Download THE NEW RACIAL JUSTICE: MOVING BEYOND THE EQUAL

THE NEW RACIAL JUSTICE:
MOVING BEYOND THE EQUAL PROTECTION CLAUSE TO
ACHIEVE EQUAL PROTECTION
EMILY CHIANG†
ABSTRACT
Since handing down Washington v. Davis and Arlington Heights v. Metropolitan Housing Development, the United States Supreme Court has significantly curtailed the ability of
plaintiffs to bring disparate impact claims under the Equal Protection Clause. Many academics continue to talk about the standards governing intent and disparate impact. Some
recent scholarship recognizes that reformers on the ground have shifted away from equalitybased claims altogether. This Article contends that civil rights advocates replaced the old
equal protection framework some time ago and that they did so deliberately and with great
success. It expands upon and refines the strategy shift some scholars have identified, with a
particular focus on racial inequality, the foundation on which equal protection rests. It does
so by focusing on three particularly timely reform movements: indigent defense reform, the
fight to end the school-to-prison pipeline, and challenges to immigration-related laws. The
Article uses these various reform movements to identify and analyze the true breadth of the
new racial justice reformers have wrought.
I. INTRODUCTION ..................................................................................................
II. THE DEATH OF THE EQUAL PROTECTION CLAUSE FOR CLAIMS PREMISED
ON RACIAL DISPARITY .......................................................................................
A. Disparate Impact After Davis .....................................................................
B. Moving Beyond the Equal Protection Clause .............................................
1. The Interplay of Liberty and Equality-Based Claims .........................
2. Moving Beyond Liberty/Equality to Achieve Equal Protection ...........
III. THE NEW RACIAL JUSTICE ................................................................................
A. Indigent Defense Reform, Other Constitutional Rights, and Purely
Disproportionate Impact .............................................................................
1. The Disparate Impact ..........................................................................
2. The Reform Response ...........................................................................
3. Beyond Public Defense Reform—Lessons for Other Claims ................
B. School-to-Prison Pipeline, Statutory Claims, and Unconscious Bias ........
1. The Disparate Impact ..........................................................................
2. The Reform Response ...........................................................................
(a) Claims Under the Equal Protection Clause .................................
(b) Other Constitutional Claims ........................................................
(c) The Statutory Alternative .............................................................
3. Beyond the School-to-Prison Pipeline—Lessons for Other Claims ......
C. Immigration, Structural Arguments, and Racial Animus .........................
1. The Disparate Impact ..........................................................................
2. The Reform Response ...........................................................................
(a) Facial Classification Based on Alienage ......................................
(b) Federal Preemption ......................................................................
(c) What Remains: Disparate Impact and a Return to the Bill of
Rights ............................................................................................
3. Beyond Immigration—Lessons for Other Claims ................................
836
838
839
841
844
847
850
851
852
853
857
860
861
864
864
866
867
871
872
874
875
875
876
877
879
† Associate Professor of Law, University of Utah S.J. Quinney College of Law. Many
thanks to Alan Chen, Cliff Rosky, Carolina Núñez, Jim Oleske, Justin Pidot, Rebecca
Scharf, and the Rocky Mountain Junior Scholars Forum. Thanks also to Stephen Dent and
Adrienne McKelvey for their invaluable research assistance.
836
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
IV. CONCLUSION .....................................................................................................
881
I. INTRODUCTION
Racial inequality endures in America. Whether the disparities—in
education,1 employment,2 incarceration,3 or any number of other areas4—constitute inequity may be a matter of debate, but the fact of
continued inequality is not.5 And yet, the Equal Protection Clause of
the Fourteenth Amendment is increasingly ill suited to address this
state of affairs. In the wake of decisions such as Washington v. Davis6
and Arlington Heights v. Metropolitan Housing Development Corp.,7
litigation seeking racial justice on the basis of disparate impact theories
of discrimination under the Constitution has been all but foreclosed.8
A number of commentators have noted that those seeking to address inequality have had to accommodate a Court and culture in1. See, e.g., infra Part III.B.1.
2. According to data provided by the Bureau of Labor Statistics, the unemployment
rate for African Americans in 2011 and 2012 was roughly double than what it was for
whites. Labor Force Statistics from the Current Population Survey, U.S. BUREAU OF LAB.
STAT., http://www.bls.gov/cps/cpsaat05.htm (last modified Feb. 5, 2013). As others have
noted, these statistics grossly underestimate the true African American unemployment
rate, because they do not include prisoners. BECKY PETTIT, INVISIBLE MEN: MASS
INCARCERATION AND THE MYTH OF BLACK PROGRESS 52-53 (2012); BRUCE WESTERN,
PUNISHMENT AND INEQUALITY IN AMERICA 97 (2006); see also MICHELLE ALEXANDER, THE
NEW JIM CROW: MASS INCARCERATION IN THE AGE OF COLORBLINDNESS 229 (The New
Press rev. ed. 2012) (noting that the true jobless rate among noncollege black men during
the 1990s was 42%, and 65% among black male dropouts). At the highest level, only thirteen African Americans have ever been chairperson or CEO of a Fortune 500 company. See
African American Chairmen & CEO’s of Fortune 500 Companies, BLACK PROFILES,
http://www.blackentrepreneurprofile.com/fortune-500-ceos/ (last updated Feb. 5, 2014).
3. In 2005, 2290 per 100,000 black people were incarcerated, compared to 412 white
people, a ratio of approximately 5.6. See George Coppolo & Kevin McCarthy, Crime Rate
and Conviction Rates Broken Down by Race, OLR RES. REP. (Jan. 18, 2008), available at
http://www.cga.ct.gov/2008/rpt/2008-R-0008.htm. African Americans make up 13% of the
general U.S. population and constitute: 28% of all arrests; 40% of all inmates held in prisons and jails; and 42% of the population on death row. See James E. Johnson et al.,
BRENNAN CTR. FOR JUST., N.Y. UNIV. SCH. OF L., RACIAL DISPARITIES IN FEDERAL
PROSECUTIONS 20 n.1 (2010), available at http://www.brennancenter.org/sites/default/files/
legacy/Justice/ProsecutorialDiscretion_report.pdf.
4. The U.S. Census Bureau reports that the 2011 median household income was
$55,412 for whites, $32,229 for blacks, and $38,624 for Hispanics. CARMEN DENAVAS-WALT
ET AL., U.S. CENSUS BUREAU, INCOME, POVERTY, AND HEALTH INSURANCE COVERAGE IN
THE UNITED STATES: 2011 8-9 (2012). The percentages of those below the poverty line were
9.8% for non-Hispanic whites, 27.6% for blacks, and 25.3% for Hispanics. Id. at 14.
5. See, e.g., Mario L. Barnes et al., A Post-race Equal Protection?, 98 GEO. L.J. 967,
982-89 (2010) (providing statistical evidence of racial disparity in poverty, income and
wealth, homeownership, employment, education, and the criminal justice system).
6. 426 U.S. 229 (1976).
7. 429 U.S. 252 (1977).
8. See infra Part II.A.
2014]
THE NEW RACIAL JUSTICE
837
creasingly discomfited by claims of group-based discrimination. This
unease has variously been described as “pluralism anxiety,”9 an “antibalkanization” perspective,10 and “the strain of difference.”11 Some
of these scholars have in turn noted a shift in the Court’s jurisprudence; in Kenji Yoshino’s words, a shift from “group-based equality
claims under the guarantees of the Fifth and Fourteenth Amendments to individual liberty claims under the due process guarantees
of the Fifth and Fourteenth Amendments.”12
The conversation about the future of equal protection and the
meaning of equality has never been more important. This Article
contends that the dialogue cannot be fully realized until we re-engage
with the paradigmatic application of equal protection principles to
race.13 It argues that racial justice advocates have engaged in a wholesale replacement of the equal protection framework over the last several decades, and that they have done so deliberately and with great
success. An understanding of the tools they used and the new racial
justice they have built is to the benefit of all who seek equality.
