Gore Present Nonjusticiable Issues? 1. The Justices’ Cryptic Opinions Interestingly, no Justice in Bush cited Baker or used the term “political question doctrine.” Nonetheless, the various opinions reveal an understanding of its potential applicability, especially in 54. For instance, many constitutional provisions appear to “lack . . . judicially discoverable and manageable standards” and to require “an initial policy determination of a kind clearly for nonjudicial discretion,” id. at 217, but that has not stopped the Court from interpreting and applying them. Indeed, Baker and its progeny provide a perfect example: The Court discovered in the Equal Protection Clause sufficiently manageable “judicial” standards to measure the validity of apportionment laws that reflected exceptionally complicated “policy determinations.” See supra notes 26-38 and accompanying text (describing these cases). Likewise, any exercise of judicial review potentially will reverse “a political decision already made,” cause the political departments “embarrassment,” and display a “lack of respect.” Baker, 369 U.S. at 217. 55. Id. at 210. 56. Perhaps the most famous example is Powell v. McCormack, 395 U.S. 486 (1969), in which the Court reviewed (and invalidated) the House of Representatives’ exercise of its Article I, Section 5, Clause 1 power to “be the Judge of the qualifications of its own Members.” Id. at 512-50. See Pushaw, Justiciability, supra note 2, at 498-99 (describing Powell and other decisions). 57. Most pertinent are cases involving state government actions that affected federal elections. See Dep’t of Commerce v. Montana, 503 U.S. 442, 456-59 (1992); Wesberry v. Sanders, 376 U.S. 1, 5-7 (1964); Gray v. Sanders, 372 U.S. 368, 370-76 (1963), discussed supra notes 34-37, 43 and accompanying text; see also Davis v. Bandemer, 478 U.S. 109, 118-27 (1986) (upholding the justiciability of a claim of illegal state gerrymandering); Reynolds v. Sims, 377 U.S. 533, 582 (1964) (reaffirming Baker’s ruling that state apportionments were not political questions), discussed supra notes 32-33, 39 and accompanying text (analyzing Reynolds). 58. 413 U.S. 1, 3-10 (1973) (dismissing a complaint that the government’s negligent training of the National Guard had caused the deaths of antiwar protestors at Kent State University). 59. 506 U.S. 224 (1993) (declining to reach the merits of a federal judge’s allegation that Article I, Section 3, Clause 6, which grants the Senate “sole Power to try all Impeachments,” requires a trial before the full Senate—not merely a hearing before a committee). 2001] REPLY TO KRENT AND SHANE 613 discussing two key constitutional provisions. First, Article II, Section 1, Clause 2 provides that “[e]ach State shall appoint [presidential electors] in such [m]anner as the Legislature thereof may direct.”60 Second, the Twelfth Amendment entrusts the counting of electoral votes to Congress, which has made laws necessary and proper to implement that power—specifically including statutes setting forth procedures for presidential election contests.61 Justice Breyer concluded that these provisions granted the states and Congress exclusive authority to resolve such disputes, with Congress’s judgment final and judicially unreviewable.62 Justice Souter apparently agreed.63 The majority, however, dispatched such concerns in two sentences: None are more conscious of the vital limits on judicial authority than are the members of this Court, and none stand more in admiration of the Constitution’s design to leave the selection of the President to the people, through their legislatures, and to the political sphere. When contending parties invoke the process of the courts, however, it becomes our unsought responsibility to resolve the federal and constitutional issues the judicial system has been forced to confront.64 2. Justiciability With Deferential Review It is impossible to say definitively which side is legally correct, because Baker made political question “law” almost entirely dependent on judicial discretion.65 For that reason, the Bush dissenters and Pro60. See, e.g., Bush v. Gore, 531 U.S. 98, 104 (2000) (interpreting this provision as granting state legislatures plenary power in choosing electors); id. at 111-22 (Rehnquist, C.J., concurring) (arguing that the Florida Supreme Court violated Article II by altering its legislature’s directions concerning the appointment of presidential electors). Conversely, the dissenters contended that the Florida court had merely interpreted (not changed) its state’s statutes and that, in any event, Article II did not contemplate review of such decisions by the United States Supreme Court. Id. at 123-29 (Stevens, J., dissenting); id. at 129-35 (Souter, J., dissenting); id. at 135-44 (Ginsburg, J., dissenting); id. at 144-58 (Breyer, J., dissenting). 61. See id. at 153-58 (Breyer, J., dissenting); see also James C. Kirby, Jr., Limitations on the Power of State Legislatures Over Presidential Electors, 27 LAW & CONTEMP. PROBS. 495, 498-500 (1962) (citing all relevant constitutional and statutory provisions). 62. See Bush, 531 U.S. at 153-58 (Breyer, J., dissenting). 63. See id. at 129 (Souter, J., dissenting) (beginning his opinion by asserting that “[t]he Court should not have reviewed” the case and suggesting that any political problems might have been resolved by Congress). An essay written long before Bush, and thus not politically biased, argued that the Court would likely apply Baker to rule that the Twelfth Amendment grants to Congress final power to resolve disputes concerning presidential electoral voting. Albert J. Rosenthal, The Constitution, Congress, and Presidential Elections, 67 MICH. L. REV. 1, 26-30 (1968). 64. Bush, 531 U.S. at 111. 65. See supra Parts I.B and I.C. For example, the Twelfth Amendment contains “a textually demonstrable constitutional commitment of the issue [presidential electoral voting] to a coordinate political department [Congress].” Baker v. Carr, 369 U.S. 186, 217 614 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:603 fessor Shane can legitimately criticize the Court only for unwisely exercising this discretion and not (as they charge) for failing to obey the law.66 And, to the extent that the content of this law consists of post-Baker decisions, they uniformly support the Bush majority.67 Professor Krent provides two further, and persuasive, justifications for federal judicial intervention in the particular situation presented in Bush. First, he cites abundant precedent demonstrating that the Justices, to protect federal rights, have often reviewed state tribunals’ interpretation of state law—including their alleged alteration of that law to defeat federal constitutional rights.68 Second, Krent defends the Court’s exercise of jurisdiction to examine the Florida judiciary’s construction of its state’s statutes governing the appointment of presidential electors.69 He shows that the Framers of Article II deliberately vested this appointment power not in Congress but in state legislatures, on the basis that the latter were less likely to form a coalition to extract promises from presidential candidates, were more receptive to state concerns, and were more directly accountable.70 Accordingly, Krent argues that Congress, state executives, or state judges could circumvent a state legislature’s instructions regarding the selection of electors unless the Supreme Court could oversee their actions.71 Krent further contends, however, that (1962). As with all constitutional provisions, however, the Twelfth Amendment is silent as to whether courts may review the exercise of this power. Moreover, the Baker Court made the unprecedented assertion that the political question doctrine concerned only separation of powers, not federalism (apparently for the purpose of justifying its intrusion into a core state political power, the composition of its legislature). Id. at 210, 226. If so, then Article II’s grant to state legislatures of power to appoint electors is simply irrelevant in determining justiciability. 66. See Bush, 531 U.S. at 144 (Breyer, J., dissenting) (“The Court was wrong to take this case.”); id. at 156 (“[F]ederal law [n]either foresees [n]or requires resolution of such a political issue by this Court.”); id. at 157 (characterizing the Court’s intervention as “legally wrong”). Professor Shane claims that the Justices in the majority abandoned their heretofore resolute commitment to the political question doctrine, as exemplified by Nixon. See Shane, supra note 5, at 580-83. But Nixon is the lone exception to the Rehnquist Court’s otherwise perfect record of finding electoral disputes (and many other political matters) to be justiciable. See supra notes 55-59 and accompanying text. Therefore, a more accurate way of expressing this criticism would be to say that conservative champions of judicial restraint should have exercised their virtually boundless discretion afforded by the political question doctrine to decline jurisdiction over the election controversy. 67. See supra notes 55-59 and accompanying text. 68. Krent, supra note 5, at 495 (citing cases). 69. Id. at 507-11. 70. Id. at 509. Professor Shane endorses Justice Breyer’s suggestion that Article II disputes should be left to Congress’s sole and unreviewable discretion under the Twelfth Amendment. See Shane, supra note 5, at 581-82. This scheme, however, would frustrate the Framers’ deliberate decision to vest the power to appoint electors in state legislatures, not Congress. See Krent, supra note 5, at 509. 71. Krent, supra note 5, at 511. 2001] REPLY TO KRENT AND SHANE 615 this precedent obliged the Bush majority to accept the Florida high court’s interpretation.72 I agree with Professor Krent that the presidential election dispute presented a justiciable issue but that the Court should have been more deferential. I reach this conclusion, however, through somewhat different reasoning. My general approach builds upon Hamilton’s insight that judicial review reflects the “natural presumption” that political officials cannot be “the constitutional judges of their own powers,” but that this presumption can be rebutted by “particular provisions in the [C]onstitution”73 examined in light of their drafting and ratification history, purpose, and place in the Constitution’s political design.74 An example of provisions that overcome the strong presumption favoring judicial review are those establishing “checks” that confer on one branch a limited share of another’s power.75 For instance, Article I grants Congress exclusive and unreviewable authority over impeachments by giving the House of Representatives the “executive” power of prosecuting them and the Senate the “judicial” power of judging them.76 Thus, the Court in Nixon correctly held impeachment to be a political question, as Professor Shane argues.77 But I disagree with Shane that Nixon dictated a similar ruling in Bush,78 because neither 72. Id. at 512. 73. THE FEDERALIST NO. 78, at 524-25 (Alexander Hamilton) (Jacob E. Cooke ed., 1961). The Marshall Court adumbrated a similar approach. See, e.g., Marbury v. Madison, 5 U.S. (1 Cranch) 137, 166-67, 176-79 (1803) (holding that the judiciary alone can independently determine whether the political departments have obeyed the constitutional restrictions on their power, but recognizing exceptions to judicial review such as the President’s conduct of foreign affairs and his nomination of executive and judicial officials). As Marbury illustrates, the presumption can be overcome only by certain constitutional clauses dealing with the government’s structure and operation, where their interpretation by the political branches does not violate individual rights. I have detailed the foregoing analysis in Pushaw, Justiciability, supra note 2, at 497-511. 74. See Pushaw, Justiciability, supra note 2, at 501, 504-11. 75. See id. at 428-30, 505-07. 76. See U.S. CONST. art. I, § 2, cl. 5, and § 3, cls. 6-7. Overwhelming historical evidence demonstrates that the Framers intended for Congress to have sole authority in this area. See Pushaw, Justiciability, supra note 2, at 429-30, 505-07 (citing historical sources). Moreover, a key purpose of impeachment—to enable Congress to remove federal judges— would be compromised if those judges had the final word on the constitutional processes designed to control them. Id. at 505-06 n.556. Finally, the Constitution’s underlying political theory of checks and balances requires that impeachment be committed exclusively to legislative discretion. Id. at 429-30, 505-07. A similar analysis applies to the Constitution’s other classic check, the executive veto, which enables the President to share in (and thus curb) the “legislative power.” See id. at 428-29, 505-06 (arguing that the President’s exercise of his veto power cannot be challenged in court). 77. See Walter Nixon v. United States, 506 U.S. 224 (1993); see also Shane, supra note 5, at 581. 78. Shane, supra note 5, at 581-83. 616 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:603 Article II nor the Twelfth Amendment contains a pure check. Rather, these two structural constitutional provisions (like many others) are justiciable yet should be interpreted by according the highest respect to the judgments of the government officials primarily entrusted with administering them.79 3. Application of the Proper Standard of Review in Bush v. Gore Article II bestows on state legislatures vast power to direct the appointment of presidential electors.80 The Court appropriately reviewed the Florida Supreme Court’s order because doing so was necessary to ensure that the latter tribunal did not alter its legislature’s statutory requirements for selecting presidential electors.81 Because 79. An example of such a provision is Article IV’s Republican Form of Government Clause. In Baker v. Carr, 369 U.S. 186 (1962), the Court erroneously assumed that all complaints under this Clause were political questions. Id. at 209-10, 217-29. Rather, federal courts presumptively can review such claims, but should yield to Congress and the President when they have determined that a state government is republican, absent an extraordinary situation such as a declaration of permanent martial law. See Luther v. Borden, 48 U.S. (7 How.) 1, 42-45 (1849); see also supra note 17 (noting that the Court adjudicated many Guarantee Clause cases between 1849 and 1912). Applying this analysis to the facts of Baker, the Court should have accepted the judgment of the federal political branches that Tennessee’s apportionment scheme was sufficiently republican, because neither the Guarantee Clause nor any other constitutional provision suggests that representation based on factors other than population (e.g., geography and economic interests) is “unrepublican.” Indeed, the Senate and the Electoral College refute that argument. See supra notes 28-31 and accompanying text. The Baker Court’s relabeling of this “Republican Form of Government” action as an “Equal Protection” one is unconvincing. The Court did not even attempt to explain how Tennessee’s apportionment statute violated the Equal Protection Clause as written and as always understood—that is, as protecting the civil rights of individuals and minorities, not political rights. See Baker, 369 U.S. at 226 (asserting that the Tennessee voters had stated a justiciable Equal Protection claim but setting forth no applicable legal standards). Although the Court later established a “one person, one vote” standard, Justice Harlan demonstrated that this principle had no constitutional foundation, and he lambasted the majority for recklessly interfering in the politics of state apportionment and voterqualification decisions. See supra note 39 and accompanying text. 80. See Bush v. Gore, 531 U.S. 98, 104 (2000) (deeming this power “plenary”). 81. The Chief Justice made precisely this point. Id. at 111-16 (Rehnquist, C.J., concurring). Article II clearly authorizes each state legislature, before the election, to set forth procedures governing the manner of appointing presidential electors. Arguably, Article II implicitly allows a legislature to protect this power by selecting electors according to these preexisting statutory standards and to disregard state judicial decisions that, in its independent judgment, have altered this law. Indeed, Florida’s legislature seemed prepared to do just that, and the Court suggested that such action would be valid. See id. at 104 (“The State, of course, after granting the franchise in the special context of Article II, can take back the power to appoint electors.”). By contrast, Profesor Krent would limit a state legislature to the initial enactment of statutes to govern the appointment of electors. If the state governor, state judges, or Congress attempted to change that law after the election, the United States Supreme Court would be available to review and thwart such attempted amendments. See Krent, supra note 5, at 511. Krent’s approach seems more sensible, for two reasons. First, the Florida Legislature itself delegated power to the executive branch to administer its election laws and to the judi- 2001] REPLY TO KRENT AND SHANE 617 the Florida Legislature had delegated to its judiciary the power to adjudicate any disputes arising under its election laws, however, the Florida Supreme Court’s interpretation of those laws was entitled to the customary extraordinary deference that the U.S. Supreme Court gives such judgments.82 As Professor Krent has shown, the Court has become exceedingly reluctant to second-guess state judges’ constructions of state law and invariably rejects claims that they “changed” that law,83 except in very rare cases where the state court’s interpretation was demonstrably arbitrary, biased, or totally unforeseeable.84 I share Krent’s opinion that this precedent required affirmance of the Florida Supreme Court, whose reading of its state’s election statutes was at least plausible—even if (as I believe) the concurring Justices’ interpretation was far more persuasive.85 ciary to decide cases arising under them. Having “tied itself to the mast,” the legislature cannot untie itself when its courts issue a displeasing judgment. Second, a basic principle of the constitutional rule of law is that the legislature which makes a law cannot interpret and apply it in specific cases. See Shane, supra note 5, at 536. Furthermore, it seems anachronistic for the Court to declare that Article II grants state legislatures “plenary power” over the selection of electors—including the right to name them directly, without any popular vote. See Bush, 531 U.S. at 102-03 (citing McPherson v. Blacker, 146 U.S. 1, 35 (1892)). Baker and its progeny created a meta-constitutional principle (located principally in the Fourteenth Amendment and Article I, Section 2 but also in the Fifteenth, Seventeenth, and Nineteenth Amendments) that representation must reflect the equally weighted vote of each citizen. The Warren Court applied this principle to curtail the state legislatures’ power over apportionment that the Constitution seemed to make as absolute as their authority over presidential elections. See Pushaw, Conservative Mirror, supra note 3. Hence, although I do not share Professor Shane’s enthusiasm for the apportionment cases, I agree that these decisions, as well as the whole post-Civil War “democratic trajectory” of the Constitution’s amendments, make the Court’s embrace of McPherson very odd. See Shane, supra note 5, at 548. Of course, this debate is really academic, because neither the Florida Legislature nor any other state legislature has unilaterally named a slate of electors for over a century. 