This Article identifies and explores three frontiers in the new racial justice: public defense, the school-to-prison pipeline, and immigration. Civil rights litigators have adapted to the Court’s restrictive
equal protection jurisprudence without ceding the battle for racial
equality in each of these areas, a fact made all the more remarkable
for the endemic disparate racial impact at the heart of these problems. Their approach to these issues serves to highlight the variety of
strategies at their disposal, many of which will be useful in other
contexts. To procure public defense reform, they have relied upon
other individual liberties protected by the Bill of Rights, the Sixth
Amendment in particular. To fight the school-to-prison pipeline, they
have used a rich federal statutory landscape to their advantage. And
to challenge restrictive immigration bills, they have mustered structural arguments, such as federal preemption.
These claims have played a critical role in reducing the disparate
racial impact of poor public defense systems, the school-to-prison
pipeline, and punitive immigration laws—and they have done so by
providing a true alternative to the Equal Protection Clause in the
9. Kenji Yoshino, The New Equal Protection, 124 HARV. L. REV. 747, 755 (2011).
10. Reva B. Siegel, From Colorblindness to Antibalkanization: An Emerging Ground
of Decision in Race Equality Cases, 120 YALE L.J. 1278, 1357 (2011).
11. Rebecca L. Brown, Liberty, the New Equality, 77 N.Y.U. L. REV. 1491, 1531 (2002).
12. Yoshino, supra note 9, at 748.
13. Cf. Kenji Yoshino, Covering, 111 YALE L.J. 769, 781 (2002) (noting that “racial
minorities and women have much to gain from a theory of discrimination that focuses on
the harms of coerced assimilation” for lesbians and gay men).
838
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
form of remedies that are structural, prospective, and class-based in
nature, not just individual and retrospective. Not only have these
claims proven more successful than traditional, explicitly race-based
claims, but they have not even been predicated upon racial inequality, much less inequity.14
Some scholars have questioned whether a race-neutral approach
to race-based problems is appropriate, criticizing it as a misguided
embrace of a “post-racial” society.15 This Article offers a response
grounded in realpolitik: litigators have not given up the fight for racial justice, but their pleadings now work around the doctrinal dead
end of explicitly race-based claims. Far from conceding the fight for
racial justice, these advocates have embraced a strategy that rewrites
the rules in their favor, resulting in real, measurable gains for equality.
Part II of this Article describes the death of the Equal Protection
Clause as a useful means of vindicating racial justice claims. It unpacks the commentary surrounding the cultural and doctrinal shifts
in the Court’s jurisprudence that have made equal protection claims
less appealing both to advocates and judges. It concludes that a return to the conversation about racial equality would enrich our understanding of equal protection for all groups and that this discussion
should begin with what advocates on the ground have been doing.
Part III of the Article identifies the primary ways in which civil
rights litigators have continued to make strides in combating racial
inequality without resorting to equal protection claims. It pairs areas
in which disproportionate racial impact is rampant with the most
meaningful alternatives to the Equal Protection Clause identified by
reformers thus far. In each area, it outlines the scope of the disparate
impact, the reform response, and the lessons each strategy has to
offer for others. This narrative provides the contours of the new racial justice, which has evolved to combat the racial inequalities of the
modern age.
II. THE DEATH OF THE EQUAL PROTECTION CLAUSE FOR CLAIMS
PREMISED ON RACIAL DISPARITY
The aftermath of Washington v. Davis, which all but foreclosed
claims of racial discrimination based upon a disparate impact theory,
has been amply documented.16 This Part will provide a brief overview
14. See infra Part II.B.
15. See, e.g., ALEXANDER, supra note 2, at 239; Kimberlé Williams Crenshaw, Twenty
Years of Critical Race Theory: Looking Back to Move Forward, 43 CONN. L. REV. 1253,
1327-40 (2011).
16. See infra Part II.A.
2014]
THE NEW RACIAL JUSTICE
839
of the doctrine as it stands today and summarize the academic commentary thus far as to the resultant state of civil rights work.
A. Disparate Impact After Davis
Any narrative of disparate impact claims must begin with Washington v. Davis17 and Arlington Heights v. Metropolitan Housing Development Corp.18 In Davis, the Court held that administration of a
test for authorities such as police officers did not violate equal protection despite the resultant exclusion of a disproportionate number of
African American applicants:
[W]e have not held that a law, neutral on its face and serving ends
otherwise within the power of government to pursue, is invalid
under the Equal Protection Clause simply because it may affect a
greater proportion of one race than of another. Disproportionate
impact is not irrelevant, but it is not the sole touchstone of an invidious racial discrimination forbidden by the Constitution. Standing alone, it does not trigger the rule that racial classifications are
to be subjected to the strictest scrutiny . . . .19
The Court concluded that absent a showing of discriminatory intent—that the state action was taken because of its disproportionate
racial impact rather than in spite of it—disparate impact is insufficient as grounds for relief.20 It noted that there is no requirement that
“the necessary discriminatory racial purpose must be express or appear on the face of the statute” and that such purpose could “often be
inferred from the totality of the relevant facts, including the fact, if it
is true, that the law bears more heavily on one race than another.”21
One term later in Arlington Heights, the Court explained that
plaintiffs must be able to offer “proof that a discriminatory purpose
has been a motivating factor in the decision” being challenged and
that absent a “stark” pattern of disparate impact, courts would have
to delve into factors such as the historical background of the decision,
the sequence of events leading to the decision, substantive and procedural departures from the normal operating procedure, and the
legislative and administrative history.22
Davis and Arlington Heights both dealt with state action in which
the state could plausibly contend the disparate impact was unfore17.
18.
19.
20.
21.
22.
426 U.S. 229 (1976).
429 U.S. 252 (1977).
Davis, 426 U.S. at 242 (internal citation omitted).
Id. at 239.
Id. at 241-42.
Arlington Heights, 429 U.S. at 265-67.
840
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
seeable, and thus in which evidence of discriminatory intent would
prove elusive. In Personnel Administrator of Massachusetts v. Feeney,23 the Court confronted state action that would inevitably and invariably lead to disparate impact, in the form of a veterans’ preference statute that “operates overwhelmingly to the advantage of
males.”24 The Court upheld the preference, noting that, “ ‘Discriminatory purpose’ . . . implies more than intent as volition or intent as
awareness of consequences.”25 Thus, although “it cannot seriously be
argued that the Legislature of Massachusetts could have been unaware that most veterans are men,” there was no equal protection
violation because there was no evidence of legislative intent to discriminate against women.26
The Court has evidenced somewhat more flexibility in its approach under two circumstances. First, the Court is more likely to
grant relief when the state action in question is more akin to state
inaction that perpetuates a pre-existing system of inequality. Thus,
where school districts pursue policies that result in continued school
segregation (as distinct from desegregation), the Court has indicated,
“actions having foreseeable and anticipated disparate impact are
relevant evidence to prove the ultimate fact, forbidden purpose.”27
Similarly, where a county seeks to maintain an at-large electoral
system in a district that dilutes the black vote, the Court has found
that “discriminatory intent need not be proved by direct evidence.”28
Second, the Court may grant relief when “the statistical disparities . . . warrant and require a conclusion that [is] irresistible, tantamount for all practical purposes to a mathematical demonstration
that the State acted with a discriminatory purpose.”29 To date, the
Court has only found statistical disparities that rose to this level
twice. In Yick Wo v. Hopkins, every single white applicant for a permit to operate a laundry in a wooden building was granted one, and
23. 442 U.S. 256 (1979).
24. Id. at 259.
25. Id. at 279.
26. Id. at 278.
27. Columbus Bd. of Educ. v. Penick, 443 U.S. 449, 464 (1979).
28. Rogers v. Lodge, 458 U.S. 613, 618 (1982); see also id. at 624 (noting district court
findings of past discrimination against African Americans, historical exclusion of African
Americans from the political process, and depressed socio-economic status of African Americans in the county).
29. McCleskey v. Kemp, 481 U.S. 279, 293-94 n.12 (1987) (internal citations and
quotation marks omitted); see also id. at 293 (noting that an exception to the general rule
that statistical evidence of disparate impact is insufficient may also exist in the “selection
of the jury venire in a particular district,” where a “stark pattern” of disparity may suffice
“as the sole proof of discriminatory intent” (quotation marks omitted)).
2014]
THE NEW RACIAL JUSTICE
841
not one of over two hundred Chinese applicants was granted one.30 In
Gomillion v. Lightfoot, the state had altered the boundaries of a
city from a square to “an uncouth twenty-eight-sided figure” that
excluded 395 of 400 black voters and not a single white voter.31 Both
of these cases pre-date Washington v. Davis; Yick Wo is more than a
century old.
The Court’s limited flexibility is thus of small comfort to plaintiffs
who wish to challenge new state action or those who cannot show the
stark statistical disparity the Court envisions. In the wake of the
Court’s jurisprudence, legislators have reacted predictably: they “do
not make a practice of justifying legislation on the grounds that it
will adversely affect groups that have historically been subject to
discrimination.”32 The Court has in turn acknowledged the difficulty
it has created, noting that “[p]roving the motivation behind official
action is often a problematic undertaking.”33
This combination of the Court’s jurisprudence and increased savviness on the part of state actors has proven nearly fatal for plaintiffs
seeking relief on the basis of racially disparate impact. The Court has
been unwilling to find discriminatory intent in claims as wideranging as those involving disparate application of the death penalty,34 juror selection,35 and a road closure that disproportionately affected African-American members of a community.36
B. Moving Beyond the Equal Protection Clause
This Article will not rehash the already familiar criticisms of the
Court’s various holdings in the area of disparate impact. Suffice it to
say that the shortcomings of a doctrine requiring proof of discriminatory intent are nearly self-evident.37 Nor will it delve into the myriad
30. 118 U.S. 356, 359 (1886).
31. 364 U.S. 339, 340-41 (1960).
32. Reva Siegel, Why Equal Protection No Longer Protects: The Evolving Forms of
Status-Enforcing State Action, 49 STAN. L. REV. 1111, 1135-36 (1997) (“[D]octrines of
heightened scrutiny have created incentives for legislators to explain their policy choices in
terms that cannot be so impugned.”).
33. Hunter v. Underwood, 471 U.S. 222, 228 (1985).
34. McCleskey, 481 U.S. at 292 (rejecting equal protection claim based on racially
disparate application of the death penalty).
35. See, e.g., Hernandez v. New York, 500 U.S. 352, 361 (1991) (deciding to remove
Spanish-speaking jurors on the grounds that they would not consider only the court interpreter’s version of testimony did not violate equal protection).
36. City of Memphis v. Greene, 451 U.S. 100, 119 (1981).
37. See, e.g., Siegel, supra note 32, at 1136 (“[T]he discriminatory purpose requirement now insulates many, if not most, forms of facially neutral state action from equal
protection challenge.”); see also Michelle Alexander, The New Jim Crow, 9 OHIO ST. J.
842
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
criticisms of the Court’s equal protection jurisprudence or suggestions for reform of the doctrine governing disparate impact.38
The dismay created by Davis and its progeny stems largely from
three observations about race in America, each of which is empirically verifiable: first, that African Americans are disproportionately
affected by laws that burden the poor or the socially disadvantaged,
because they are disproportionately poor and socially disadvantaged;39 second, that most of the racism that remains in America is of
the subconscious variety, as opposed to the explicit state-driven Jim
Crow variety;40 and third, that racial inequalities (in a purely numerCRIM. L. 7, 19 (2011) (“The Court has, as a practical matter, closed the door to claims of
racial bias in the criminal justice system. It has immunized the new caste system from
judicial scrutiny for racial bias, much as it once rallied to legitimate and protect slavery
and Jim Crow.”); Mario L. Barnes & Erwin Chemerinsky, The Once and Future Equal
Protection Doctrine?, 43 CONN. L. REV. 1059, 1066 (2011) (“[T]he Court has largely ignored
the history of the moment that produced the Reconstruction Amendments and created a
framework for equal protection analysis that all but ensures only a narrow group of discrimination claims will be actionable or succeed.”); Kenneth L. Karst, The Supreme Court,
1976 Term–Foreword: Equal Citizenship Under the Fourteenth Amendment, 91 HARV. L.
REV. 1, 51 (1977) (“A legislature oblivious to [the] existing stigma of caste will nonetheless
reinforce the stigma when it produces racially discriminatory effects through ostensibly
‘neutral’ legislation.”); Charles R. Lawrence III, The Id, the Ego, and Equal Protection:
Reckoning with Unconscious Racism, 39 STAN. L. REV. 317, 319 (1987) (describing criticisms of Davis as falling into two categories, those who object to the motive requirement
and those who argue that racial inequality exists regardless of motive).
38. See, e.g., Lawrence, supra note 37, at 324 (proposing new “test to trigger judicial
recognition of race-based behavior”); see also Barbara J. Flagg, “Was Blind, But Now I See”:
White Race Consciousness and the Requirement of Discriminatory Intent, 91 MICH. L. REV.
953, 960 (1993) (proposing requirement that government justify all facially neutral decision
making criteria that have disparate effects); Suzanne B. Goldberg, Equality Without Tiers,
77 S. CAL. L. REV. 481, 491 (2004) (proposing single standard of review for Equal Protection
claims); Karst, supra note 37, at 52 n.287 (proposing that government take into account
principle of equal citizenship and justify racially disproportionate results); Larry G. Simon,
Racially Prejudiced Governmental Actions: A Motivation Theory of the Constitutional Ban
Against Racial Discrimination, 15 SAN DIEGO L. REV. 1041, 1111 (1978) (arguing government should have to produce credible explanations when confronted with disparate impact); David A. Strauss, Discriminatory Intent and the Taming of Brown, 56 U. CHI. L. REV.
935, 956 (1989) (proposing rigorous application of intent requirement, as requiring government to act as if it does not know the race of those affected by the decision).
39. See, e.g., Michael J. Perry, The Disproportionate Impact Theory of Racial Discrimination, 125 U. PA. L. REV. 540, 557 (1977).
40. There is a whole body of literature, both in legal commentary and in scientific
studies, that identifies and explores the phenomenon of unconscious racism. See, e.g., Lawrence, supra note 37, at 336-44 (discussing unconscious racism and citing studies); see also
IAN AYRES, PERVASIVE PREJUDICE? UNCONVENTIONAL EVIDENCE OF RACE AND GENDER
DISCRIMINATION 19-44, 165-314 (2001) (gathering evidence of disparate treatment in the
retail sales of new cars, disparate impact in access to kidney transplantation, and the
setting of bail rates); Flagg, supra note 38, at 983-85 (gathering studies indicating unconscious racial bias in employment, mortgage lending, retail bargaining, psychiatric diagnoses, and other settings); Linda Hamilton Krieger, The Content of Our Categories: A Cognitive Bias Approach to Discrimination and Equal Employment Opportunity, 47 STAN. L.
REV. 1161, 1161 (1995) (discussing unconscious bias in the context of Title VII). A number
of scholars have concluded that the Court’s disregard for this research has resulted in a
2014]
THE NEW RACIAL JUSTICE
843
ical sense of the word “inequality”) persist in America regardless of
why anyone thinks this may be the case.41 Much has been written
about whether these inequalities constitute inequity, that is, whether
they reflect structural or institutional racism, but that conversation
is beyond the scope of this Article.42
Neither does this Article engage with the question of what should
replace the current antidiscrimination framework.43 Some have answered this question in a purely normative way, with various proposals for the Court to modify or discard pieces of its equal protection
jurisprudence.44 Others have answered this question more descriptively, advocating a particular approach (such as a move to libertydoctrine “that does not reflect prevailing understandings of the ways in which racial or
gender bias operates.” Siegel, supra note 32, at 1138; cf. Richard Delgado, Centennial Reflections on the California Law Review’s Scholarship on Race: The Structure of Civil Rights
Thought, 100 CAL. L. REV. 431, 440-41 (2012) (noting the difference between the paradigm
of racial liberalism, which believes that racism is not dead and that discrimination can
take many forms, and that of racial conservatism). Much has also been made of the Implicit Association Test, which purports to identify and quantify implicit bias, or unconscious
racism. See PROJECT IMPLICIT, https://implicit.harvard.edu/implicit/iatdetails.html (last
visited June 29, 2014). See, e.g., Anna Roberts, (Re)forming the Jury: Detection and Disinfection of Implicit Juror Bias, 44 CONN. L. REV. 827, 827 (2012) (discussing unconscious
racism in context of jury selection); see also Alexander R. Green et al., Implicit Bias Among
Physicians and Its Prediction of Thrombolysis Decisions for Black and White Patients, 22 J.