82. See infra notes 83-85 and accompanying text. 83. Krent, supra note 5, at 511-26. The reasons are both theoretical (e.g., concerns for federalism) and practical (e.g., the difficulty of determining which judicial applications of the law to new circumstances unreasonably “change” it). Id. at 524-26. 84. For example, the Court reversed a South Carolina tribunal’s creative interpretation of its state’s trespass law that was designed to deprive blacks of their federal Due Process rights. See Bouie v. City of Columbia, 378 U.S. 347, 348-349 (1964); see also Krent, supra note 5, at 512-21 (discussing Bouie and the few other cases overruling state judicial constructions of state law, and establishing that the Court has ceased questioning any state decisions that may affect federal Contracts Clause rights). 85. See Krent, supra note 5, at 497 (“The Florida Supreme Court’s construction of Florida law, while in no way dictated by precedent or the plain language of the statutory scheme, was at a minimum, plausible. The U.S. Supreme Court has consistently failed to disturb far more questionable state court decisions . . . . ”). Indeed, the very fact that the four dissenting Justices would have deferred to the Florida court’s interpretation indicates that it was at least reasonable. See id. at 501-02. Nonetheless, the concurring Justices concluded that Bush was an exceptional case in which a state court had actually changed its state’s laws so as to frustrate federal interests. See Bush, 531 U.S. at 111-22 (Rehnquist, C.J., concurring). Specifically, they argued 618 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:603 The second crucial provision, the Twelfth Amendment, gives Congress broad discretion in counting—and hence determining the validity of—electoral votes.86 The Court should always affirm Congress’s decisions, absent some plain and egregious violation of the Twelfth Amendment87 or some other constitutional provision.88 Unfortunately, the Court did not wait for Florida to complete the process of appointing electors or for Congress to review the validity of Florida’s electoral votes. Thus, the critical justiciability issue in Bush was ripeness, not the political question doctrine. Here I concur with that the Florida Supreme Court had violated Article II by overturning the Secretary of State’s reasonable exercise of her delegated statutory discretion to (1) enforce the legislature’s deadline for certification and ignore untimely ballots, and (2) construe “legal vote” as not requiring the counting of improperly marked ballots. Id. at 118-20 (Rehnquist, C.J., concurring). However, the Chief Justice acknowledged that the provisions of Florida’s election code “may well admit of more than one interpretation.” Id. at 114 (Rehnquist, C.J., concurring). Yet if other constructions (presumably including that of the Florida court) were possible, then under longstanding precedent the U.S. Supreme Court should have deferred to them. See id. at 135-44 (Ginsburg, J., dissenting); Krent, supra note 5, at 530-33. Like Professor Krent, I find it ironic that the concurring Justices charged a state court with “altering” rather than “interpreting” existing law in a case in which they were making up new Equal Protection law. See Krent, supra note 5, at 497. Significantly, three distinguished conservative legal thinkers have concluded that, although the Court’s Equal Protection analysis was dubious, it nonetheless properly intervened to thwart the Florida Supreme Court’s nakedly political decisions and thus to preserve the state legislature’s Article II power to direct the selection of presidential electors. See Robert H. Bork, Sanctimony Serving Politics: The Florida Fiasco, 19 NEW CRITERION 4, 8-11 (2001); Richard A. Epstein, “In such Manner as the Legislature Thereof May Direct”: The Outcome in Bush v. Gore Defended, 68 U. CHI. L. REV. 613 (2001); Richard A. Posner, Florida 2000: A Legal and Statistical Analysis of the Election Deadlock and the Ensuing Litigation, 2000 SUP. CT. REV. 1. Professor McConnell echoed these arguments but also asserted that the majority had correctly applied Equal Protection law, which prohibited the Florida Supreme Court from authorizing wholly arbitrary vote-counting standards developed for partisan political advantage. See Michael W. McConnell, Two-and-a-Half Cheers for Bush v. Gore, 68 U. CHI. L. REV. 657 (2001). 86. See supra notes 61-63. 87. For instance, assume the following four facts. First, Congress effectuates its Twelfth Amendment power by enacting a statute providing that, absent evidence of criminal or fraudulent misconduct, it will count the electoral votes certified by each state. Second, the Republican Party controls Congress. Third, the Republican candidate for President would have won a very close election but for a third-party candidate who beat him in Utah (with the Democratic candidate finishing a distant third there). Fourth, in counting the electoral votes, Congress determines that the votes for third-party candidates are not valid and thus awards Utah’s electoral votes to the Republican candidate. In that situation, the Supreme Court could reverse Congress’s decision. Admittedly, this scenario is farfetched, but that is exactly the point: In all other circumstances the Court must defer to Congress’s judgments about validity where the question is open to differing views (e.g., whether certain electoral votes were procured fraudulently, whether they were received by the deadline, etc.). 88. An example of such a palpable and extreme violation would be Congress’s refusal to count electoral votes because they were cast by women or Hispanics. 2001] REPLY TO KRENT AND SHANE 619 Professor Shane and Justice Souter that the Court’s intervention was premature.89 III. FOURTEENTH AMENDMENT ISSUES A. Equal Protection Seven Justices held that the Florida Supreme Court had violated the Equal Protection Clause by failing to set forth specific standards to ensure uniform application of the general statutory “intent of the voter” touchstone.90 Professor Shane correctly deplores “the shallowness of the majority’s analysis.”91 The Equal Protection Clause does not require states to be rigorously consistent in administering elections and counting votes, and it certainly does not prohibit the “intent of the voter” standard for determining ballot validity that many states use.92 Indeed, our constitutional system of federal elections virtually guarantees some unequal treatment of voters in order to promote the competing goals of federalism and decentralization. As explained above, Article II authorizes each state legislature to regulate the procedures for selecting presidential electors.93 The states, in turn, have 89. See Bush, 531 U.S. at 129 (Souter, J., dissenting) (speculating that, if the Court had not intervened, the state and Congress may have worked out the problem and “there would ultimately have been no issue requiring our review”). Professor Shane also argues that the Court intervened too soon and adds that Congress probably would have done no worse than the Court in sorting out this mess—and would have been held directly accountable for any mistakes. Shane, supra note 5, at 537. In the Bush majority’s defense, however, the Warren Court made ripeness a discretionary determination. It directed judges to evaluate “the fitness of the issues for judicial decision and the hardship to the parties of withholding court consideration.” Abbott Labs. v. Gardner, 387 U.S. 136, 149 (1967). The Bush majority might reasonably have concluded that (1) the Article II and Equal Protection questions presented needed no further development, and (2) postponement may have caused hardship by permitting the recount to continue and the victor to be announced, only to have the Court invalidate the outcome—or perhaps not to have had enough time to issue a judgment, given the tight deadlines for selecting presidential electors. See Pushaw, Conservative Mirror, supra note 3. 90. See Bush, 531 U.S. at 103-08 (per curiam opinion joined by Rehnquist, C.J., and O’Connor, Scalia, Kennedy, and Thomas, JJ.); id. at 134 (Souter, J., dissenting); id. at 14546 (Breyer, J., dissenting). 91. Shane, supra note 5, at 584. 92. See Bush, 531 U.S. at 124-26 (Stevens, J., dissenting); see also id. (maintaining that the Florida Supreme Court, by ordering a manual recount pursuant to an “intent of the voter” standard subject to review by a judge to reconcile any discrepancies, did not run afoul of the Equal Protection Clause). 93. See supra notes 80-81 and accompanying text. Professor Shane argues that Section 2 of the Fourteenth Amendment gave individual citizens the right to vote in presidential elections, thereby limiting Article II’s absolute grant to state legislatures of the power to select electors however they pleased (including by direct appointment). See Shane, supra note 5 at 572-73. Although Shane’s claim is supportable, I think that if the Amendment’s drafters had intended such a radical change, they would have done so far more directly. Moreover, such textual vagueness is important because the historical evidence is conflicting. See, e.g., Rey- 620 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:603 allowed each county considerable autonomy over the details, such as ballot design and voting machinery.94 As Professor Shane points out, such disparities will likely have a greater impact on election results than allowing each county to use different standards in determining “intent of voter.”95 Thus, applying searching Equal Protection scrutiny to state election laws would cause massive instability.96 Nonetheless, Bush’s “shallow” reasoning is an improvement over Baker’s, which was nonexistent. The Warren Court deliberately declined to set forth any substantive legal standards under the Equal Protection Clause, but rather simply held the voters’ unprecedented claim to be justiciable.97 Only later did the Court fabricate the “one person, one vote” maxim, itself a simplistic slogan that ignored the many nonpopulation interests involved in legislative representation.98 These cases effectively put federal judges in charge of restructuring state legislatures, thereby dealing a devastating blow to federalism.99 Indeed, the very fact that Baker and its progeny had already been decided makes Bush more acceptable.100 Generally speaking, the apportionment cases created a “precedent” for the Justices to interpret the Equal Protection Clause very creatively to solve perceived nationwide electoral crises, even if doing so interferes with state autonnolds v. Sims, 377 U.S. 533, 589-614 (1964) (Harlan, J., dissenting) (contending that the Fourteenth Amendment did not affect the states’ preexisting power over their citizens’ voting qualifications, and noting that Section 2 expressly recognized each state’s authority to “deny” or “abridge” the right to vote and attached a penalty for doing so—suffering a proportional reduction in the basis of its representation). 94. See Bush, 531 U.S. at 126 (Stevens, J., dissenting). 95. Shane, supra note 5, at 572-74. 96. The Court, however, was careful not to question statutory discrepancies in voting systems among and within the states, and indeed it acknowledged that local entities “may develop different systems for implementing elections.” Bush, 531 U.S. at 109. Rather, the majority stressed that “a state court with the power to assure uniformity has ordered a statewide recount with minimal procedural safeguards.” Id. The Court underscored this point by limiting its holding “to the present circumstances, for the problem of equal protection in election processes generally presents many complexities.” Id. Presumably, “the present circumstances” refers to court-ordered recounts in contested elections, not to all state statutory disparities in voting matters. If by “the present circumstances” the majority meant the facts of Bush only, then the Court engaged in fiat rather than adjudication, which requires the consistent application of legal principles in successive cases. See Pushaw, Conservative Mirror, supra note 3; see also Shane, supra note 5, at 580 (deeming “suspect” the “self-declaration that the Justices may be writing a ticket for one ride only”). By contrast, even though the Warren Court made up the “one person, one vote” rule, it intended to (and did) apply it in all later decisions. See supra notes 32-38 and accompanying text; see also Shane, supra note 5, at 583 (contending that judicial activism can be constructive but only if the new constitutional values can be articulated in “adequately neutral principles to be persuasively applicable to foreseeable future cases”). 97. See supra notes 26-27 and accompanying text. 98. See supra notes 32-39 and accompanying text. 99. See id. 100. See Pushaw, Conservative Mirror, supra note 3. 2001] REPLY TO KRENT AND SHANE 621 omy in administering elections.101 More specifically, the Rehnquist Court, unlike the Warren Court, actually had some precedent that at least colorably justified its ruling. First, the majority found that, because the Florida Legislature had exercised its electoral appointment power by authorizing its citizens to vote, Reynolds gave them a fundamental right which could not be infringed by arbitrary treatment that valued one person’s vote over another’s.102 Second, Gray barred states from disparately treating voters in different counties.103 The majority held that, contrary to these cases, the Florida Supreme Court had authorized “standardless manual recounts,”104 with the result that two voter ballots marked exactly the same way would not necessarily be counted identically.105 In fact, “the standards for accepting or rejecting contested ballots might vary not only from county to county but indeed within a single county from one recount team to another.”106 It is this last discrepancy, I suspect, that the Court found particularly galling. Canvassing boards and counting teams in heavily Democratic counties switched from their established (and thus politically neutral) written guidelines for determining the legal validity of ballots to progressively more lenient standards (for example, from completely detached to partially detached to merely indented chads).107 Millions of Americans watching on television, and virtually all Republicans, concluded that these constant changes were made to help Al Gore get more votes.108 Moreover, the majority apparently found it more than a mere coincidence that the solidly Democratic Florida Supreme Court happened to decide every major issue in favor of the Democratic candidate.109 These Justices would find disingenuous Professor Shane’s comment that “it is hard to predict who [would 101. See supra notes 20-38 and accompanying text. 102. Bush, 531 U.S. at 104-05 (citing Reynolds v. Sims, 377 U.S. 533, 555 (1964)); see supra notes 32-33 and accompanying text. 103. Bush, 531 U.S. at 107 (citing Gray v. Sanders, 372 U.S. 368 (1963)); see supra notes 34-35, 57 and accompanying text. 104. Bush, 531 U.S. at 103. 105. Id. at 104-09. 106. Id. at 106. For instance, in Miami-Dade County, three canvassing board members used three separate standards in defining a legal vote. Id. Similarly, Palm Beach County “changed its evaluative standards during the counting process”: It “began . . . with a 1990 guideline which precluded counting completely attached chads, switched to a rule that considered a vote to be legal if any light could be seen through a chad, changed back to the 1990 rule, and then abandoned any pretense of a per se rule, only to have a court order that the county consider dimpled chads legal.” Id. at 106-07. 107. See supra note 106. 108. The Justices simply noted the changes and did not mention the partisanship that motivated them, but it is not too hard to connect the dots. 109. Again, the majority did not make any direct accusations, but Justice Stevens found them implicit. See Bush, 531 U.S. at 128 (Stevens, J., dissenting) (criticizing the Court for an “unstated lack of confidence in the impartiality and capacity of the state judges who would make the critical decisions if the vote count were to proceed”). 622 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:603 have been] hurt by” the Florida Supreme Court’s order allowing different standards for evaluating contested ballots, limiting recounts to “undervotes” in some counties but counting all ballots in others, and permitting partial recounts from some counties but complete totals from others.110 In short, this evidence suggests that the Bush majority genuinely believed they were upholding the integrity of the votes cast and thus promoting democratic ideals. Consequently, I reject Professor Shane’s charge that the Court sought to thwart the right to suffrage and thereby “betrayed” democracy.111 I also have no basis for doubting that the Warren Court Justices sincerely felt that they were furthering democracy by crafting a “one person, one vote” standard.112 The crucial point is that Baker and its progeny repose trust in a majority of Justices to subjectively determine what is “democratic.” Under the “law” of Baker, that gut call will sometimes produce a liberal political outcome, and sometimes a conservative one.113 110. Shane, supra note 5, at 552. On the contrary, it is pretty easy to predict that Bush would have been harmed. First, permitting different standards to determine a legal vote appeared to help Gore, especially in light of the seemingly partisan changing of standards. See supra notes 104-06 and accompanying text. Second, the Florida Supreme Court’s order selectively confining recounts in heavily Democratic Miami-Dade County to 9000 undervotes, Bush, 531 U.S. at 102-03, would likely have injured Bush because the majority of those votes were presumably Democratic, and the canvassing board members there were using multiple standards. See supra notes 106-08 and accompanying text. Third, moving beyond the realm of educated guesses, the Florida court’s decision to accept partial recounts from Miami-Dade actually yielded 168 Gore votes. Bush, 531 U.S. at 102-03. So did its addition of 215 Gore votes from Palm Beach County, even though they were submitted after the Florida Supreme Court’s own November 26 deadline. Id. 111. See Shane, supra note 5, at 583. Of course, I do not deny that the perceptions of some of the Republican Justices may have been influenced by their own political biases. But neither Professor Shane nor anyone else has presented any evidence that those Justices deliberately decided Bush according to their political predilections in order to deprive people of their right to vote and undermine democracy. 