GEN. INTERNAL MED. 1231, 1235 (2007) (concluding that implicit bias in physicians influences treatment decisions); Anthony G. Greenwald & Linda Hamilton Krieger, Implicit
Bias: Scientific Foundations, 94 CAL. L. REV. 945, 966 (2006) (concluding that “implicit race
bias is pervasive and is associated with discrimination against African Americans” that
results in disparate impact); Jerry Kang et al., Implicit Bias in the Courtroom, 59 UCLA L.
Rev. 1124, 1126-27 (2012) (applying science of implicit bias to the courtroom). The premise
of this Article does not depend upon the existence or extent of unconscious racism; it merely depends upon the existence and extent of racial inequality.
41. See infra Parts III.A.1, III.B.1 & III.C.1.
42. See William M. Wiecek, Structural Racism and the Law in America Today: An
Introduction, 100 KY. L.J. 1, 13 (2012) (“Because of its invisibility, structural racism does
its work in the Potemkin village of ‘race-neutral’ policies.”); cf. KWAME TURE & CHARLES V.
HAMILTON, BLACK POWER: THE POLITICS OF LIBERATION IN AMERICA 4 (Vintage Books
1992) (1967) (“Racism is both overt and covert.”); Ian F. Haney López, Institutional Racism:
Judicial Conduct and a New Theory of Racial Discrimination, 109 YALE L.J. 1717, 1723
(2000) (building “a theory of racism that explains organizational activity that systematically harms minority groups even though the decision-making individuals lack any conscious
discriminatory intent”).
43. See Barnes & Chemerinsky, supra note 37, at 1084-85 (arguing Washington v.
Davis should be overturned).
44. See id. at 1075; see also John Hasnas, Equal Opportunity, Affirmative Action, and
the Anti-Discrimination Principle: The Philosophical Basis for the Legal Prohibition of
Discrimination, 71 FORDHAM L. REV. 423, 427 (2002) (examining the Equal Protection
Clause and Civil Rights Act of 1964 through the lens of morality); Charles R. Lawrence III,
Forbidden Conversations: On Race, Privacy, and Community (A Continuing Conversation
with John Ely on Racism and Democracy), 114 YALE L.J. 1353, 1382 (2005) (arguing that
the Equal Protection Clause reflects a commitment to a “new substantive value of ‘nonslavery’ and antisubordination to replace the old values of slavery and white supremacy”).
844
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
based claims) and supporting that position with the legal strategies
adopted in some cases.45
This Article’s contention is that reformers on the ground have already replaced the current antidiscrimination framework. They have
done so deliberately and steadily over the last several decades, and
they have done so with great success. It contends, moreover, that this
success is attributable to the fact that reformers no longer overtly
identify or describe their work as being explicitly concerned with antidiscrimination, or at least not in their legal pleadings. Reformers
have absorbed both the Court’s and the culture’s discomfort with explicitly race-based claims and have turned instead to combating racial
inequality using a different framework and vocabulary altogether.46
1. The Interplay of Liberty and Equality-Based Claims
The complementary nature of liberty and equality-based claims
has long been noted.47 Using Lawrence v. Texas as a focal point, Lawrence Tribe described the relationship between these two families of
claims with customary aplomb: he speaks of “a narrative in which
due process and equal protection, far from having separate missions
and entailing different inquiries, are profoundly interlocked in a legal
double helix. It is a single, unfolding tale of equal liberty and increasingly universal dignity.”48 Lawrence, in Tribe’s view, is a case that
“presupposed and advanced an explicitly equality-based and relationally situated theory of substantive liberty.”49
This Part identifies three overlapping strands in the recent scholarship on liberty- and equality-based claims. The first focuses on a
45. See infra Part II.B.1.
46. In one of the few empirical studies conducted on disparate impact racial discrimination claims, Theodore Eisenberg and Sheri Lynn Johnson express surprise at how
few such claims are filed. They state: “The Supreme Court’s standard takes its toll not
through an unusually high loss rate for those plaintiffs reaching trial or appeal, but by
deterring victims from even filing claims.” Theodore Eisenberg & Sheri Lynn Johnson, The
Effects of Intent: Do We Know How Legal Standards Work?, 76 CORNELL L. REV. 1151, 1153
(1991). Although this Article focuses largely upon systemic reform cases rather than individual claims, its contention is that rather than being deterred from filing claims at all,
victims of this type of racial discrimination have simply found alternatives to the Equal
Protection Clause.
47. See, e.g., William N. Eskridge, Jr., Destabilizing Due Process and Evolutive Equal
Protection, 47 UCLA L. REV. 1183, 1183 (2000) (“The Due Process Clause secures libertarian protections at the retail (individual) level that are important when the group is socially
despised, while the Equal Protection Clause potentially offers minorities wholesale (group)
level protection when (or if) the Court recognizes their legitimacy as partners in American
pluralist democracy.”).
48. Laurence H. Tribe, Lawrence v. Texas: The “Fundamental Right” That Dare Not
Speak Its Name, 117 HARV. L. REV. 1893, 1898 (2004).
49. Id.
2014]
THE NEW RACIAL JUSTICE
845
perceived shift in the Court’s jurisprudence driven by a desire to
avoid inter-group conflict in an increasingly diverse society. The second describes and advocates in favor of one type of claim as opposed
to the other. And the third, “realist” position, takes the position that
courts and reformers should simply use whichever claim fits best.
In the first group are Reva Siegel, Kenji Yoshino, and Rebecca
Brown. Each of these scholars provides a psychological profile of sorts
on the Court. Siegel identifies an antibalkanization perspective on
the Court, which is concerned about threats to social cohesion.50 She
contends that moderates on the Court who hold this perspective are
willing to:
allow government to engage in race-conscious efforts to integrate,
providing that government proceeds in ways that lower the salience of race in its interactions with the public. Antibalkanization . . . is distinctively concerned about the appearance of raceconscious interventions – the risk that race-conscious civil rights
interventions will heighten conflict or resentment.51
Where Siegel refers to an antibalkanization perspective, primarily
in the context of race-based decision making, Yoshino identifies “pluralism anxiety” as being responsible for a larger shift in the Court’s
jurisprudence “away from traditional group-based identity politics in
its equal protection and free exercise jurisprudence.”52 He explores a
more general narrative in which equality-based claims writ large (as
opposed to liberty-based claims) make the Court nervous, because
they force the Court to pick favorites among groups; “[l]iberty claims,
in contrast, emphasize what all Americans . . . have in common.”53
Brown also alludes to the problem of “representation of an increasingly heterogeneous population for which there can be no serious
contention that the interest of some is necessarily the interest of
all.”54 Instead of an antibalkanization perspective or pluralism anxiety, Brown discusses “the strain of difference:”
[T]he shared sense of values does not exist for everything that all
people value. Nor does it exist for the increasing, yet still small,
number who may wish to enjoy common freedoms, but in ways
50.
51.
52.
53.
54.
Siegel, supra note 10, at 1278.
Id. at 1357.
Yoshino, supra note 9, at 755.
Id. at 796.
Brown, supra note 11, at 1528.