112. See supra notes 32-38 and accompanying text. 113. Thus, commentators like Professor Shane cannot consistently laud Baker and its progeny while condemning Bush. Conversely, conservatives cannot fairly criticize Baker as a symbol of Warren Court political excess while cheering Bush as “valiant” and “legitimate in law.” See Bork, supra note 85, at 8-11. Rather, legal consistency requires either accepting both decisions or (as I do) rejecting both. Although some regard consistency as “the hobgoblin of little minds,” I think it is essential to the legitimacy of adjudication and hence the rule of law. I cling to the quaint view that the Court should apply principles of “law” rooted in the Constitution’s clauses as illuminated by (1) the history of their drafting, ratification, and early implementation, (2) underlying structural principles such as separation of powers and federalism, (3) Federalist political theory, and (4) precedent based on these materials and gradually adapted over the years. See, e.g., Pushaw, Case/Controversy, supra note 2, at 448, 468-517. Accordingly, I reject the modern Court’s assertion of untethered discretion in many critical areas of constitutional law. See, e.g., Robert J. Pushaw, Jr., The Inherent Powers of Federal Courts and the Structural Constitution, 86 IOWA L. REV. 735 (2001) (setting forth historically based legal standards to govern federal judges’ exercise of “inherent authority,” which is currently invoked to rationalize virtually any action related to managing litigation or imposing sanctions); Grant S. Nelson & Robert J. Pushaw, Jr., Rethinking the Commerce Clause: Apply- 2001] REPLY TO KRENT AND SHANE 623 B. Due Process and Related Concerns Professor Shane develops an elaborate, and defensible, argument that the Due Process Clause required hand recounts in the four challenged counties.114 I am not persuaded, however, that this Clause mandates such a specific remedy. Indeed, to the extent that the Florida Supreme Court ordered manual recounts but did not ensure fair and consistent standards, that failure may have violated the Due Process Clause.115 More specifically, Shane characterizes Bush as a type of mass administrative adjudication of individual claims, like Social Security.116 Although there are certain similarities, I do not think that Bush was primarily about vindicating individual rights. Rather, the case really concerned the interest of all Americans in ensuring the integrity of the electoral process in selecting their only single national leader. CONCLUSION The panelists have made several persuasive arguments. For instance, Professor Krent correctly contends that precedent authorized the Court to exercise jurisdiction but obliged it to defer to the Florida Supreme Court’s decision. Likewise, Professor Shane properly exposes the shortcomings of Bush v. Gore’s Equal Protection analysis. Nonetheless, I cannot agree with either of them that the Court’s decision in Bush is wholly unprecedented. Rather, Baker is directly on point, even if none of the Justices mentioned it. ing First Principles to Uphold Federal Commercial Regulations but Preserve State Control Over Social Issues, 85 IOWA L. REV. 1 (1999) (developing a legal test for the Commerce Clause that is faithful to history yet workable in resolving modern disputes). 114. Shane, supra note 5, at 553-68. 115. Bush v. Gore, 531 U.S. 98, 103, 110 (2000) (alluding to this alternative basis for its holding of unconstitutionality). 116. Shane, supra note 5, at 562-63. THE REGULATORY ROLE OF STATE CONSTITUTIONAL STRUCTURAL CONSTRAINTS IN PRESIDENTIAL ELECTIONS JAMES A. GARDNER* I. THE BUSH DECISIONS .......................................................................................... A. Bush I ............................................................................................................ B. Bush II .......................................................................................................... II. STATE POWER IN A FEDERAL SYSTEM ................................................................. III. STATE CONSTITUTIONS AND THE SELECTION OF PRESIDENTIAL ELECTORS ....... IV. DECENTRALIZATION OF STATE ELECTORAL ADMINISTRATION............................ CONCLUSION ........................................................................................................ 627 627 629 630 640 651 658 In our federal system of government, state officials are normally understood to draw their legitimate authority from affirmative grants of power under their state’s constitution. Just as it grants official power, however, a state constitution is normally understood to limit and condition it as well in ways that conclusively bind state actors. In Bush v. Palm Beach County Canvassing Board (Bush I)1 and Bush v. Gore (Bush II),2 the U.S. Supreme Court unjustifiably departed from this model, a model drawn from salutary principles of federalism that play a significant role in the protection of popular liberty. State constitutions serve three main functions: they authorize and empower the state government to achieve the public good; they grant the state power sufficient to resist abuses of national authority by the national government; and they control the state power thus granted by establishing institutions of governmental self-restraint so that the state government does not become an undue threat to its own people. Any state constitution thus strikes a balance between empowerment and restraint of state government. The way any state polity chooses to strike this balance will depend upon its trust in government generally and its relative trust of state and national power in particular. The U.S. Constitution, however, has virtually nothing to say about how a state may strike this balance. Although it protects a state’s citizens from their own state government by establishing a minimal level of individual rights, the U.S. Constitution leaves to state polities, through their state constitutions, any decisions about additional structural limitations on state power. * Visiting Professor of Law, State University of New York, University at Buffalo School of Law; Professor of Law, Western New England College School of Law. I thank Bob Williams for commenting on an earlier draft. 1. 531 U.S. 70 (2000) (per curiam). 2. 531 U.S. 98 (2000) (per curiam). 625 626 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:625 In the Bush cases, the Supreme Court interpreted Article II, Section 1 and the Equal Protection Clause of the U.S. Constitution to create a different system in presidential elections. According to the Court in Bush I, a state legislature engaged in the process of selecting presidential electors, or determining how they will be selected, is free from any constraints that the state constitution might otherwise impose on the exercise of legislative power. The Court’s ruling in Bush II further invaded state autonomy over the internal structure of state government by subjecting Florida’s decision to decentralize the administration of statewide elections to what amounts to a federal nondelegation doctrine for states. These rulings unjustifiably disable state constitutions from serving in presidential elections the significant, liberty-protective regulatory function they serve on every other occasion in which state power is exercised: restraining the organs of state government from abusing important state powers. Part I of this paper briefly discusses the Bush cases. Part II then lays out the role that state constitutions normally play in implementing the American structure of federalism, a system the principal goal of which is the adequate protection of liberty. Part III turns to whether state constitutions should be understood to play a diminished role in the protection of state liberty during presidential elections. Although Article II authorizes state legislatures to appoint presidential electors,3 I argue, it by no means follows that in so doing Article II unmoors state legislatures from the state constitutional restraints upon legislative power that attend its use in every other circumstance. In fact, it is far from unusual for the national Constitution to grant authority which state constitutions then limit or revoke, and there is no reason to suspect a different arrangement here. Moreover, if state constitutions do not permit state legislatures to exercise legislative power in a way that comports with the requirements of Article II, then Article II provides an obvious remedy: exclusion of the state’s electors from the presidential selection process. I argue in Part IV that the Court’s ruling in Bush II, although couched in the language of equal protection, responds to similarly misplaced concerns about autonomous state decisions regarding the internal structure of state government. The Court’s equal protection analysis is entirely unsatisfactory. Its main concern is apparently with what would ordinarily be characterized as a due process objection to the vagueness of the “intent of the voter” standard. Vagueness, however, is a concern under the Due Process Clause only with respect to laws addressed to individuals, not government officials. When laws are too vague for government officials reliably to follow, the problem is better cast in terms of nondelegation, a separation of 3. U.S. CONST. art. II, § 1. 2001] REGULATORY ROLE 627 powers doctrine. Yet the U.S. Constitution has nothing to say about the way a state chooses to divide power on the state level, including its choice not to observe a strict nondelegation doctrine. Moreover, a state’s decision to decentralize the administration of elections is also a legitimate internal structural decision. To the extent it is based on a distrust of state as compared to local power, such a vertical dispersion of state power serves as a mechanism by which the state polity may protect itself from centralized state abuses of power during statewide elections. The Court’s misguided equal protection analysis unjustifiably invades this essential area of state autonomy. In reaching these decisions, the Court has acted in a way that strips from the presidential election process two important, liberty-protective safeguards. I. THE BUSH DECISIONS A. Bush I Article II, Section 1 of the U.S. Constitution provides that presidential electors from each state shall be selected “in such Manner as the Legislature thereof may direct.” The Florida Legislature, like every other state legislature in the union, has long redelegated that responsibility to the people of the state, who elect slates of electors pledged to particular candidates.4 The first count of Florida’s popular vote in the November 7, 2000, presidential election showed an exceedingly close result—George Bush led Al Gore by fewer than 1,800 votes out of nearly six million cast.5 Under Florida law, the closeness of the result triggered an automatic machine recount,6 further reducing Bush’s lead. At the request of Florida Democratic Party officials, manual recounts of one percent of the ballots were conducted in several counties.7 These recounts further narrowed Bush’s lead, causing county officials to determine that “an error in the vote tabulation which could affect the outcome of the election”8 had occurred, thus warranting full, countywide, manual recounts.9 At this point, questions arose concerning the statutory deadline by which counties conducting manual recounts had to certify their results to the Florida Secretary of State for final statewide tabulation.10 The Secretary of State, relying on an advisory opinion issued by the Florida Division of Elections, announced that she would not accept 4. FLA. STAT. § 103.011 (2000). 5. Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d 1220, 1225 (Fla. 2000), vacated sub nom. Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000). 6. FLA. STAT. § 102.141(4) (2000) (amended 2001). 7. Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d at 1225. 8. FLA. STAT. § 102.166(5)(c) (2000) (amended 2001). 9. Id. § 102.166(5) (amended 2001). 10. See Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d at 1225-26. 628 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:625 vote totals submitted later than seven days after the election,11 as provided in one section of Florida’s election code.12 Some county boards of elections, concerned that the seven-day statutory deadline would not provide them with enough time to carry out recounts that they were obliged to conduct under other sections of the election code, filed suit seeking a reconciliation of their duties.13 Several related cases were then consolidated before the Florida Supreme Court, which ruled on November 21, 2000.14 In its opinion, the Florida Supreme Court interpreted the contested provisions of Florida’s election code.15 It held first that the statutory phrase “error in the vote tabulation” includes “a discrepancy between the number of votes determined by a voter tabulation system and the number of voters determined by a manual count of a sampling of precincts,”16 thus affirming the counties’ decisions to conduct full manual recounts.17 It then reconciled several seemingly inconsistent provisions of the Florida election code concerning the timing of manual recounts,18 holding that the seven-day deadline was not a final deadline, and that the Secretary of State must accept manual recount totals unless they are so late as to preclude unsuccessful candidates from filing election contests.19 Finally, given the lateness of the hour and the tangled course of events, the court invoked its equitable powers to fashion a remedy in which the Secretary of State was directed to accept manual recount totals from counties until November 26, 2000.20 In reaching this result, the Florida Supreme Court looked frequently to the Florida Constitution for guidance. 21 The court’s analysis was guided, it held, by a paramount obligation to effectuate the state constitutional right to vote of the voters, whom it regarded as the real parties in interest.22 In reviewing the applicable law, the court listed first two sections of the Florida Constitution dealing with elections, including one that authorizes the state legislature to regulate elections by law.23 Finally, in construing conflicting electoral statutes and settling on a remedy, the court noted that the Florida 11. Id. at 1226. 12. FLA. STAT. § 102.111(1) (2000) (amended 2001). 13. See Palm Beach County Canvassing Bd. v. Harris, 772 So. 2d. at 1226. 14. Id. at 1220, 1226. 15. Id. at 1220. 16. Id. at 1229. 17. See id. 18. See id. at 1231-39. 19. Id. at 1239. 20. Id. at 1240. 21. See id. 22. Id. at 1227-28. 23. Id. at 1230 (citing FLA. CONST. art. VI, § 1 (“Registration and election shall . . . be regulated by law . . . .”)). 2001] REGULATORY ROLE 629 Constitution prohibits the legislature from regulating elections in ways that impose “‘unreasonable or unnecessary’ restraints on the right of suffrage.”24 Bush appealed this ruling to the U.S. Supreme Court, which, in Bush I, vacated the Florida Supreme Court’s ruling and remanded for clarification as to the actual basis for the lower court decision.25 The Court was unsure, it said, “as to the extent to which the Florida Supreme Court saw the Florida Constitution as circumscribing the legislature’s authority under Art. II, § 1, cl. 2.”26 Although the Court technically went no further than identifying this ambiguity, it suggested strongly throughout the opinion that the Florida Supreme Court had committed reversible error to the extent it had viewed the Florida Constitution as constraining in any way the authority of the Florida Legislature under Article II, Section 1 of the U.S. Constitution.27 The Court quoted dicta from its 1892 decision in McPherson v. Blacker,28 suggesting that Article II’s specific grant of authority to the “state legislature,” as opposed to the state itself, not only “operat[es] as a limitation upon the State in respect of any attempt to circumscribe the legislative power” but also “cannot be held to operate as a limitation on that power itself.”29 Relying on this dictum, the Court went on to observe that the Florida Supreme Court had said things in its opinion “that may be read to indicate that it construed the Florida Election Code without regard to the extent to which the Florida Constitution could, consistent with Art. II, § 1, cl. 2, ‘circumscribe the legislative power.’”30 The implication of this sentence is clearly that the Florida Constitution may not be so read. B. Bush II Whereas Bush I arose out of the “protest” phase of Florida’s 2000 presidential election, Bush II arose out of the subsequent “contest” phase. Proceeding subject to the Florida Supreme Court’s decision in Palm Beach County Canvassing Board v. Harris, the Florida Election Canvassing Commission, after receiving various county manual recount totals, certified George Bush the winner of Florida’s electoral votes by a margin of 537 votes.31 Gore then filed a statutory contest proceeding, arguing that votes had been illegally counted in numbers 24. Id. at 1236-37 (quoting Treiman v. Malmquist, 342 So. 2d 972, 975 (Fla. 1977)). 25. Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70, 78 (2000) (per curiam). 26. Id. 27. See id. 28. 146 U.S. 1 (1892). 29. 531 U.S. at 76 (quoting McPherson, 146 U.S. at 25). 30. Id. at 77. 31. Gore v. Harris, 772 So. 2d 1243, 1247 (Fla. 2000) (per curiam), rev’d sub nom. Bush v. Gore, 531 U.S. 98 (2000). 630 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:625 sufficient to change the outcome of the election.32 In conducting an appellate review of various aspects of the counting of presidential votes, the Florida Supreme Court interpreted the state election code to require that certain disputed ballots be counted in Gore’s total.33 In addition, the court ruled that approximately 9,000 additional ballots (the so-called “undercounts”)34 in Miami-Dade County had to be recounted manually and the totals included in the final count.35 Finally, the court held that similarly undercounted ballots throughout the state might also need to be manually counted and remanded the case for such a determination.36 In conducting any recounts, the court held, canvassers should be guided by the statutory standard of attempting to effectuate the “clear . . . intent of the voter.”37 On appeal in Bush II, the U.S. Supreme Court reversed.38 After noting that the U.S. Constitution grants no one the right to vote for presidential electors,39 the Court went on to observe that the Florida Legislature’s decision to permit citizens to vote for electors gave rise to an obligation under the Equal Protection Clause of the U.S. Constitution to avoid valuing any individual’s vote arbitrarily over the vote of another.40 The Florida Supreme Court’s recount decision, the Court held, did not satisfy this obligation because the standard it directed to be applied—the “intent of the voter” standard—was too unspecific “to ensure its equal application.”41 As a result, the Court said, “the standards for accepting or rejecting contested ballots might vary not only from county to county but indeed within a single county from one recount team to another.”42 This process was “inconsistent with the minimum procedures necessary to protect the fundamental right of each voter” in the recount process.43 II. STATE POWER IN A FEDERAL SYSTEM To get our bearings, it will be useful to review some basic principles of federalism. Our system of government is founded on two fundamental premises that are in considerable tension with one another: first, government is necessary; second, it is dangerous. That 32. Id. (relying on FLA. STAT. § 102.168(3)(c) (2000)). 33. Id. at 1248. 34. Undercounted ballots are those on which a voter attempted to punch a vote for President but did so incompletely or in a way that caused the tabulating machine to record the ballot as registering no vote for President. 35. Gore v. Harris, 772 So. 2d at 1262. 36. Id. 37. Id.; see FLA. STAT. § 101.5614(5) (2000) (amended 2001). 38. Bush v. Gore, 531 U.S. 98, 110 (2000) (per curiam). 39. Id. at 104. 40. Id. at 104-05. 41. Id. at 106. 42. Id. 43. Id. at 109. 2001] REGULATORY ROLE 631 we have a system of government at all shows, of course, that Americans have decided that the potential benefits of government outweigh the potential risks. If citizens were able to achieve all their goals without the application of organized governmental power, presumably they would attempt to do so without creating and empowering a government, with its attendant risk of official tyranny. The fact that the American polity has seen fit to create not one but many governments, much less to endow them with the enormous powers they currently possess, suggests that organized, collective social power wielded by a governmental apparatus is essential to the achievement of collective popular goals, whatever they may be. American governments, it thus seems safe to say, require significant affirmative powers if they are to achieve the apparently ambitious goals that the people have established for them, and in so doing to effectuate the public good as citizens conceive it. That kind of power, however, is dangerous, and one of the chief sources of danger is the possibility that it may become excessively concentrated and centralized. As Madison famously observed, “[t]he accumulation of all powers . . . in the same hands . . . may justly be pronounced the very definition of tyranny.”44 Federalism, of course, is a method of structuring government power that recognizes the necessity of government while attempting to reduce the risk of official tyranny. Federalism protects liberty by giving each level of government, state and federal, substantial powers sufficient to allow each to monitor and check the abuses of the other. As with the horizontal separation of powers, which divides governmental power into legislative, executive, and judicial branches, each piece of government in this fragmented system is given the power and incentive to struggle against the others: “[a]mbition,” as Madison put it, “must be made to counteract ambition.”45 The result is a compound federal republic in which power is deeply fragmented, reducing as far as possible by structural means the likelihood that a tyrannical measure of power can be accumulated in a single set of hands: In a single republic, all the power surrendered by the people, is submitted to the administration of a single government; and usurpations are guarded against by a division of the government into distinct and separate departments. In the compound republic of America, the power surrendered by the people, is first divided between two distinct governments, and then the portion allotted to each, subdivided among distinct and separate departments. Hence a double security arises to the rights of the people. The different 44. THE FEDERALIST NO. 47, at 324 (James Madison) (Jacob E. Cooke ed., 1961). 45. THE FEDERALIST NO. 51, at 349 (James Madison) (Jacob E. Cooke ed., 1961). 632 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:625 governments will controul [sic] each other; at the same time that each will be controuled [sic] by itself.46 The multiplicity of power centers in the American scheme can create the impression that the system is chaotic—a pure, Hobbesian war of all against all without any purpose other than the accumulation of power. This is not the case, or at least need not be the case. What unifies the dispersion of governmental power, in the Framers’ view, is the people, for the entire system is designed to assure as far as possible that their wishes will be done and their liberties left intact. “The Federal and State Governments,” Madison observes, “are in fact but different agents and trustees of the people, instituted with different powers and designated for different purposes.”47 Federalism is thus more than a passive institutionalization of political conflict. It is a dynamic system that is designed to be manipulated by the people to produce results they desire. Hamilton put this point clearly: [I]n a confederacy the people, without exaggeration, may be said to be entirely the masters of their own fate. Power being almost always the rival of power; the General Government will at all times stand ready to check the usurpations of the state governments; and these will have the same disposition towards the General Government. The people, by throwing themselves into either scale, will infallibly make it preponderate. If their rights are invaded by either, they can make use of the other, as the instrument of redress.48 Thus, the Framers expressly contemplated that popular allegiances to any government would not be fixed organically but would ebb and flow according to their instrumental value to the populace at any given time. Each level of government would be given sufficient power to check any tyrannical tendencies of the other, but when and whether to activate this checking power would be up to the people. The “double security” of which Madison spoke,49 then, does not arise so much from some complicated scheme of complementary powers, as is so often supposed, but from a conceptually much simpler arrangement in which the state and national governments independently police much of the same turf. To protect that turf, the people make use 46. Id. at 350-51. 47. THE FEDERALIST NO. 46, at 315 (James Madison) (Jacob E. Cooke ed., 1961). 48. THE FEDERALIST NO. 28, at 179 (Alexander Hamilton) (Jacob E. Cooke ed., 1961). See also THE FEDERALIST NO. 46, supra note 47, at 317, in which Madison, after remarking that Americans place their faith and trust primarily in their state governments, observes: If . . . the people should in future become more partial to the federal than to the State governments, the change can only result, from such manifest and irresistible proofs of a better administration, as will overcome all their antecedent propensities. And in that case, the people ought not surely to be precluded from giving most of their confidence where they may discover it to be most due . . . . 49. See supra text accompanying note 46. 2001] REGULATORY ROLE 633 of whichever level of government is more capable and accommodating at any particular moment. This scheme of federalism clearly contemplates that states will have significant powers. First, even apart from their functions in monitoring the exercise of national power, states will likely need significant powers if only to accomplish directly the public good, however the people of the states may understand it. The powers of the national government are impressive, but they are also limited; many collective social goals are largely beyond the power of Congress to effectuate. As we might expect, contemporary state governments possess significant powers which they can use affirmatively in pursuit of the public good. Indeed, the scope of state power tends to be (although it need not be) broader than the scope of national power, reaching many areas of ordinary life that the national government is usually understood to lack the power to regulate.50 For example, state law overwhelmingly provides the controlling substantive rules in the laws of tort, contract, commercial transactions, crimes, property, wills, and family formation—the areas of law that are most likely to affect individuals directly in the most significant aspects of their daily lives. Second, it is clear, particularly given the immense power now wielded by the national government, that state governments must themselves possess considerable power if federalism is to work as intended. States cannot possibly serve as effective counterweights to national power if they are too weak to erect effective obstacles to abuses of national power. States presently have numerous means at their disposal to check abusive exercises of national power. These range from illegal or extraconstitutional methods such as secession, the threat of force, or peaceful defiance of national law; to quasi-legal strategies such as deliberate failure fully to comply with or to enforce binding federal law; to fully legal means such as the use of political pressure, the exercise of ordinary state power where it is not preempted, refusal of national financial incentives, lawsuits against the national government in federal court, and setting higher standards of state conduct by granting more generous constitutional or statutory rights. All of these means of resistance, however, require the state to possess and exercise considerable governmental power. Because state power in a well-functioning federal system must be significant, it inevitably presents a real and substantial risk of abuse. For this reason, state power must be not only granted for legitimate purposes, but somehow restrained so that it will not be used for illegitimate ones. State power may be so restrained in one of two 50. See, e.g., Walter F. Dodd, The Function of a State Constitution, 30 POL. SCI. Q. 201 (1915). 634 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:625 ways. Federalism itself furnishes the first method. Just as federalism contemplates that state power will be deployed to check abuses of national power, it also contemplates that national power will be deployed to check abuses of state power. Thus, the national government has at its disposal numerous means by which abuses of state power may be curbed. These include such familiar mechanisms as preemption of undesirable state laws; the use of financial incentives to induce states to alter their laws or practices; protection of individual rights through application of the national Bill of Rights and the Fourteenth Amendment; and federal judicial review of the constitutionality of state laws. In addition, the national government may back these practices with the implicit (or sometimes explicit) threat of force. It is at least conceivable that some state polity might think that the methods of restraint available to the national government amount to a set of external safeguards entirely sufficient to prevent a state from using its powers tyrannically against its own citizens, and consequently that no additional, internal measures were needed. In practice, however, no state polity has ever been willing to trust its freedom from state tyranny entirely to the national government. On the contrary, the citizenry of every state has thought it necessary to create internal safeguards, implemented by means of a state constitution, which deploy state power against itself to institutionalize a form of state self-restraint. The most common means of internally restraining state government include partially disabling it by dividing power horizontally, vertically or in both ways; creating a state bill of rights; and establishing state judicial review as an enforcement mechanism. States, then, must possess significant power both to achieve the public good directly and to monitor and check abusive exercises of national power; yet the granting of such power poses risks to liberty that require state power to be hobbled by institutions of internal selfrestraint. This tension imposes upon state constitutions a structural trade-off whose parameters are defined by the effectiveness of state power, on one hand, and the effectiveness of limitations on those powers, on the other. At one end of the spectrum, a state polity might create an extremely potent state government enjoying ample authority to impede national tyranny and facing few internal impediments to the effective exercise of those powers. Such a state government would be capable of protecting the liberty of its people vigorously against invasion by the national government, but only at the cost of leaving its citizens relatively vulnerable to invasions of liberty by the state itself. At the other end of the spectrum, a state polity might create a state government enjoying relatively few powers or facing significant internal self-restraints that impair its ability effectively to 2001] REGULATORY ROLE 635 exercise granted powers. Such a state government, because it would be weak, would pose few direct threats to the liberty of the state’s citizens. It would, however, leave them relatively vulnerable to invasions of their liberty originating at the national level. There is no limit to the number of different ways in which a state polity may choose to strike a balance between state empowerment and restraint, and between state and national power. Under federalism, however, the exact nature of that balance is decidedly for the people of each state to determine. Although any number of factors may influence the way in which a state polity chooses to balance these considerations, it seems certain that one extremely important factor will be the particular patterns of distrust of governmental power that prevail among the people of the state. American constitutional theory rests on the bedrock proposition that no government is entitled to the people’s complete trust and faith.51 Yet it does not follow that all governments are to be distrusted equally. There are numerous reasons why a state polity might be less willing to trust one level of government than the other. Size alone, for example, may affect political dynamics in different ways on the state and national levels. As Madison argued in The Federalist No. 10, state political processes may well be more susceptible to a destructive, factional politics of self-interest than are national political processes simply because states are smaller, and contain fewer people, than the nation.52 State governments, on the other hand, are often said to be more responsive to public opinion than the national government, which might make them more rather than less trustworthy. Another factor that might affect the trust the people of a state feel for the national and state governments concerns the relative competencies of each. Not every government has the power, resources or incentives to do every job equally well. Finally, a state polity might feel different degrees of trust for the state and national governments as the result of its actual historical experience of life under both.53 This is not a factor amenable to structural analysis, but rather a contingent matter that depends entirely upon the particular actions each level of government may have taken in the past, the effects of those actions on the state’s citizens, and the ways in which the citizenry has chosen to interpret those actions. If the people of 51. THE FEDERALIST NO. 51, supra note 45, at 349: If men were angels, no government would be necessary. If angels were to govern men, neither external nor internal controuls [sic] on government would be necessary. In framing a government which is to be administered by men over men, the great difficulty lies in this: You must first enable the government to controul [sic] the governed; and in the next place oblige it to controul [sic] itself. 52. THE FEDERALIST NO. 10, at 63-64 (James Madison) (Jacob E. Cooke ed., 1961). 53. See, e.g., C. VANN WOODWARD, THE BURDEN OF SOUTHERN HISTORY 17-21 (rev. ed. 1960). 636 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:625 the state feel that the national government has, on the whole, treated them better than the state government, they may be more inclined to trust national power than state power, and vice versa. Whatever the choices made by a state polity, it is clear that a critical function of a state constitution is to embody and institutionalize those choices through the establishment, empowerment, and restraint of concrete institutions of state governance. A state constitution, then, is neither a piece of legal decoration nor some kind of optional add-on that state polities may choose to create for purposes that are of no concern to the national polity or its government. On the contrary, state constitutions are critical tools in operationalizing and institutionalizing the principles of federalism. They are an indispensable component of a system that must function effectively at every level if liberty is to be protected adequately at any level. Normally, the U.S. Constitution is understood to have little influence over these state decisions: it treats the protection of a state’s citizens from their own state government largely as a matter of internal state self-governance.54 Thus, how much power state governments are to possess, how that power is to be structured and divided, and what rights citizens are to hold against state governments are all questions for the people of each state to answer for themselves. The national Constitution does, to be sure, contain some mechanisms that protect citizens from state tyranny.55 It establishes, for example, a minimal level of protection for individual rights which states are obliged to respect.56 The Constitution also provides the national government with certain tools for intervening to relieve state citizens from the burdens of abusive state actions, preemption, financial inducements, and judicial review of state laws, for example. But these protections are quite limited. Congress can hardly monitor every law 54. See, e.g., Dreyer v. Illinois, 187 U.S. 71, 83-84 (1902). A local statute investing a collection of persons not of the judicial department, with powers that are judicial and authorizing them to exercise the pardoning power which alone belongs to the Governor of the State, presents no question under the Constitution of the United States. The right to the due process of law prescribed by the Fourteenth Amendment would not be infringed by a local statute of that character. Whether the legislative, executive and judicial powers of a state shall be kept altogether distinct and separate, or whether persons or collections of persons belong to one department may, in respect to some matters, exert powers which, strictly speaking, pertain to another department of government, is for the determination of the State. Id.; see also Mayor of Philadelphia v. Educ. Equality League, 415 U.S. 605, 615 n.13 (1974) (“The Constitution does not impose on the States any particular plan for the distribution of governmental powers.”). 55. This occurs by operation of the Fourteenth Amendment’s “incorporation doctrine.” 56. See, e.g., Duncan v. Louisiana, 391 U.S. 145 (1968); Griswold v. Connecticut, 381 U.S. 479 (1965); Adamson v. California, 332 U.S. 46 (1947); Palko v. Connecticut, 302 U.S. 319 (1937). 2001] REGULATORY ROLE 637 enacted in every state,57 and the protections furnished by the national Bill of Rights, though hardly inconsiderable, have never been thought to represent the last word on the relationship between individuals and the state. If the U.S. Constitution provides some protection against state tyranny by establishing baseline levels of individual rights and by authorizing external intervention by the national government, it provides virtually (and perhaps literally) no such protection when it comes to the question of how state power ought to be structured and deployed—a question long understood in our system of constitutional government to be critical to the protection of liberty.58 Here, state polities are left almost entirely to their own devices. A few provisions of the U.S. Constitution, by referring to a state legislature, state executive or state judiciary, clearly presuppose the existence of a state government divided into three branches.59 Whether they require such a division is altogether a different question, and is in any event beside the point. Even if the U.S. Constitution requires state constitutions to establish three distinct government branches, it says nothing at all about what powers the branches ought to possess and how those powers ought to be distributed among them, leaving it silent on the one issue that actually counts in determining whether a system of constitutional separation of powers will protect liberty adequately and permanently.60 The only provision of the U.S. Constitution that might plausibly be thought to establish specific, liberty-protective requirements relating to the internal structure of state government is the Guarantee Clause, which requires the United States to “guarantee to every State in this Union a Republican Form of Government.”61 The mean57. The Court’s Dormant Commerce Clause jurisprudence recognizes this reality in the economic realm. 