846
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
that can be understood as distinguishable from the manner in
which the many enjoy them.55
Yoshino and Brown both also participate in the second strand of
the conversation, describing and advocating a shift towards libertybased claims. Yoshino observes that the Court has restricted equality-based claims but opened another avenue for relief, using “liberty
analysis to mitigate its curtailment of group-based equality analysis.”56 Brown also urges a new look at liberty claims, but her analysis
is grounded in an understanding of equality as having already been
largely attained: “As equality was to the last century, so should liberty be to the next. Equality jurisprudence, after all, has achieved the
stunning accomplishment of reconciling a robust judicial enforcement
with the demands of democratic constitutional theory.”57 As Brown
and others58 tell the story, equality claims have largely succeeded
while liberty claims have foundered.59
Others, such as William Eskridge, would respond that equalitybased claims fill a role that liberty-based claims cannot, because they
can provide relief for an entire group at a time:
[R]egular equal protection and due process scrutiny might be either interchangeable or interdependent at the retail level, that is,
in challenges to particular discriminations, especially penaltybased ones. But the Equal Protection Clause alone offers a minority group a potential constitutional jackpot at the wholesale level,
that is, in challenges to an array of interconnected discriminations
in state benefits as well as burdens. 60
Finally, some scholars take the approach closest to the hearts of
litigation-minded reformers: the best type of claim is that which fits
your agenda. Richard Delgado, for example, states:
[A]ttention to human needs, problems, deprivation, and flourishing
may proceed under one of two banners, individual rights or equal
55. Id. at 1531; see also Mary D. Fan, Post-Racial Proxies: Resurgent State and Local
Anti-“Alien” Laws and Unity-Rebuilding Frames for Antidiscrimination Values, 32
CARDOZO L. REV. 905, 908-09 (2011) (“Equality norms can be framed and vindicated in a
more palatable, legally tenable form, and tied to other interests to appeal more widely and
ameliorate estrangement in a polarized polity.”).
56. Yoshino, supra note 9, at 776.
57. Brown, supra note 11, at 1492.
58. See, e.g., Carlos A. Ball, Why Liberty Judicial Review Is as Legitimate as Equality
Review: The Case of Gay Rights Jurisprudence, 14 U. PA. J. CONST. L. 1, 3-4 (2011) (identifying the roots of the liberty/equality binary as a “legitimacy disparity” between the two,
wherein judicial review founded on equality principles is lauded, as in the case of Brown v.
Board of Education, 347 U.S. 483 (1954), but review founded on liberty principles is suspect, as in the case of Roe v. Wade, 410 U.S. 113 (1973)).
59. Brown, supra note 11, at 1494.
60. Eskridge, supra note 47, at 1216.
2014]
THE NEW RACIAL JUSTICE
847
protection. Both approaches aim at the same goal, both result in
heightened judicial scrutiny, and the choice to proceed under one
banner or the other is largely a matter of tactics, ideological commitment, or perceived public sentiment.61
Each of these scholars has contributed invaluable and groundbreaking insight on the dynamic interplay between liberty and equality claims, and each offers a unique and valuable perspective on recent developments in that dynamic. Siegel provides a pinpoint identification of the problem that reformers concerned with racial equality
confront, not just before the Supreme Court, but also in lower courts
and in the court of public opinion: race continues to matter, but state
actors (like school administrators) must pretend it does not. Yoshino
and Brown provide a compelling description of the shift in the Court’s
jurisprudence away from equality-based claims. Delgado’s work resonates with reformers on the ground who will gladly adopt any claim
that serves their purposes. And Eskridge offers us a powerful reminder that structural problems demand structural solutions—
solutions that are wholesale, not retail.
To this important body of literature, this Article offers several
insights. First, critical work remains to be done, even when it comes
to the most basic of equalities. Second, there is a rich world beyond
the binary of liberty- and equality-based claims to help tackle that
work. And third, reformers on the ground are already making use of
a multitude of strategies to address racial disparity.
2. Moving Beyond Liberty/Equality to Achieve Equal Protection
This Article urges a return to the conversation about claims for
racial equality, which remain salient in the national discourse but
often are no longer identified as such by their proponents. The inclination to move beyond race is visible on several fronts: in society, in
the Court’s decisions, and within the academy. Thus, the debate over
affirmative action is at once avoidant of race and yet indelibly
stamped by it, taking place against the backdrop of a Court and society that increasingly seems to believe—rightly or wrongly—that
“[t]he way to stop discrimination on the basis of race is to stop discriminating on the basis of race.”62 Most contemporary academic
61. Delgado, supra note 40, at 450.
62. Parents Involved in Cmty. Schs. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 748
(2007); c.f. Barnes & Chemerinsky, supra note 37, at 1064 (“Substantive racial equality
clearly mattered to the Court once.”).
848
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
commentary on equal protection has, in turn, moved away from race
and towards gay rights and same-sex marriage in particular.63
This Article contends that the discussion about equal protection is
a richer one when informed by the traditional, racial underpinnings
of the Equal Protection Clause. We must be careful, both as a society
and an academy, not to be too quick to leave race behind. From a
purely empirical perspective, race indisputably still matters.64 But
even more importantly, the scope of protection afforded under the
Equal Protection Clause should matter not just to the old groups already under its ambit, but also to new groups seeking to join. Membership in the club might be less valuable if the benefits are not as
extensive as one had assumed.
A return to the academic and social conversation about racial
equality is not the same as a return to explicitly race-based claims in
litigation; indeed, the primary descriptive insight this Article contributes is that such claims are largely doomed to fail.65 But this refusal to revisit a doctrinal dead-end from a litigation standpoint is
not an endorsement of a “post-racial” society.66 Kimberlé Williams
63. See, e.g., Katie R. Eyer, Marriage This Term: On Liberty and the “New Equal
Protection”, 60 UCLA L. REV. DISCOURSE 2, 6 (2012) (arguing the LGB movement should
continue to root its claims in equality, as opposed to liberty); Yoshino, supra note 9, at 778
(centering his claim that equality-based claims are giving way to liberty-based claims by
citing Lawrence v. Texas, 539 U.S. 558, 574 (2003)). Not coincidentally, gay marriage is on
the Court’s mind as well. See United States v. Windsor, 133 S. Ct. 2675 (2013) (striking
down the Defense of Marriage Act); see also Hollingsworth v. Perry, 133 S. Ct. 2652 (2013)
(holding proponents of California’s Proposition 8 did not have standing to appeal district
court’s order finding the proposition unconstitutional).
64. See supra notes 39-42 and accompanying text. Racial inequality is unlikely to
disappear anytime soon, and the sordid history of race in America will likely continue to
manifest itself in differentiation between legislation with racial impact and legislation with
impact on minorities of other types. See Robert M. Cover, The Origins of Judicial Activism
in the Protection of Minorities, 91 YALE L.J. 1287, 1303 (1982) (“In contrast to the deep
societal roots of governmental action against Blacks—the close fit between private terror,
public discrimination, and political exclusion, directed against Blacks for a century—action
against other minorities has usually been sporadic, transitory, and local.”); see also id. at
1308 (“[T]he apparently neutral structural characteristics of the Constitution had never
been neutral concerning race.”). Further, although the Court’s holdings have adversely
affected the ability of other groups to bring disparate impact claims, such as those based
upon gender inequalities, nowhere has the effect been greater than on claims involving
racial disparity.
65. Cf. Barnes et al., supra note 5, at 1000 (urging a change in “the way in which
equality advocates frame the discussion” to “shift the underlying premise of equality arguments from a compensatory to a distributive justice rationale”); Fan, supra note 55, at 909
(arguing that “alternate approaches [to equality-based claims] must enfold antidiscrimination concerns and norms into the analysis rather than altogether elide address of the concerns”). This Article contends that successful reformers have avoided this strategy because
it would defeat the purpose. They have, in other words, already reframed the discussion,
and the new frame has nothing whatsoever to do with racial equality.