58. The original U.S. Constitution, after all, did not contain a Bill of Rights; the Framers apparently believed that structural protections were the most important means for protecting liberty. 59. See, e.g., U.S. CONST. art. I, § 2, cl. 1 (voters for Representative “shall have the Qualifications requisite for Electors of the most numerous Branch of the State Legislature”); id. art. I, § 2, cl. 4 (vacancies in House to be filled by writs of election issued by state’s “Executive Authority”); id. art. I, § 3, cl. 1 (Senators from each state to be chosen by “the Legislature thereof ” ); id. art. II, § 1, cl. 2 (presidential electors to be appointed in each state “in such Manner as the Legislature thereof may direct”); id. art. VI, cl. 2 (“the Judges in every State shall be bound” by the U.S. Constitution). For a more detailed discussion, see Michael C. Dorf, The Relevance of Federal Norms for State Separation of Powers, 4 ROGER WILLIAMS U. L. REV. 51 (1998). 60. See THE FEDERALIST NOS. 47, 48 (James Madison) (Jacob E. Cooke ed., 1961). 61. U.S. CONST. art. IV, § 4. Another provision, Article I, Section 10, Clause 1, prohibits states from granting any “Title of Nobility,” which might be construed to forbid a monarchical or aristocratic form of government. I ignore it here because its applicability seems extremely limited and too remote from contemporary considerations of governmental structure. 638 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:625 ing of the Clause is obscure, however, because it has been neither authoritatively construed nor judicially enforced: the Supreme Court has consistently held that questions arising under the Guarantee Clause raise nonjusticiable political questions.62 Even if the Clause were enforceable, however, there are good reasons to think that it imposes structural requirements so minimal as to provide no meaningful constraints on a state’s ability to structure state government in whatever way the people of the state think best protects them from state tyranny. According to Madison, all that is required for a government to qualify as republican is that its officers be elected, even indirectly, by the people.63 The U.S. Supreme Court, moreover, has held that the Guarantee Clause was not intended to call into question any form of state government existing in 1787.64 Given that Rhode Island and Connecticut then operated under their original royal charters, and that the rights to vote and to hold office were severely restricted in many states, it is difficult to imagine even a vigorously enforced Guarantee Clause requiring much of state governmental structure. State polities, then, are generally quite free to experiment in their state constitutions by tailoring internal structures of self-governance to their particular tastes and notions about how liberty is best to be protected. This freedom has been frequently exercised, and in ways that often seem to reflect popular determinations about the trustworthiness or untrustworthiness of state as compared to national power. For example, many state constitutions seem to reflect a popular trust in state power by granting state governments far greater power than the U.S. Constitution grants the national government, and often giving it to them in more concentrated, and thus potentially more dangerous, forms. Thus, unlike the U.S. Constitution, which enumerates government powers, state constitutions tend to grant power to each branch of state government in one immense, undifferentiated and unlimited block. Some state constitutions, moreover, authorize a much more relaxed separation of powers than does the U.S. Constitution. For example, the Rhode Island Constitution permits state legislators to appoint members of executive branch agencies, including 62. See, e.g., Baker v. Carr, 369 U.S. 186, 208-10 (1962); Colegrove v. Green, 328 U.S. 549, 556 (1946); Pacific States Tel. & Tel. Co. v. Oregon, 223 U.S. 118, 150-51 (1912); Luther v. Borden, 48 U.S. (7 How.) 1, 47 (1849). But see New York v. United States, 505 U.S. 144, 183-85 (1992) (noting disagreement as to the scope of nonjusticiability under the Guarantee Clause and concluding that the question need not be resolved in the case at hand); Minor v. Happersett, 88 U.S. (21 Wall.) 162, 175-76 (1874) (refusing to recognize suffrage rights for women under the Guarantee Clause). 63. THE FEDERALIST NO. 39, at 251 (James Madison) (Jacob E. Cooke ed., 1961). The original U.S. Constitution used indirect election for Senators, the President, and the Vice President. U.S. CONST. art. I, § 3; id. art. II, § 1. 64. Minor, 88 U.S. (21 Wall.) at 175-78. 2001] REGULATORY ROLE 639 self-appointment65—forms of cross-branch appointments forbidden under the U.S. Constitution.66 Several state constitutions permit the state legislature to exercise a legislative veto over regulations promulgated by state executive departments,67 a device that violates the constitutional separation of powers laid out for the national government.68 In other cases, state constitutions impose constraints on state governments that have no counterpart on the national level, suggesting heightened popular distrust of state power. For example, many state constitutions divide and disperse state power in ways that the U.S. Constitution does not. Such measures include a highly splintered executive branch in which executive power is divided among multiple officials, each independently elected,69 and a stricter horizontal separation of powers among branches of state government, such as a vigorous nondelegation requirement.70 Many state constitutions restrict the exercise of state legislative power to a much greater degree than the U.S. Constitution restricts congressional power by imposing limitations on taxation and spending, as well as a host of technical requirements such as the single-subject rule or restrictions on special or local legislation.71 A number of state constitutions reveal a deep distrust of state government by creating numerous mechanisms of direct popular control such as referendums, initiatives, recall, and rotation in office.72 If states are free to experiment to this degree, it is difficult to see anything in the U.S. Constitution that might prevent them from going even further afield from the standard, national model of governmental structure. The parliamentary system of government, for example, is the world’s most popular structural form of democratic government. If a state wanted to adopt such a system, it is hard to imagine what provision of the U.S. Constitution might bar it from doing so. Only the extremely unlikely prospect of a radical expansion of no- 65. In re Advisory Opinion to the Governor (Rhode Island Ethics Commission— Separation of Powers), 732 A.2d 55 (R.I. 1999). 66. See U.S. CONST. art. I, § 6, cl. 2; Buckley v. Valeo, 424 U.S. 1, 120-24 (1976). 67. See, e.g., CONN. CONST. art. II; IOWA CONST. art. III, § 40; NEV. CONST. art. III, § 1(2)(c); N.J. CONST. § art. V, & 6; Mead v. Arnell, 791 P.2d 410 (Idaho 1990). For a discussion, see Jim Rossi, Institutional Design and the Lingering Legacy of Antifederalist Separation of Powers Ideals in the States, 52 VAND. L. REV. 1167, 1212-16 (1999). 68. See INS v. Chadha, 462 U.S. 919, 941-43 (1983). 69. See 32 THE BOOK OF THE STATES 33-34 (1998). Regarding Florida in particular, see Deborah K. Kearney, The Florida Cabinet in the Age of Aquarius, 52 FLA. L. REV. 425 (2000). 70. See Rossi, supra note 67, at 1193-1201. 71. ROBERT F. WILLIAMS, STATE CONSTITUTIONAL LAW 796-823, 917-93 (3d ed. 1999). 72. See James A. Gardner, Devolution and the Paradox of Democratic Unresponsiveness, 40 S. TEX. L. REV. 759, 766-67 (1999). 640 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:625 tions of due process73 could, it seems, derail such a state decision concerning the details of its constitutional structure of internal selfgovernance. If the picture I have painted of state autonomy over internal government structure is correct, the question raised by the Bush cases is this: are things any different during presidential elections merely because Article II gives state legislatures a role to play in the selection of presidential electors? The answer, I suggest, is no. III. STATE CONSTITUTIONS AND THE SELECTION OF PRESIDENTIAL ELECTORS For the reasons I have described, state constitutional restraints on the powers of state legislatures represent sober, potentially valuable, liberty-protective measures that may impair a legislature’s ability successfully to abuse its often considerable powers. Article II, Section 1 of the U.S. Constitution grants state legislatures a power that their own state constitutions do not and could not grant: it specifically authorizes state legislatures to select presidential electors, or to determine the manner in which they shall be selected. Like any other power granted to a state legislature, this one is susceptible of abuse. Madison’s theory of self-aggrandizement assumes that legislatures will habitually act so as to increase their own powers at the expense of other political actors,74 but this need not always be the case. History furnishes numerous examples of representative bodies voluntarily turning over power to tyrants of varying sorts. The Athenian Assembly relinquished all its power to the Four Hundred.75 The Roman Senate on several occasions turned over its power to a single dictator.76 In our time, the German Reichstag voted in 1933 to turn over all legislative power to an executive cabinet led by Hitler.77 There is no reason to suppose that American state legislatures are 73. That is, the Court might have to find some kind of individual due process right to be subject only to laws made by a three-branch government whose powers are balanced in a way sufficiently close to the balance established by the U.S. Constitution. The theory for such a holding would be that only this arrangement, and no other, adequately protects personal liberty. Before the Bush decisions, I would have found this prospect laughable. Now I am not so sure. 74. THE FEDERALIST NO. 48, at 333 (James Madison) (Jacob E. Cooke ed., 1961) (“The legislative department is every where extending the sphere of its activity, and drawing all power into its impetuous vortex.”); see also THE FEDERALIST NOS. 10, 47, 51 (James Madison). 75, THUCYDIDES, THE PELOPONNESIAN WAR 525-26 (Rex Warner trans., Penguin Books 1954). 76. See LIVY, THE EARLY HISTORY OF ROME 124-25, 137-38, 213 (Aubrey de Selincourt trans., Penguin Books 1960). 77. WILLIAM L. SHIRER, THE RISE AND FALL OF THE THIRD REICH: A HISTORY OF NAZI GERMANY 198-99 (1960). 2001] REGULATORY ROLE 641 somehow immune from the temptations that have led other legislatures to support tyrants.78 It is far more likely, of course, that any abuse of the power to choose presidential electors would be less dramatic. Legislators might, for example, barter away their votes for political favors or even for cash. This is not an idle possibility. Before the ratification of the Seventeenth Amendment, when U.S. Senators were selected by state legislatures, bribery of state legislators by senatorial candidates was a widespread problem. In 1899, for example, one candidate for Senator handed out over $140,000 of his own money to state legislators in exchange for votes.79 Much is at stake in presidential elections. The office of President has attained a kind of preeminence that the Framers did not foresee and might well have thought undesirably reminiscent of the kingly authority they overthrew. Yet nothing in Article II itself sets any limits on the ways in which state legislatures may select electors. If that process is to be conducted and disciplined by law, that law must, it seems, be supplied by state constitutions. There is little reason, moreover, to think that the people of any state would be particularly inclined to trust their state legislature to perform the critical task of selecting presidential electors completely free of popular guidance and constraints applied by constitutional means. First, there is a long American history of popular distrust of state legislatures and of corresponding efforts to use state constitutions to curb undesirable legislative behavior. A striking number of typical state constitutional provisions represent deliberate public responses to specific acts of state governmental malfeasance. For example, many state constitutions sharply restrict the state’s ability to incur debt.80 These restrictions date mostly from the middle third of the nineteenth century and were adopted in response to a series of disastrous public works expenditures on canals and railroads that caused serious financial difficulty for numerous states.81 Many state constitutions require that the title of a bill accurately reflect its sub78. The Framers clearly believed that a national legislature could not be trusted to resist such temptations. During debates at the 1787 convention concerning presidential selection, a proposal was rejected that would have permitted Congress to select the President. On several occasions, a concern was expressed that foreign nations would intrigue with members of Congress to corrupt the selection process, shaping it to their malevolent whims. See 4 THE FOUNDERS’ CONSTITUTION 536-50 (Philip B. Kurland & Ralph Lerner eds., 1987). 79. C.H. HOEBEKE, THE ROAD TO MASS DEMOCRACY: ORIGINAL INTENT AND THE SEVENTEENTH AMENDMENT 91 (1995); see also Jay S. Bybee, Ulysses at the Mast: Democracy, Federalism, and the Sirens’ Song of the Seventeenth Amendment, 91 NW. U. L. REV. 500, 538-41 (1997). 80. Stewart E. Sterk & Elizabeth S. Goldman, Controlling Legislative Shortsightedness: The Effectiveness of Constitutional Debt Limitations, 1991 WIS. L. REV. 1301, 1305. 81. Id. at 1306-10. 642 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:625 ject. These provisions grew out of the infamous Yazoo scandal of 1795, in which the Georgia state legislature enacted a law whose innocuous title did not accurately reflect the fact that the law’s main purpose was to sell public lands to private speculators at an unconscionably low price.82 Second, popular distrust of state legislative power has typically been strongest in matters relating to the selection of government officials, a phenomenon reflected in an American electoral history of almost continuous democratization. The United States may have been founded by republicans, but it was soon inherited by populists.83 The common current of American populist ideology in its various incarnations holds that the people are at least the equal of their representatives in respect of their competence to assess and to choose among competing public policies. However, populism holds unequivocally that the people are decidedly superior to their representatives in their innate virtue.84 American populist movements, from the Jacksonians through the Progressives, have thus tended to see American politics as a ceaseless struggle by a virtuous and competent citizenry to prevent its will from being obstructed by corrupt officials or unsuitable governmental structures and institutions. As a result, populists have turned frequently to constitutional amendment, most often at the state level, to strengthen popular control over what they understand to be unreliable government agents. This trend reached its height during the Progressive movement, whose leading historian described its purpose as opposing “the control of government by special interests and the prostitution of government to serve the needs of a small minority.”85 Progressives fought this domination by waging largely successful campaigns for reforms such as expanded suffrage, secret ballots, direct primaries, initiatives, referenda, recall, campaign finance disclosure, and easier constitutional amendment, among others, many of which were eventually implemented.86 The purpose of these reforms was, according to the Progressives themselves, “to dissolve the unholy alliance between corrupt business and corrupt politics”87 by giving “a majority of the 82. Millard H. Ruud, No Law Shall Embrace More Than One Subject, 42 MINN. L. REV. 389, 391-92 (1958). 83. Rossi, supra note 67, at 1171-74. 84. See James A. Gardner, Madison’s Hope: Virtue, Self-Interest, and the Design of Electoral Systems, 86 IOWA L. REV. 87, 130-36 (2000). 85. BENJAMIN PARKE DE WITT, THE PROGRESSIVE MOVEMENT 46 (Richard T. Ely ed., 1915). 86. See, e.g., id. chs. 10-11; WILLIAM ALLEN WHITE, THE OLD ORDER CHANGETH (1910), reprinted in THE PROGRESSIVE MOVEMENT, 1900-1915, at 133-136 (Richard Hofstadter ed., 1963); THE PROGRESSIVE PARTY PLATFORM OF 1912, reprinted in id. at 129. See generally John Dinan, Framing a “People’s Government”: State Constitution-Making in the Progressive Era, 30 RUTGERS L.J. 933 (1999). 87. THE PROGRESSIVE PARTY PLATFORM OF 1912, supra note 86, at 128. 2001] REGULATORY ROLE 643 people . . . an easy, direct, and certain control over their government.”88 There are good reasons, then, to think that state constitutional restraints could serve a potentially valuable, liberty-protective function by regulating state processes of elector selection. There are also good reasons to think that state polities are likely to value highly their ability to impose such constraints on what would otherwise be unrestricted state legislative power. Conversely, and contrary to the Supreme Court’s intimations in Bush I, there is no good reason to think that Article II impairs the ability of state polities to use state constitutions for just this purpose. Our theory of government is positivist. It holds that the organs of any government owe their existence, their legitimacy, and their powers to some founding act of popular sovereignty, the terms of which are usually embodied in a constitution.89 This theory is generally thought to apply as much to state governments as to the national government.90 The American system of federalism, to be sure, blurs some of the distinctions between state and national sovereignty by constituting a system of overlapping authority, and by making state power subordinate in some respects to national power. It does not, however, go so far as to undermine the positivist premise that state governments are created, and their powers granted and restrained, by state polities in state constitutions rather than by the national polity in the U.S. Constitution. The Supreme Court’s decision in Bush I suggests that these otherwise valid principles do not apply to state legislative power during presidential elections because the power to select presidential electors is one granted to state legislatures directly and exclusively by the somewhat unusual provisions of Article II of the U.S. Constitution. A power granted to an organ of state government by the national Constitution, the Court apparently believes, cannot be modified by a mere state constitution, which is by definition subordinate. Yet the interplay between national and state constitutional authority under Article II is neither as unique, nor as straightforwardly hierarchical, as the Court assumes. In fact, the national Constitution 88. DE WITT, supra note 85, at 196. 