66. Cf. Crenshaw, supra note 15, at 1313-46 (describing and critiquing the phenomenon of post-racialism).
2014]
THE NEW RACIAL JUSTICE
849
Crenshaw offers a powerful critique of the strategies described herein
in a variation of the adage: “The master’s tools will never dismantle
the master’s house.”67 Crenshaw criticizes not just the concept of
post-racialism, the idea that race no longer matters in our society,
but also the “post-racial pragmatism . . . that urges scaling racial
obstacles while declining to name them.”68 She argues that such
pragmatism entraps racial justice advocates and constituencies; silences racial justice advocacy; threatens to make the racial disparities described in this Article “unremarkable features of the postracial world;” and renders civil rights advocates “yesterday’s news—
irrelevant, delusional and unsophisticated.”69
In a similar vein, Michelle Alexander thoroughly documents the
racial disparity in America’s mass incarceration, which she describes
as “the New Jim Crow.”70 Alexander criticizes the reform response to
this phenomenon on several fronts: for being insufficient in scale and
scope; focusing overly on litigation instead of grassroots reform; and
considering race-neutral grounds for reform.71 Specifically, Alexander
argues, “The prevailing caste system cannot be successfully dismantled with a purely race-neutral approach.”72 She notes that “opportunities for challenging mass incarceration on purely race-neutral
grounds have never been greater,” but urges racial justice advocates
not to take the “tempting bait.”73
This Article’s response to these powerful and persuasive criticisms
is grounded more in realpolitik than ideology. While Alexander is
undoubtedly correct that racial inequality will not be eradicated with
any single lawsuit or legal strategy, litigation will continue to play an
important role and litigators must take notice of doctrinal reality.
Rather than cede the notion of the “reigned in” and narrowed “field of
contestation” that Crenshaw depicts, reformers have broadened the
legal grounds on which to fight.74 They would surely agree that “there
67. AUDRE LORDE, SISTER OUTSIDER 112 (rev. ed. 2007).
68. Crenshaw, supra note 15, at 1332.
69. Id. at 1327-1340. Reformers may find themselves stuck between a rock and a hard
place in the academic commentary. Richard Thompson Ford criticizes those who perhaps
belong to an older school for which Crenshaw may feel nostalgia. See RICHARD THOMPSON
FORD, RIGHTS GONE WRONG: HOW LAW CORRUPTS THE STRUGGLE FOR EQUALITY 24-25, 27
(2011) (“Civil rights have effectively ‘occupied the field’ of social justice, crowding out alternative ways of thinking and new solutions. . . . The civil rights tradition encourages us to
look at disputes through a lens that is designed to focus on discrimination.”).
70. ALEXANDER, supra note 2.
71. Id. at 225-39.
72. Id. at 239.
73. Id.
74. Crenshaw, supra note 15, at 1343.
850
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
are limits to the degree that racial justice can be finessed,”75 but the
fight for racial justice is not over, and the advocates who continue to
wage the fight are far from irrelevant, delusional, or unsophisticated
in their tactics. For example, they are engaged in active battles over
the school-to-prison pipeline and the disproportionate impact of the
mortgage foreclosure crisis—two of the areas of racial disparity
Crenshaw worries will become “unremarkable features of the postracial world.”76
Civil rights advocates do not have the luxury of ignoring equal
protection doctrine post-Washington v. Davis, which has essentially
reified the concept of post-racialism in the courtroom; but neither
have they walked away from the challenge.77
III. THE NEW RACIAL JUSTICE
This Part begins the descriptive project of identifying and analyzing the ways in which reformers have begun to combat racial inequality without resort to the Equal Protection Clause.78 I have identified
three main areas in which they have done so, each of which serves to
illuminate two larger groups of claims, one having to do with the nature of the right being vindicated and the other having to do with the
nature of the disproportionate impact. The claims range from those
rooted in the specific enumerations of the Bill of Rights to those that
stem from the federalist structure of our legal system. The nature of
the disproportionate impact addressed by these claims is similarly
wide-ranging, from that which is caused with no explicit racial ani75. Id. at 1346.
76. Id. at 1337-40.
77. 426 U.S. 229 (1976). The civil rights movement has a long history of pragmatism,
dating back to Plessy v. Ferguson, which was brought as a test case by reformers who presented the Court with a plaintiff who was seven-eighths white. See Plessy v. Ferguson, 163
U.S. 537 (1896); see also, e.g., Miss. Univ. for Women v. Hogan, 458 U.S. 718 (1982) (bringing gender discrimination claim on behalf of men); Frontiero v. Richardson, 411 U.S. 677,
681-82 (1973) (same).
78. Others have written about circumventing the intent requirement to vindicate
racial inequality, most notably in the area of environmental law. See, e.g., Julie H. Hurwitz
& E. Quita Sullivan, Using Civil Rights Laws to Challenge Environmental Racism: From
Bean to Guardians to Chester to Sandoval, 2 J.L. SOC’Y 5, 9-10 (2001); Suzanne Smith,
Current Treatment of Environmental Justice Claims: Plaintiffs Face a Dead End in the
Courtroom, 12 B.U. PUB. INT. L.J. 223, 249-50 (2002) (discussing administrative proceedings and private rights of action under section 602 of Title VI as possible alternatives to
using the Equal Protection Clause to seek redress); Sandra L. Geiger, An Alternative Legal
Tool for Pursuing Environmental Justice: The Takings Clause, 31 COLUM. J.L. & SOC.
PROBS. 201, 204 (1998); see also, e.g., Andrea Brenneke, Civil Rights Remedies for Battered
Women: Axiomatic & Ignored, 11 LAW & INEQ. 1, 4 (1992) (discussing statutory alternative
to equal protection challenges on behalf of battered women); Angela J. Davis, Prosecution
and Race: The Power and Privilege of Discretion, 67 FORDHAM L. REV. 13, 18 (1998) (discussing equal protection challenges to racially biased decisions to prosecute).
2014]
THE NEW RACIAL JUSTICE
851
mus, to that which nears explicit racial bias, but falls short of the
Supreme Court’s definition of discriminatory intent.
The first group of claims in Part III.A, indigent defense reform,
represents not only those claims that implicate other explicitly enumerated constitutional rights, such as ones involving the criminal
justice system and criminal procedure, but also claims in which any
other enumerated constitutional right (other than equal protection) is
at stake. These cases also serve to inform a discussion of disproportionate impact that stems purely from the economics of poverty, with
no evidence of explicit animus.
The second constellation of claims in Part III.B, having to do with
efforts to redress the school-to-prison pipeline, stand in for claims
implicating statutory as opposed to constitutional rights, and claims
in which the source of the disproportionate impact is perhaps less
clear. These are claims for which the Equal Protection Clause might
have held some promise in the absence of a discriminatory intent
requirement, that is, where there is ample statistical evidence of racial disparity in treatment and perhaps some evidence of intent, but
not enough to meet the requirements laid out in Arlington Heights.
Finally, Part III.C turns to immigration-related claims, challenges
to state and local regulations directed at undocumented workers, or
so-called “illegal immigrants.” This group of claims represents structural interests beyond enumerated individual rights, such as the
separation of powers. This group of claims also involves disproportionate impact in which racial animus likely plays a role but falls
short of the explicit discriminatory intent the Court demands.
A. Indigent Defense Reform, Other Constitutional Rights, and Purely
Disproportionate Impact
The problems that plague the state public-defense systems in this
country are not new; much has been written to catalog them and this
Article will take them as a given.79 This Part uses public defense reform as a lens through which to view a particular type of racial justice claim: those that implicate criminal procedure, or more broadly,
79. See, e.g., NORMAN LEFSTEIN, ABA STANDING COMM. ON LEGAL AID & INDIGENT
DEFENDANTS, SECURING REASONABLE CASELOADS: ETHICS AND LAW IN PUBLIC DEFENSE
12-24 (2011), available at www.americanbar.org/content/dam/aba/publications/books/ls_scl
aid_def_securing_reasonable_caseloads_supplement.pdf; NAT’L RIGHT TO COUNSEL COMM.,
JUSTICE DENIED: AMERICA’S CONTINUING NEGLECT OF OUR CONSTITUTIONAL RIGHT TO
COUNSEL 49-101 (2009), available at http://www.constitutionproject.org/pdf/139.pdf; Emily
Chiang, Indigent Defense Invigorated: A Uniform Standard for Adjudicating PreConviction Sixth Amendment Claims, 19 TEMP. POL. & CIV. RTS. L. REV. 443, 447-50 &
nn.20-39 (2010).