89. This foundational proposition runs through the canonical texts of American political theory such as the Declaration of Independence, the preamble of the U.S. Constitution, and McCulloch v. Maryland, 17 U.S. 316 (1819). 90. See Sturges v. Crowninshield, 17 U.S. 122, 193 (1819) (“[I]t was neither necessary nor proper to define [in the U.S. Constitution] the powers retained by the States. These powers proceed, not from the people of America, but from the people of the several States.”). For more general discussions, see James A. Gardner, Federalism and the Problem of Political Subcommunities, in TO PROMOTE THE GENERAL WELFARE: A COMMUNITARIAN LEGAL READER 271, 282-83 (David E. Carney ed., 1999); G. ALAN TARR, UNDERSTANDING STATE CONSTITUTIONS 69-70 (1998). 644 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:625 routinely grants power to the organs of state government in many situations in which no one seems to question the authority of state constitutions to guide and limit the ways that state governmental actors may exercise the granted authority. For example, the U.S. Supreme Court frequently rules that particular state or national laws do not violate any provision of the Bill of Rights or Fourteenth Amendment. Each such ruling authorizes every legislature in the land to enact laws identical to the one upheld, free of national constitutional interference. Yet such rulings have never been understood to insulate state legislatures from restrictions and conditions on the exercise of legislative power imposed by state constitutions. The fact that the U.S. Constitution permits a state legislature to enact a law by no means frees the legislature from the host of prohibitions and restrictions, whether substantive or procedural, that state constitutions often impose. Indeed, the U.S. Supreme Court itself has often observed that state polities are free to bind state legislatures to more demanding constitutional standards for observing individual rights than the national Constitution requires of them.91 Similarly, the U.S. Constitution authorizes Congress to use its spending power to provide financial inducements to states to take actions that Congress desires.92 A necessary corollary of the Constitution’s authorization to Congress to dangle financial inducements before state governments is the complementary authorization to state governments both to comply with congressionally imposed conditions and to accept the money. Yet no one, I take it, would dispute the authority of state constitutions to limit, condition, or even countermand this nationally granted authority to take actions necessary to receive national funds.93 If a state constitution bars the legislature from enacting the federally required law, or subjects the collection or use of state revenues to budgetary or appropriations requirements, or imposes upon the state legislature procedural limitations that prevent it from accepting the offered gift, surely these limitations are not displaced simply because the national Constitution has attempted to create some new form of state legislative authority. It might be objected that the kinds of national authorization just described are indirect, whereas Article II consists of a direct grant of 91. The most prominent such case is PruneYard Shopping Center v. Robins, 447 U.S. 74 (1980). The most influential statement of this principle by a sitting Justice, however, is Justice Brennan’s 1977 article, William J. Brennan, Jr., State Constitutions and the Protection of Individual Rights, 90 HARV. L. REV. 489 (1977). 92. See South Dakota v. Dole, 483 U.S. 203 (1987). 93. See North Carolina ex rel. Morrow v. Califano, 445 F. Supp. 532 (E.D.N.C. 1977) (three-judge court); William Van Alstyne, “Thirty Pieces of Silver” for the Rights of Your People: Irresistible Offers Reconsidered as a Matter of State Constitutional Law, 16 HARV. J.L. & PUB. POL’Y 303 (1993). 2001] REGULATORY ROLE 645 state legislative authority. Yet the U.S. Constitution also grants authority directly to organs of state government in circumstances where state constitutions routinely limit or modify it. For example, the U.S. Constitution directly authorizes state courts to hear and decide questions of federal law.94 This grant of authority, however, has long been understood to occur subject to state constitutional rules establishing the jurisdiction and powers of state courts. The U.S. Supreme Court has held, for instance, that state constitutions may establish standing rules broader than those permitted to federal courts under Article III and that state courts may hear federal law claims brought by parties who would lack standing to bring them in federal court.95 Indeed, the Court has gone further: it has held that its own jurisdiction may be expanded by standing rules established by state constitutions, thus permitting the Court to hear appeals of federal questions from state courts utilizing standing rules that would require dismissal if the same parties had brought the case up through the federal judicial system.96 The Supreme Court’s position in Bush I is all the more surprising because the text and context of Article II suggest a ready alternative construction that would be far more consistent with the practices discussed above. If a state’s constitution forbids its legislature to exercise the power to select presidential electors in a way that comports with the requirements of Article II, the obvious remedy is to exclude the state’s electors from the presidential selection process—a power that Congress has already formalized by statute.97 There is simply no reason for the Court, particularly in light of the anti-commandeering principles it has articulated under the Commerce Clause,98 to construe Article II as dictating directly the content and organization of state legislative power when a system of structural incentives could easily achieve the same goals by penalizing the inappropriate use of otherwise autonomous state power.99 Such a construction of Article II would give effect to the vital state constitutional regulatory function, while at the same time recognizing the requirement that national policy reign supreme in matters concerning national elections. 94. See Martin v. Hunter’s Lessee, 14 U.S. (1 Wheat.) 304, 342 (1816). 95. See ASARCO, Inc. v. Kadish, 490 U.S. 605, 612-25 (1989). 96. Id. at 618-24. 97. 3 U.S.C. § 15 (1994). 98. See Printz v. United States, 521 U.S. 898 (1997); New York v. United States, 505 U.S. 144 (1992). 99. Section 2 of the Fourteenth Amendment furnishes an appealing model in these circumstances. That provision encourages states to grant the franchise as widely as possible, without requiring them to do so. States that disenfranchise males over the age of twenty-one are instead penalized by suffering a proportionate reduction in congressional representation. U.S. CONST. amend. XIV, § 2. 646 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:625 This conclusion would be in no way inconsistent with some of the Court’s interpretations of other provisions of the U.S. Constitution that structure national political processes, most notably the Qualifications Clauses and the amendment procedures of Article V. In U.S. Term Limits, Inc. v. Thornton100 the Court held that the list of qualifications for members of Congress set out in Article I of the U.S. Constitution101 is not subject to alteration by state constitutional provisions imposing additional or different qualifications on candidates running from the state in question. And in Hawke v. Smith,102 the Court invalidated a state constitutional provision requiring approval by popular referendum of state legislative ratifications of amendments to the U.S. Constitution on the ground that Article V requires that ratification be performed by the state legislature itself.103 These cases, however, deal with provisions of the national Constitution that call upon American citizens to act exclusively in their capacity as citizens of the nation, in the former case by electing members of Congress, and in the latter by deciding whether to amend the national Constitution. Article V makes use of the state legislative form for this purpose, but it is clear that the state legislature is functioning in this circumstance only as a convenient vehicle for the state-by-state aggregation of national opinion rather than as an agent of a distinct state polity.104 Not all national political processes, however, require Americans to act in their capacity as national citizens. As Madison noted, the U.S. Constitution is partly national and partly federal.105 In the former capacity, the Constitution creates political processes that are truly national and that treat citizens as a unified national polity. The Qualifications Clauses and Article V fall into this category. In its federal capacity, however, the Constitution builds some of its political structures directly on the backs of state processes in which citizens act not as a single, national citizenry but as disaggregated members of the various state polities.106 The preeminent example of 100. 514 U.S. 779 (1995). 101. See U.S. CONST. art. I, § 2, cl. 2 (qualifications of Representatives); id. art. I, § 3, cl. 3 (qualifications of Senators). 102. 253 U.S. 221 (1920). 103. Id. at 229-31. 104. As the Court observed in Hawke, “ratification by a State of a constitutional amendment is not an act of legislation within the proper sense of the word.” Id. at 229. The same can be said of the original 1789 ratification of the Constitution by specially designated ratifying conventions. See U.S. CONST. art. VII; U.S. Term Limits, Inc., 514 U.S. at 821-22; McCulloch v. Maryland, 17 U.S. (4 Wheat.) 316, 402 (1819). 105. THE FEDERALIST NO. 39, supra note 63, at 253-57. 106. That, at least, is the theory. In practice, things have not turned out that way. Today, it seems to me, when Americans act in a disaggregated way—which they frequently do—the cleavages along which they divide bear no meaningful relation to state boundaries. In contrast, when they act pursuant to formal constitutional processes that contemplate 2001] REGULATORY ROLE 647 this is undoubtedly the U.S. Constitution’s treatment of the right to vote. As the U.S. Supreme Court has often pointed out, the Constitution grants no one the right to vote in national elections.107 Instead, it adopts by reference the voter qualifications established by the states: “the Electors in each State shall have the Qualifications requisite for Electors of the most numerous Branch of the State Legislature.”108 In this setting, state constitutional provisions establishing and regulating the right to vote clearly control and are simply made a part of the national constitutional scheme. State constitutional regulation of state voting qualifications is not somehow displaced in this situation merely because the U.S. Constitution sees fit to make use of existing state processes as a step toward accomplishing national political objectives. A national constitutional structure even more directly on point is the pre-Seventeenth Amendment apparatus governing selection of United States Senators.109 Under the original scheme, U.S. Senators were elected by state legislatures: Article I, Section 3 of the U.S. Constitution expressly granted the power to conduct senatorial elections to “the Legislature” of each state. Moreover, Article I, Section 4 expressly provides that “The Times, Places and Manner of holding Elections for Senator . . . shall be prescribed in each State by the Legislature thereof . . . .”110 Thus, the power to conduct and regulate senatorial elections was allocated expressly, in two separate provisions, to the state legislature specifically rather than the state generally just as it is in Article II. Despite this doubly direct delegation of authority, no one appears to have assumed that state legislatures, in performing a vital national political function, somehow became immune in the performance of that function from the normal course of regulation by state constitutions. Indeed, the assumption appears to have been quite the opposite. Georgia’s 1798 Constitution, for example, the first it adopted following ratification of the national Constitution in 1789, imposed direct, albeit minimal, regulations on the legislative process for electing senators: disaggregation along state lines, they act almost always in response to forces that are national in scope. See Gardner, supra note 90; Larry D. Kramer, Putting the Politics Back Into the Political Safeguards of Federalism, 100 COLUM. L. REV. 215 (2000). 107. Rodriguez v. Popular Democratic Party, 457 U.S. 1, 9 (1982) (quoting San Antonio Indep. Sch. Dist. v. Rodriguez, 411 U.S. 1, 35 n.78 (1973)); Minor v. Happersett, 88 U.S. (21 Wall.) 162, 178 (1875). 108. U.S. CONST. art. I, § 2, cl. 1; see also id. amend. XVII, § 1 (same wording). 109. Id. art. I, § 3, cl. 1, amended by id. amend. XVII (1913). 110. Id. art. I, § 4 (emphasis added). Congress, however, is given the superseding power to regulate congressional elections. The full provision reads: “The Times, Places and Manner of holding Elections for Senators and Representatives, shall be prescribed in each State by the Legislature thereof; but the Congress may at any time by Law make or alter such Regulations, except as to the Places of chusing Senators.” Id. art. I, § 4, cl. 1. 648 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:625 All elections by the general assembly shall be by joint ballot of both branches of the legislature; and when the senate and house of representatives unite for the purpose of electing, they shall meet in the representative chamber, and the president of the senate shall in such cases preside, receive the ballots, and declare the person or persons elected.111 Vermont’s 1793 Constitution also contained a provision regulating elections for U.S. Senator: All elections, whether by the people or the Legislature, shall be free and voluntary: and any elector who shall receive any gift or reward for his vote, in meat, drink, monies or otherwise, shall forfeit his right to elect at that time, and suffer such other penalty as the law shall direct . . . .112 Numerous other state constitutional provisions also imposed modest housekeeping requirements on legislative elections for U.S. Senator.113 Florida’s 1868 Constitution was even more direct: “The legislature shall elect United States Senators in the manner prescribed by the Congress of the United States and by this constitution.”114 Presidential elections possess this same federal, rather than national, character. The Constitution creates a system in which Presidents are selected after a process of aggregating one by one the preferences of individual state polities. The Framers did not believe that a single chief executive could be selected using a process continental in scope, if only because they believed no one could know the qualities of good candidates from around an immense nation,115 and they did not try to construct such a system. Instead, they assumed that 111. GA. CONST. art. IV, § 2 (1798). It is clear that this provision applies to elections, rather than ordinary legislative voting, because the provision goes on to state: “In all elections by the people the electors shall vote viva voce until the legislature shall otherwise direct.” Id. 112. VT. CONST. ch. II, § 34 (1793) (emphasis added). 113. A fairly common type of regulation in early state constitutions provided: “In all elections by the general assembly, the members thereof shall vote viva voce, and the votes shall be entered on the journals.” E.g., ALA. CONST. art. VI, § 6 (1819). To similar effect were: ARK. CONST. art. IV, § 19 (1836); CAL. CONST. art. IV, § 38 (1849); FLA. CONST. art. VI, § 17 (1838); IND. CONST. art. II, § 13 (1851). 114. FLA. CONST. art. V, § 29 (1868). This provision was carried forward into the 1885 constitution, FLA. CONST. art. III, § 31 (1885), which remained in force until ratification of the Seventeenth Amendment. 115. For example, at the 1787 convention, Elbridge Gerry argued that the people should play no role in presidential selection because they were “too little informed of personal characters in large districts . . . .” 3 THE FOUNDERS’ CONSTITUTION, supra note 78, at 536 (statement of June 2). Roger Sherman similarly argued that the people “will never be sufficiently informed of characters . . . .” Id. at 537 (July 17). Colonel Mason also believed that “[t]he extent of the Country renders it impossible that the people can have the requisite capacity to judge of the respective pretensions of the Candidates.” Id. at 538 (July 17). But see id. at 537-38 (statement of Governour Morris, July 17): “[t]hey will not be uninformed of those great & illustrious characters which have merited their esteem & confidence.” 2001] REGULATORY ROLE 649 presidential selection would be dominated by local knowledge and local biases, which led them to require that the presidential and vicepresidential candidates be from different states.116 Although national political parties have now extraconstitutionally coordinated the presidential selection process,117 the Constitution itself does not contemplate any fully national aspect to presidential selection. The electors from each state meet and deliberate only among themselves, for example, without deliberating with electors from other states.118 And, of course, if the Electoral College produces no majority winner, the selection process moves to the House of Representatives, where each state gets a single vote.119 In selecting presidential electors, then, or in providing for their selection, state legislatures act not as agents of the national citizenry but as agents of the citizenry of their own states. In granting this authority to state legislatures, the U.S. Constitution thus makes a grant of national power not to an agency of the nation but to an organ of state government, functioning as an organ of state government. When the national Constitution grants power in this particular way, it takes the organs of state government as it finds them— subject to control by the state polity and acting under whatever instructions, embodied in the state constitution, that the state polity has seen fit to impose for the protection of its liberty.120 To take a relatively easy case, suppose, for example, that a state legislature were to use its Article II power to delegate the selection of presidential electors not to the people of the state but to some other organ of state government, such as the state supreme court. If state constitutional separation of powers principles prohibited state judges from exercising this function,121 Article II of the U.S. Constitution would not somehow override the state constitutional limitation on judicial authority. By the same token, it seems to me entirely within the reach of legitimate state constitutional power either to require the legislature to select presidential electors itself, or to forbid it from doing so and to require the legislature instead to provide for the selection of electors by popular election. This is, perhaps, not so far 116. See U.S. CONST. amend. XII. 117. See Kramer, supra note 106; see also E.E. SCHATTSCHNEIDER, PARTY GOVERNMENT 1-12 (Dr. Phillips Bradley ed., 1942). 118. U.S. CONST. amend. XII. 119. Id. 120. This is the very position the Court took in Smiley v. Holm, 285 U.S. 355, 367-68 (1932), where it rejected an argument that Article II forbids state laws regulating presidential elections to be submitted to the governor for signature or veto pursuant to the state’s normal legislative processes established in its state constitution. 