852
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
those for which a constitutional guarantee other than equal protection applies. This analysis is also intended to shed light on the fight
against other similar problems that attend our criminal justice system as a whole.
1. The Disparate Impact
Although problems with state public defense systems do not solely
affect people of color—and in some jurisdictions may not even primarily affect people of color—the effects of poorly functioning systems across the U.S. are disproportionately borne by people of
color. First, problems with public defense systems are by definition
borne by the poor, and greater percentages of African Americans
and Hispanics live below the poverty line than do whites.80 Second,
all aspects of the criminal justice system disproportionately affect
people of color and African-American men in particular.81 Finally,
statistical evidence indicates that a higher percentage of incarcerated
African Americans and Hispanics had publicly appointed counsel
than whites.82
This Part will assume that there is generally no evidence of racial
animus in state and local government decisions to underfund or
otherwise neglect their public defense systems, and certainly no
evidence of discriminatory intent that rises to the level required
by Arlington Heights. In other words, traditional equal protection
80. DENAVAS-WALT ET AL., supra note 4, at 8-9. In 2011, the percentages of those
below the poverty line were 9.8% for non-Hispanic whites, 27.6% for blacks, and 25.3% for
Hispanics. Id. at 14.
81. See Dorothy E. Roberts, The Social and Moral Cost of Mass Incarceration in African American Communities, 56 STAN. L. REV. 1271, 1272-73 (2004) (discussing the “community-level harms” that flow from “grossly disproportionate” rates of incarceration of
African-American men); Michael Tonry & Matthew Melewski, The Malign Effects of Drug
and Crime Control Policies on Black Americans, 37 CRIME & JUST. 1, 7 (2008) (surveying
the empirical literature on the disproportionate burdens facing people of color in arrests,
conviction, and sentencing, concluding that “political and ideological exigencies of the last
quarter century have conduced to the adoption of crime control policies of unprecedented
severity, the primary burdens of which have been borne by disadvantaged blacks (and,
increasingly, Hispanics)”); Katherine J. Rosich, Race, Ethnicity, and the Criminal Justice
System, AM. SOC. ASS’N 2-3 (Sept. 2007), http://www.asanet.org/images/press/docs/pdf/
ASARaceCrime.pdf (surveying social science research on race and crime and identifying
numerous areas of racial disparity); see also Alexander, supra note 37, at 18-19 (“Law
enforcement officials are largely free to discriminate on the basis of race today, so long as
no one admits it. That’s the key.”).
82. “While 69% of white state prison inmates reported they had lawyers appointed by
the court, 77% of blacks and 73% of Hispanics had publicly financed attorneys. In federal
prison black inmates were more likely than whites and Hispanics to have public counsel:
65% for blacks, 57% for whites and 56% for Hispanics.” Indigent Defense Systems, BUREAU
OF JUST. STAT., http://bjs.ojp.usdoj.gov/index.cfm?ty=tp&tid=28#defendants (last visited
June 29, 2014); see also CAROLINE WOLF HARLOW, U.S. DEP’T OF JUST., DEFENSE COUNSEL
IN CRIMINAL CASES 9 (2000), available at http://bjs.ojp.usdoj.gov/content/pub/pdf/dccc.pdf.
2014]
THE NEW RACIAL JUSTICE
853
claims to challenge this particular type of state action (or inaction)
are foreclosed.
2. The Reform Response
Despite the lack of availability of traditional equal protection
claims, indigent defense reform has been a veritable hotbed of
litigation and other activity over the last decade, and a number of the
lawsuits have been filed by organizations explicitly dedicated to racial justice.83 For many, the connection between the pursuit of indigent defense reform and racial justice is explicit: “[T]he dispropor83. See, e.g., Duncan v. State, 784 N.W.2d 51, 53 (Mich. 2010), vacated, 790 N.W.2d
695 (2010) (alleging indigent criminal defendants are being denied their right to counsel
and “effective assistance of counsel,” filed by the ACLU’s Racial Justice Project); White v.
Martz, No. CDV-2002-133, 2006 Mont. Dist. LEXIS 136, at *1 (Mont. Dist. Ct. Jan. 25,
2006) (alleging public defender programs in Montana counties lack the resources to provide
statutorily and constitutionally adequate representation of indigent clients, filed by the
ACLU’s Racial Justice Project); Hurrell-Harring v. State, 930 N.E.2d 217, 232 (N.Y. 2010)
(alleging a claim for ineffective counsel of indigent criminal defendants, filed by the New
York Civil Liberties Union); Brief for NAACP Legal Defense and Educational Fund, Inc. as
Amici Curiae Supporting Appellants, Richmond v. Dist. Court of Md., No. 24-C-06-009911
CN, 2007 WL 5446238 (Cir. Crt. Md. Dec. 7, 2007); see also SARAH GERAGHTY & MIRIAM
GOHARA, ASSEMBLY LINE JUSTICE: MISSISSIPPI’S INDIGENT DEFENSE CRISIS 6-8 (2003),
available at http://www.americanbar.org/content/dam/aba/migrated/legalservices/download
s/sclaid/indigentdefense/ms_assemblylinejustice.authcheckdam.pdf. The Brennan Center
for Justice at NYU School of Law has also identified indigent defense reform as a focus
within the area of “racial justice.” Racial Justice, BRENNAN CTR. FOR JUST.,
http://www.brennancenter.org/content/section/category/racial_justice/ (last visited June 29,
2014); see also Brief for Brennan Center et al. as Amici Curiae Supporting Appellees, DeWolfe v. Richmond, 76 A.3d 1019 (Md. 2013) (No. 24-C-06-009911 CN), 2011 WL 4585691;
Brief for National Association of Criminal Defense Lawyers et al. as Amici Curiae Supporting Appellees, DeWolfe v. Richmond, 76 A.3d 1019 (Md. 2013) (No. 24-C-06-009911 CN),
2011 WL 4585691; Brief for Brennan Center et al. as Amici Curiae Supporting Appellees,
Duncan v. State, 784 N.W.2d 51 (Mich. 2010) (No. 07-242-CZ), 2011 WL 4585691; Brief for
National Association of Criminal Defense Lawyers et al. as Amici Curiae Supporting Appellees, Duncan v. State, 784 N.W.2d 51 (Mich. 2010) (No. 07-242-CZ), 2011 WL 4585691;
Brief of Former Prosecutors Michael A. Battle et al. as Amici Curiae Supporting Appellants, Hurrell-Harring v. State, 930 N.E.2d 217 (N.Y. 2010) (No. 2010-0066), 2010 WL
1775135. In a 2008 report, the Brennan Center discussed guidelines to determine eligibility for publicly appointed counsel. DAVID UDELL ET AL., BRENNAN CTR. FOR JUST., ELIGIBLE
FOR JUSTICE: GUIDELINES FOR APPOINTING DEFENSE COUNSEL 5-26 (2008), available at
http://www.brennancenter.org/sites/default/files/legacy/publications/Eligibility.Report.pdf; see
also THOMAS GIOVANNI & ROOPAL PATEL, BRENNAN CTR. FOR JUST., GIDEON AT 50: THREE
REFORMS TO REVIVE THE RIGHT TO COUNSEL 8-9 (2013), available at
http://www.brennancenter.org/sites/default/files/publications/Gideon_Report_040913.pdf. The
Charles Hamilton Houston Institute for Race & Justice at Harvard Law School has a project entitled “MyGideon,” designed to provide resources to indigent and capital defense
attorneys. MyGideon, THE CHARLES HAMILTON HOUS. INST. FOR RACE & JUST., HARV. L.