121. Cf. Mistretta v. United States, 488 U.S. 361 (1989) (holding under U.S. Constitution that Article III judges may not be given duties incompatible with their judicial duties or that undermine the integrity of the judicial branch). 650 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:625 from the position the Florida Constitution takes when it directly authorizes the state legislature to regulate elections, including presidential elections, by law.122 Apparently, Congress understands these principles better than the Court does, for it has never claimed the power to regulate directly the process by which state legislatures select presidential electors, a power we might expect Congress to have if state legislatures, in choosing electors, were free from any controls originating at the state level.123 Instead, such regulations as Congress has enacted all deal with how electoral votes are canvassed and counted;124 they deal, that is to say, solely with the output of state legislative processes. Significantly, the only occasion on which Congress has intruded directly into the state legislative chamber to regulate state legislative processes is in a pre-Seventeenth Amendment statute establishing procedures for senatorial elections.125 However, Article I, Section 4 of 122. See FLA. CONST. art. VI, § 1. 123. See Burroughs & Cannon v. United States, 290 U.S. 534 (1934), in which the Court rejected a challenge to the Federal Corrupt Practices Act of 1925 regulating contributions by political committees. The defendants, charged with violating the Act, moved to dismiss their indictments on the ground that Congress lacked the power to regulate the way in which states select presidential electors, claiming that congressional power was limited by Article II, Section 1, Clause 4 to setting the date and time of voting for electors. The Court rejected this argument as taking too “narrow a view of the powers of Congress”: The congressional act under review seeks to preserve the purity of presidential and vice-presidential elections. Neither in purpose nor in effect does it interfere with the power of a state to appoint electors or the manner in which their appointment shall be made. It deals with political committees organized for the purpose of influencing elections in two or more states . . . . Its operation, therefore, is confined to situations which, if not beyond the power of the state to deal with at all, are beyond its power to deal with adequately. Id. at 544-45. While presidential electors are not officers or agents of the federal government, they exercise federal functions under, and discharge duties by virtue of authority conferred by, the Constitution of the United States. The President is vested with the executive power of the nation. The importance of his election and the vital character of its relationship to and effect upon the welfare and safety of the whole people cannot be too strongly stated. To say that Congress is without power to pass appropriate legislation to safeguard such an election from the improper use of money to influence the result is to deny to the nation in a vital particular the power of self protection. Congress, undoubtedly, possesses that power, as it possesses every other power essential to preserve the departments and institutions of the general government from impairment or destruction, whether threatened by force or by corruption. 124. See 3 U.S.C. §§ 1-18 (1994). 125. 14 Stat. 243, ch. 245, § 1 (1866). Each house [of the State Legislature] shall openly, by a viva voce of each member present, name one person for senator in Congress from said State, and the name of the person so voted for, who shall have a majority of the whole number of votes cast in each house shall be entered on the journal of each house by the clerk or secretary thereof; but if either house shall fail to give such majority to any person on said day, that fact shall be entered on the journal. At twelve o’clock, meridian, of the day following . . . the members of the two houses shall convene in joint assembly and the journal of each house shall then be read, and if the same person shall have received a majority of all the votes in each house, such person shall be declared duly elected senator . . . ; but if the same person 2001] REGULATORY ROLE 651 the U.S. Constitution, unlike Article II, Section 1, grants this power expressly to Congress: “The Times, Places and Manner of holding Elections for Senators and Representatives, shall be prescribed in each State by the Legislature thereof; but the Congress may at any time by Law make or alter such Regulations, except as to the Places of chusing Senators.”126 Thus Congress, appropriately, has never interpreted its own powers to include regulatory oversight of state legislative selection of presidential electors. Conversely, state constitutions, for their part, have never sought to regulate the processes by which national officers are certified and installed. All parties except the Court, then, seem to have understood that there is a boundary dividing state and national authority in presidential elections. This understanding makes the Court’s decision in Bush I all the more dangerous, for if Congress is correct that it lacks the power to regulate state legislative processes; and if the Court is prepared to cut off state legislatures engaged in selecting presidential electors from state constitutional oversight; then state legislatures are subject to no legal constraints at all. The national system still would not be without recourse: Congress might still reject a state’s electoral votes if they have been somehow tainted. But it seems both unnecessary and unwise to place that protection solely in the hands of congressional political judgments when state constitutions are available to do what they do best: restrain and control state legislative power by providing binding legal standards, judicially enforceable by independent state courts. IV. DECENTRALIZATION OF STATE ELECTORAL ADMINISTRATION Having embarked on the enterprise of displacing state internal structural decisions in Bush I, the Court proceeded to insert itself into another such decision in Bush II. Though somewhat better disguised, the Court’s equal protection analysis in Bush II provides cover for an intrusive ruling that constitutionalizes the Court’s disapproval of Florida’s decision to decentralize the administration of presidential and other statewide elections. As discussed earlier, the principal function of a state constitution is to determine how much power the state government will exercise and the ways in which it will be distributed. These decisions affect the liberty of the state citizenry in three ways. First, by allocating to shall not have received a majority of the votes in each house, . . . the joint assembly shall then proceed to choose, by a viva voce vote of each member present a person for the purpose aforesaid, and the person having a majority of all the votes of the said joint assembly, a majority of all the members elected to both houses being present and voting, shall be declared duly elected . . . . Id. 126. U.S. CONST. art. I, § 4, cl. 1 (emphasis added). 652 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:625 the state government a certain degree of affirmative power, the state polity determines the ability of the state government to use its powers directly to achieve the public good. Second, the decision determines how much power the state government will have at its disposal to monitor and, if necessary, check abuses of national power by the federal government. Third, decisions concerning the internal structure of state power influence the degree to which state power may be deployed against itself. Pitting state power against itself institutionalizes a form of governmental self-restraint that reduces the threat to liberty created by granting the state government powers sufficient to achieve the first two objectives. To make these constitutional decisions, a state polity must weigh its fears of state and national power, respectively, and balance those fears against the hopes and ambitions that caused it to create and empower a government in the first place. A state polity that fears state power—and all do to some degree— has at its disposal numerous methods for restraining that power. As we have seen, one such method involves dispersing state power. State power may be dispersed horizontally, of course, through the standard practice of separating power into three branches. Indeed, most states have gone beyond the national model of separation of powers and have dispersed power horizontally to an even greater degree by subdividing the executive power and providing for the independent election of sub-gubernatorial executive branch officials.127 Like national power, however, state power can also be dispersed vertically. The state level analogue to federalism involves the equally familiar distribution of power between state and local governments such as counties, towns, or cities. Exactly how much power a state polity chooses to withdraw from the centralized state government and allocate to decentralized local governments depends in part, as it does on the national level, on how much the polity is inclined to trust each level of government. In many states, local governments are granted little autonomous power. Municipal governments have often been regarded in the United States as administrative subdivisions of the states, “created as convenient agencies for exercising such of the governmental powers of the State as may be entrusted to them,”128 and most state constitutions grant the state legislature plenary power to create or dissolve local governments and to delegate to them whatever powers it chooses. Dillon’s Rule, traditionally followed throughout the United States, holds additionally that the scope of 127. New Jersey, which vests all executive power exclusively in its governor and elects no sub-gubernatorial officials (not even a lieutenant governor), is the prominent exception. See N.J. CONST. art. V. 128. Hunter v. Pittsburgh, 207 U.S. 161, 178 (1907). 2001] REGULATORY ROLE 653 powers granted to local governments should be narrowly construed.129 State governments that follow these practices thus may possess the power to decide which local governments will exist, if any; their precise configuration; the amount of power such governments will possess; and the circumstances in which that power may be used—all of which may be revoked or modified by the legislature at its pleasure. From the point of view of vertical separation of powers, this arrangement is functionally the equivalent of allowing Congress to decide whether and how much authority state governments should exercise. It is not an arrangement, in other words, that can be expected to produce an effective check from below on abuses of state power. To allocate state power in this way is to grant power to state governments in a far more concentrated form than the form in which similar powers are allocated to the national government. Yet power concentrated, as Madison warned, is far more dangerous than power dispersed. That a state polity chooses nevertheless to forgo a vertical check and grant power to the state government in this concentrated form suggests a degree of popular confidence in state government that has no counterpart on the national level. Not every state follows this approach, however. In some states, local governments are granted often considerable power through constitutional home rule provisions—provisions that place these local powers beyond the authority of state legislatures to invade or displace.130 In these states, popular trust in local government may be far greater, at least with respect to the purposes enumerated in home rule provisions. Florida, it seems, has chosen a middle approach toward its local governments. The Florida Constitution strikes the balance between state and local power generally in favor of the state, but it sets aside a few areas in which power is given unequivocally to local governments. Municipalities, for example, have presumptive independence under the Florida Constitution, which provides that “[m]unicipalities shall have governmental, corporate and proprietary powers to enable them to conduct municipal government, perform municipal functions and render municipal services, and may exercise any power for municipal purposes except as otherwise provided by law.”131 Yet nothing in this provision requires any municipalities to exist at all, and even 129. See Richard Briffault, Our Localism: Part I—The Structure of Local Government Law, 90 COLUM. L. REV. 1, 8 (1990). 130. See 1 CHESTER JAMES ANTIEAU & JOHN MICHAEL ANTIEAU, ANTIEAU’S LOCAL GOVERNMENT LAW § 3.14 (1996). 131. FLA. CONST. art. VIII, § 2(b). 654 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:625 existing municipalities, though granted potentially broad powers,132 may see those powers revoked or altered at the pleasure of the legislature. Things stand differently in Florida regarding counties. The Florida Constitution expressly requires that “[t]he state shall be divided by law into political subdivisions called counties.”133 It then proceeds to grant expressly to counties the exclusive power to conduct all statewide and county elections.134 Under the Florida scheme, then, the state itself has been stripped of the power to administer elections not only of officials of local governments, but of its own officials. This is about as clear an indication as one can expect to find that the citizens of Florida do not trust state officials with the administration of elections, or at least do not trust them to perform this function as much as they trust their local officials to do so. It is not entirely clear what particular considerations or historical experiences caused the people of Florida to allocate power in this way, but it is easy to imagine why a polity might think such an arrangement more conducive to its liberty than a centrally administered state electoral system. First, a decentralized electoral system is more difficult to bring under the control of a single person, group, or party. Second, unlike state election officials, local election officials are ready at hand, visible, well known in their communities, and accountable to the local populace. Third, allowing state officials direct control over the very electoral system that could keep them in power might provide them with the means and the temptation to manipulate the system so as to secure their perpetuation in office. Whatever the reason, Floridians have made a decision about how best to protect their liberty from invasions at the hands of their own state government, a decision that was theirs to make and is not subject to review under any principle of national constitutional law. Or at least that was the case prior to Bush II. In my view, the Court’s equal protection ruling in Bush II represents an attack on Floridians’ decision to protect their liberty by constitutionally decentralizing the administration of statewide elections. The Court’s ruling reacts to this entirely legitimate, internal, vertical separation of state powers by applying to Florida’s system, under cover of the 132. See State v. Panama City Beach, 529 So. 2d 250 (Fla. 1988); TALBOT D’ALEMBERTE, THE FLORIDA STATE CONSTITUTION: A REFERENCE GUIDE 124-25 (G. Alan Tarr ed., 1991). 133. FLA. CONST. art. VIII, § 1(a). 134. Id. art. VI, § 5(a) (“A general election shall be held in each county on the first Tuesday after the first Monday in November of each even-numbered year to choose a successor to each elective state and county officer whose term will expire before the next general election . . . . ”). 2001] REGULATORY ROLE 655 Equal Protection Clause, what amounts to a federal nondelegation doctrine for states. My conclusion rests to a great extent on the facial implausibility of the Court’s equal protection analysis, which I shall briefly rehearse. To begin with, there is no suggestion in Bush II that intentional discrimination played any role in the structuring of Florida’s vote-counting processes.135 This in itself is not fatal, however, since intentional discrimination has never been required in cases falling within the fundamental rights strand of equal protection analysis.136 What has always been required in every equal protection case, however, is the identification of some class of persons disadvantaged by the challenged government action. Neither the parties nor the Court ever identified such a class, and indeed they could not do so. No class of voters within any county could have been systematically disadvantaged by that county’s choice of vote-counting method, since the ballots of all voters within the county were counted by the same means. The essence of Bush’s constitutional claim, moreover, was that the counting and invalidation of ballots fell essentially randomly, and was thus unconstitutionally arbitrary,137 a premise inconsistent with the existence of an identifiable disadvantaged class. The only plausible candidate for a disadvantaged class might consist of the class of all voters for statewide candidates who resided in counties that utilized the less-reliable methods of vote counting. Residents of these counties were at a disadvantage compared to residents of counties that used more reliable methods of vote counting because their ballots were disqualified at a higher rate, thus disproportionately diminishing the voices of such voters in determining the outcome of statewide races. The Court did not speak in these terms, but even if it had, such a claim contains a fatal flaw. When the people of a Florida county, through their elected representatives, adopt a substandard method of counting votes, the wound is self-inflicted, through ordinary processes of collective self-governance. It is, in other words, nothing more than an ordinary instance of interjurisdictional variation in the law, a commonplace phenomenon that has never been thought to raise equal protection concerns.138 135. Allegations were later made, in other forums, of intentional discrimination against minority voters. See Dana Canedy, Rights Panel Begins Inquiry Into Florida’s Voting System, N.Y. TIMES, Jan. 12, 2001, at A20. Even so, I do not believe that anyone has gone so far as to accuse the state of deliberately utilizing a system of punch-card balloting in heavily minority districts for the purpose of reducing the number of countable votes from those districts. 136. See Ira C. Lupu, Untangling the Strands of the Fourteenth Amendment, 77 MICH. L. REV. 981 (1979). 137. Bush v. Gore, 531 U.S. 98, 105-06 (2000). 138. See Gerald L. Neuman, Territorial Discrimination, Equal Protection, and SelfDetermination, 135 U. PA. L. REV. 261 (1987). 656 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:625 Having before it no cognizable class of systematically disadvantaged voters, the Court resorts to different language, the language of arbitrary deprivation of the right to vote. “[T]he State,” says the Court, “may not, by . . . arbitrary and disparate treatment, value one person’s vote over that of another.”139 However, the right to be free from arbitrary deprivations of constitutionally protected interests is one that sounds in due process, not equal protection. The Court attempts nevertheless to shoehorn the problem into equal protection terms, by citing Harper v. Virginia Board of Elections140 and Reynolds v. Sims141 to support its analysis. Neither of these cases helps, unfortunately, because both involved disadvantageous treatment of identifiable classes of voters: those who could not afford a poll tax in the former case, and those who lived in populous urban districts in the latter. Here, there is no such identifiable class. The Court’s insistence on an equal protection analysis grows more puzzling as the opinion unfolds. Florida’s recount mechanisms, the Court asserts, “do not satisfy the minimum requirement for nonarbitrary treatment of voters . . . .”142 This was so, the Court explains, because the “intent of the voter” standard to be applied under Florida law contains an “absence of specific standards to ensure its equal application.”143 It is, in other words, too vague. This vagueness then causes the arbitrary treatment of voters, according to the Court. Yet both of these related problems—arbitrariness and vagueness—have historically been treated by the Court as raising questions of due process, not equal protection. Both deal with the direct relationship of the law to individuals, not the relative position under the law of one individual compared to another. The right to nonarbitrary treatment under the law is, after all, the essence of due process rationality review.144 When arbitrary treatment results from a law’s vagueness, the problem has always been treated as one of due process.145 But even if the problem the Court identifies is couched more appropriately as one of due process vagueness, the claim still contains fatal flaws. Under the Due Process Clause, vagueness in a law raises 139. Bush v. Gore, 531 U.S. at 104-05. 140. 383 U.S. 663 (1966). 141. 377 U.S. 533 (1964). 142. Bush v. Gore, 531 U.S. at 105. 143. Id. at 105-06. 144. See Williamson v. Lee Optical, 348 U.S. 483 (1955). 145. Vagueness, under the Due Process Clause, becomes a problem when a law is so vague that persons “of common intelligence must necessarily guess at its meaning and differ as to its application . . . .” Connally v. Gen. Constr. Co., 269 U.S. 385, 391 (1926). There is another, more specialized vagueness doctrine that arises under the First Amendment to combat statutory vagueness which chills speech, Grayned v. City of Rockford, 408 U.S. 104 (1972), but that doctrine is inapplicable here. 2001] REGULATORY ROLE 657 constitutional problems only when the law is so vague that individuals become unable to understand it and thus to conform their conduct to the law.146 That is why legislative vagueness most often rises to the level of a due process problem in the criminal setting. The vagueness here is very different: the recount law requiring application of the intent-of-the-voter standard is addressed not to individuals, but to government officials. There is no claim in Bush II that voting instructions provided to individual voters were too vague. Rather, the claim is that Florida law provides county election officials with insufficient guidance as to the proper method of counting ballots, allowing such officials to exercise inappropriately unconstrained discretion, which in turn “has led to unequal evaluation of ballots . . . .”147 “[E]ach of the counties,” the Court complains, “used varying standards to determine what was a legal vote. Broward County used a more forgiving standard than Palm Beach County, and uncovered almost three times as many new votes . . . .”148 There is a constitutional doctrine that covers situations where legal rules meant to guide government officials instead vest them with excessive discretion, but it is a doctrine neither of equal protection nor due process. It is the nondelegation doctrine. The nondelegation doctrine prohibits legislatures from delegating power to executive branch officials in terms so vague as to give them essentially no legislative guidance.149 That is precisely the problem here. The defect in Florida’s election procedure, according to the Court, is the lack of “adequate statewide standards for determining what is a legal vote . . . .”150 Had this election been conducted by national administrative officials under laws enacted by Congress, the Court might have some business deciding that the intent-of-the-voter standard gave those officials a degree of discretion so great as to violate the nondelegation doctrine, and thus violated the constitutionally mandated separation of powers. Obviously, however, the U.S. Supreme Court had no business holding that the State of Florida violated any kind of separation of powers doctrine. The internal structure of Florida government, including its vertical separation of powers, is for the state’s citizenry to decide. Moreover, as shown earlier, states have exercised this freedom by choosing a wide variety of structures that allocate govern- 146. See, e.g., City of Chi. v. Morales, 527 U.S. 41 (1999); Papachristou v. City of Jacksonville, 405 U.S. 156 (1972). 147. Bush v. Gore, 531 U.S. at 106. 148. Id. at 107. 149. See Yakus v. United States, 321 U.S. 414 (1944); A.L.A. Schechter Poultry Corp. v. United States, 295 U.S. 495 (1935); J.W. Hampton, Jr. & Co. v. United States, 276 U.S. 394 (1928). 150. Bush v. Gore, 531 U.S. at 110. 658 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:625 ment authority in different ways and which separate the allocated powers with different degrees of strictness. In fact, the strength of nondelegation principles is an area in which state constitutional practices differ widely, not only from one state constitution to another but as between the state and national constitutions.151 The Court took pains to disclaim any intention of interfering with state discretion in this area: “The question before the Court is not whether local entities, in the exercise of their expertise, may develop different systems for implementing elections.”152 Yet that is pretty much what the Court has done. By basing its ruling, albeit indirectly, on its objection to Florida’s decentralization of election administration, the Court has interfered with, and for all practical purposes invalidated, a measure crafted by the state polity for the better protection of its liberty. Such a ruling fails to pay adequate respect to the national constitutional structure of federalism, a structure that allows state polities alone to decide how best to balance state power against national and local power. CONCLUSION The Court’s decisions in Bush I and Bush II may be criticized on many grounds, as the papers in this symposium reveal. But one of the worst aspects of the opinions is the way they invade state autonomy over internal structures of state self-governance. The Court’s decisions conflict with basic principles of federalism, which envision a shared, federal process of presidential selection in which state and national power each play distinct roles. The Bush cases transform the governance of presidential elections from a shared enterprise of the national and state legislative and judicial branches to one conducted exclusively by national actors. The result is a centralization of power in the national government—and particularly in the national judiciary—considerably at odds with the purposes and methods of federalism. In other recent contexts, the U.S. Supreme Court has taken great pains to assert the importance of federalism to the protection of individual liberty. It has gone so far as to curb severely some major pieces of national civil rights legislation on the ground that they may not be enforced against the states without their consent153—on the ground, that is to say, that the direct protection of individual liberty through specifically targeted legislation must take a backseat to the indirect protection of individual liberty through structural federal151. See Rossi, supra note 67, at 1191-1201. 152. Bush v. Gore, 531 U.S. at 109. 153. See Bd. of Trs. v. Garrett, 531 U.S. 356 (2001); Kimel v. Fla. Bd. of Regents, 528 U.S. 62 (2000); Alden v. Maine, 527 U.S. 706 (1999). 2001] REGULATORY ROLE 659 ism. It is difficult, to say the least, to square the Court’s rulings in the Bush cases with this position. CONCEPTIONS AND MISCONCEPTIONS OF STATE CONSTITUTIONAL LAW IN BUSH V. GORE ROBERT A. SCHAPIRO* I. ARTICLE II AND STATE LAW................................................................................. A. Text and Context........................................................................................... B. The Blacker Precedent.................................................................................. C. The Framers’ View of State Constitutions and State Legislatures............. II. SUPREME COURT REVIEW OF STATE COURT DETERMINATIONS OF STATE LAW . A. The Concurrence’s Standard of Review: “Independent” and/or “Deferential” .................................................................................................. B. Precedents for Federal Review of State Court Determinations of State Law................................................................................................................ 1. Bouie and Fair Notice............................................................................ 2. General Common Law ........................................................................... 3. A Federal Common Law of Election Procedure .................................... III. THE FEDERAL INTEREST IN STATE PRESIDENTIAL ELECTION PROCEDURES ...... A. Federal Protection of State Legislatures...................................................... 1. State Constitutions and State Legislatures .......................................... 2. State Courts and State Legislatures ..................................................... B. Diversity in State Presidential Election Procedures ................................... C. Legislative Intent and the State Constitution ............................................. 1. Determining Legislative Intent.............................................................. 2. Contingency and Context ....................................................................... IV. IMPLICATIONS FOR STATE CONSTITUTIONAL LAW .............................................. 663 665 669 671 672 673 674 675 675 676 677 678 679 680 681 683 685 688 688 Although the United States Supreme Court’s disposition of Bush v. Gore (Bush II)1 ultimately turned on the Equal Protection Clause of the Fourteenth Amendment, an alternative theory, based on Article II of the United States Constitution, has garnered significant academic support in the year following the decision.2 The Article II theory was suggested in the initial per curiam opinion (Bush I)3 and in the questions of the Justices during the oral arguments, and was fully embraced by the concurrence in Bush II, which was written by Chief Justice Rehnquist and joined by Justices Scalia and Thomas. The reliance on Article II entailed a distinctive vision of the role of state constitutions in the state governmental process and of the place of federal courts in overseeing that role. The concurrence explicitly adopted the view that the Constitution provided federal constraints * Associate Professor of Law, Emory University School of Law. J.D., Yale Law School, 1990. I am grateful to Jim Rossi and the other organizers and participants in this symposium. William W. Buzbee, Joseph P. Helm, Michael L. Wells, and Robert F. Williams provided valuable advice and assistance. 1. Bush v. Gore, 531 U.S. 98 (2000). 2. See, e.g., RICHARD A. POSNER, BREAKING THE DEADLOCK: THE 2000 ELECTION, THE CONSTITUTION, AND THE COURTS (2001); Richard A. Epstein, “In such Manner as the Legislature Thereof May Direct”: The Outcome in Bush v. Gore Defended, 68 U. CHI. L. REV. 613 (2001). 3. Bush v. Palm Beach County Canvassing Bd., 531 U.S. 70 (2000) (per curiam). 661 662 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:661 on the authority of state courts to construe state procedures governing presidential elections.4 Under this conception, the concurrence insisted on a central role for federal courts in policing the relationship between state courts and state legislatures. The concurrence understood Article II both to grant plenary power to state legislatures and to mandate federal oversight of a state court’s interpretation of state law. Under this theory, the federal courts not only review whether state law, as interpreted by state courts, violates the Federal Constitution, but also review whether the state court correctly interpreted state law. The concurrence justified this extraordinary assertion of federal authority based on the need to protect the state legislature from the state courts. The concurrence further contended that uniformity in the construction of state election procedures was desirable and justified federalizing the interpretation of state law. Underlying the concurrence’s interpretation of Article II was a conception of the state constitutional system. The concurrence envisaged a unitary model of state constitutionalism that involved a strong dichotomy between state constitutions and other forms of state law. The concurrence apparently conceived of state constitutions and state statutes as relatively autonomous with little interpenetration. In the concurrence’s vision, state constitutions and state statutes may be divided in a fairly straightforward manner without any particular need to examine the context of a specific state’s constitutional system. The concurrence further assumed the predominance of uniform principles of the allocation of interpretive authority at the state level. At various points, the concurrence also evinced skepticism about the ability of state judges to reach fair and reasonable decisions and a corresponding confidence in the ability of federal courts to discern appropriate benchmarks against which to measure state judicial deviation. This Article argues that the flawed nature of the concurrence’s understanding of state constitutional systems fatally undermined its conclusions. The concurrence’s homogenizing conception failed to capture important features of state constitutions. An understanding of the role of state constitutions in the state law process requires an appreciation of the characteristics of a particular state’s constitution. The attempt to fit all state constitutions into a particular mold will necessarily fail, and the complexity of each state’s constitutional dynamic suggests that the United States Supreme Court should not attempt to create uniform rules of interpretation governing the role of state constitutions in presidential election disputes. The principle of separation of powers in state governments takes a variety of forms 4. Bush v. Gore, 531 U.S. at 111-22 (Rehnquist, C.J., concurring). 2001] STATE CONSTITUTIONAL LAW 663 and often differs substantially from the federal model. Without an understanding of a particular state’s system, it is impossible to comprehend the appropriate relationship between state courts and state legislatures. Moreover, the diversity of state statutory regimes belies the concurrence’s apparent confidence in the existence of a single, correct method of interpreting state statutes and its related assertion that it, rather than the Florida Supreme Court, best understood the Florida Election Code. Part I reviews the concurrence’s theory that Article II of the Federal Constitution grants special, plenary authority to the state legislature. I argue that this interpretation did not rest on firm foundations of text, precedent, or history. Rather, the concurrence’s conception could be justified, if at all, only by resort to unarticulated federal interests. Part II analyzes the concurrence’s related, but distinct conclusion that Article II mandates that federal courts independently review the correctness of state courts’ constructions of state laws governing presidential elections. As with the concurrence’s notion of plenary legislative power, I argue that the concurrence’s scrutiny of state courts’ interpretation of state law can be justified only by the existence of extraordinary federal interests, which remain unexplored in the opinion. Part III turns directly to an account of the federal interests at stake. This Part examines potential interests, such as the need to protect the state legislature from unprincipled judicial activism, the need for uniformity in the interpretation of presidential election codes, and the relative competence of the federal courts in interpreting state election laws. I argue that a proper understanding of state constitutional systems demonstrates the absence of any of these potential justifications for federal judicial intervention in this case. Part III further contends that the diversity of state constitutional contexts undermines the concurrence’s efforts to formulate uniform interpretive rules for state election codes. Part IV concludes with some more general reflections on the implications of the concurrence’s flawed conception of state constitutional law. I. ARTICLE II AND STATE LAW Article II provides that “[e]ach State shall appoint, in such Manner as the Legislature thereof may direct,” electors for President and Vice President.5 In Bush I, the United States Supreme Court suggested that this language did not merely specify that state law would govern the conduct of presidential elections but in fact constituted a special kind of delegation of authority to the state legislature.6 In 5. U.S. CONST. art. II, § 1, cl. 2. 6. The per curiam opinion stated: 664 FLORIDA STATE UNIVERSITY LAW REVIEW [Vol. 29:661 Bush II, the concurring opinion of Chief Justice Rehnquist, joined by Justices Scalia and Thomas, explicitly adopted this position. The concurrence asserted that Article II “leaves it to the legislature exclusively to define the method” of selecting presidential electors and that a “significant departure from the legislative scheme,” therefore, raises a federal constitutional question.7 Indeed, from the remainder of the opinion, it is clear that under the concurrence’s theory, a significant departure from the legislative scheme not only raises a constitutional question but actually violates the provisions of Article II. Accordingly, the concurrence undertook the task of determining whether the Florida Supreme Court’s interpretation of Florida law distorted the intent of the legislature, thus transgressing Article II. This section examines some of the bases for the theory that Article II confers plenary power on the state legislature. I explore in particular the justification for concluding that Article II frees the state legislature from the constraints that the state constitution otherwise would impose. The state constitution assumes particular importance in the discussion because, with the exception of the United States Constitution, the state constitution generally serves as the sole check on state legislative power. The primary target of a plenary power reading of Article II is the state constitution. As became apparent in Bush II, the concurrence’s theory targets the state judiciary as well. In defending its understanding of Article II, the concurrence relied primarily on the text of the constitutional provision and on language drawn from McPherson v. Blacker.8 As the dissenters pointed out, however, neither the word “legislature” nor the Blacker precedent can bear the weight that the concurrence places on it.9 A possible source for the interpretation of Article II on which the concur[I]n the case of a law enacted by a state legislature applicable not only to elections to state offices, but also to the selection of Presidential electors, the legislature is not acting solely under the authority given it by the people of th