SCH. (July 26, 2012), http://www.charleshamiltonhouston.org/portfolio/my-gideon/; see also
Catherine V. Beane, Indigent Defense: Separate and Unequal, CHAMPION, May 2004, at 54,
55 (“NACDL’s Indigent Defense Committee welcomes your suggestions on ways that we
can better address racial justice issues and the disproportionate impact that inadequate
indigent defense has on communities of color.”).
854
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
tionate presence of racial minorities among the indigents relying
on public defender services reinforces the need to ensure that all defendants receive competent representation, else the criminal justice
system will create further disparities in treatment of persons of
different races.”84
And yet no mention of racial justice is made in the reform work
itself; the legal documents filed rely almost exclusively upon the
guarantees of the Sixth Amendment. This Part will focus upon
some of the most recent developments in this area and highlight the
successes reformers have had in a variety of cases with different
procedural postures.85
First, advocates have continued to file classic class action suits
that seek wholesale reform at the state or county level. Litigation
of this type builds upon successes like the statewide public defense
system implemented by Montana in response to an ACLU lawsuit.86
These claims typically request injunctive and declaratory relief on
behalf of a class of indigent criminal defendants, alleging that the
public defense systems in question are inadequately resourced
and supervised.87
Litigation seeking state-wide reform was most recently successful
in Michigan, where a suit filed by the ACLU in 2006 resulted in passage of a bill that created a public defense commission tasked with
collecting data and implementing standards.88 Similar litigation is
84. Charles J. Ogletree, Jr., An Essay on the New Public Defender for the 21st Century, 58 LAW & CONTEMP. PROBS. 81, 93 (1995).
85. The increase in litigation and other reform activity in this area has been so vast
that this Article will not even attempt a comprehensive review. There have been a number
of non-litigation successes as well, however, and they too have not resorted to explicit
claims of racial injustice. See, e.g., Order, In the Matter of the Review of Issues Concerning
Representation of Indigent Defendants in Criminal and Juvenile Delinquency Cases (Nev.
2008) (ADKT No. 411) (implementing comprehensive performance standards); H.R. 483,
51st Leg., 1st Sess. (N.M. 2013) (creating independent public defense commission in New
Mexico); H.R. 147, 62d Leg., 1st Reg. Sess. (Idaho 2013) (establishing presumptive indigency guidelines in Idaho); H.R. 148, 62d Leg., 1st Reg. Sess. (Idaho 2013) (prohibiting attorneys from serving as both lawyer and guardian ad litem for children); H.R. 149, 62d Leg.,
1st Reg. Sess. (Idaho 2013) (preventing juveniles from making uninformed waivers of counsel); Texas Fair Defense Act, S.B. 7, 77th Leg., Reg. Sess. (Tex. 2001), 2001 Tex. Gen. Laws
1800-01 (codified as amended at TEX. CODE. CRIM. PROC. ANN. art. 1.051 (West 2004))
(providing for state funding, requiring minimum standards in counties, guaranteeing
access to necessary and sufficient support services for attorneys, and creating centralized
data collection system in Texas).
86. White, 2006 Mont. Dist. LEXIS 136, at *5.
87. Amended Complaint at 5, White v. Martz, No. CDV-2002-133, 2006 Mont. Dist.
LEXIS 136 (Mont. Dist. Ct. Apr. 1, 2002).
88. Michigan Indigent Defense Commission Act, Act No. 93, 2013 Mich. Pub. Acts 53.
The author worked on this litigation while at the Brennan Center for Justice, and subsequently at the ACLU.
2014]
THE NEW RACIAL JUSTICE
855
still pending in New York.89 Advocates filed suit in Georgia90 and entered into a consent decree that revised the procedures for responding to requests for conflict-free appellate counsel, provided for additional full-time staff attorneys and workload controls for the appellate division of the public defender office, monitoring of contract attorneys working for the appellate division, and data collection.91
County or municipality-based lawsuits are pending in Washington,92
Texas,93 Georgia,94 and Pennsylvania.95 Not one of these lawsuits includes a federal equal protection claim.
Second, some public defenders have successfully brought suits
themselves, seeking to limit their own caseloads. These cases have a
more mixed record than the prototypical class actions described
above and some have yet to play out fully, but they remain a type of
claim tethered to the Sixth Amendment that reformers may consider.96 The Florida Supreme Court recently found in favor of the MiamiDade County Public Defender’s Office’s right to move to decline appointment in future cases due to excessive caseloads.97 Similarly, a
suit filed by the Public Defender of Mojave County, Arizona, resulted
in a ruling that permitted the office to withdraw from thirty-nine
cases, and warned the county that future motions to withdraw would
also be granted “until the court is convinced that the reasons for do-
89. Hurrell-Harring v. New York, 914 N.Y.S.2d 367, 372 (N.Y. App. Div. Jan. 6 2011)
(reversing lower court’s denial of class certification and granting class certification).
90. Petition for Writ of Mandamus and Verified Complaint for Injunctive and Declaratory Relief, Flournoy v. Georgia, No. 2009CV178947 (Ga. Super. Ct. Dec. 15, 2009), 2009
WL 8728189 (seeking relief on behalf of class represented by state public defenders office’s
appellate division).
91. Consent Decree, Flournoy v. Georgia, No. 2009CV178947 (Ga. Super. Ct. Mar. 12,
2012), 2012 WL 5885196.
92. Order Denying Defendants’ Motions for Summary Judgment and Plaintiffs’ Motion for Preliminary Injunction, Wilbur v. City of Mt. Vernon, No. C11-1100RSL (W.D.
Wash. Dec. 4, 2013), 2012 WL 600727; see also Best v. Grant Cty., No. 04-2- 00189-0 (Super. Ct. Wash. Aug. 26, 2004) .
93. Heckman v. Williamson Cty., No. 10-0671, slip op. at 2-3 (Tex. June 8, 2012)
(seeking relief on behalf of class facing misdemeanor charges).
94. Verified Complaint at 4, 14, Cantwell v. Crawford, No. 09EV275M (Ga. Super. Ct.
Apr. 7, 2009), 2009 WL 1043789 (seeking relief on behalf of class charged with felony offenses in five counties in Georgia).
95. Class Action Complaint, Flora v. Luzerne Cty., No. 2072 CD 2013 (Pa. Commw.
Ct. Apr. 10, 2012).
96. For a general discussion of this type of litigation and for specific information about
these cases, see LEFSTEIN, supra note 79, at 161-90.
97. Pub. Defender, Eleventh Jud. Cir. of Fla. v. State, 115 So. 3d 261, 274 (Fla.
2013) (noting the court is “reaffirm[ing] that aggregate/systemic motions to withdraw are
appropriate in circumstances where there is an office-wide or wide-spread problem as to
effective representation”).
856
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
ing so no longer exist.”98 But relief was denied to defenders seeking
relief in New Orleans99 and Knoxville.100 And although the Missouri
Public Defender Commission fought successfully for the right to limit
the caseloads of its attorneys,101 a bill introduced in the 2013 legislative session sought to resolve the caseload problem by privatizing
services for all non-serious felony cases through low-bid contracts.102
Claims brought by public defenders based explicitly upon a disparate impact theory under the Equal Protection Clause have not
met with success at all, however. In Idaho, for example, contract attorneys filed suit against a county for breach of contract when the
county sought to terminate their contract. The attorneys included an
equal protection claim, alleging that the county’s attempts to jettison
its contract with them in favor of a low-bid contract would have a
disproportionate impact on Idaho’s racial minorities because 20% of
Idaho’s racial minorities and 25% of Idaho’s total Hispanic population
reside in the county in question.103 Their claim was denied.104
Finally, advocates of improved public defense services have had
notable success before the Supreme Court. The Court has continually
expanded entitlement to counsel under the Sixth Amendment, requiring in recent years, for example, the appointment of counsel at
pretrial interrogations105 and in misdemeanor cases where there is a
substantial likelihood of incarceration.106 The Court has also recently
held that failure to communicate a plea offer to a defendant constitutes deficient performance by counsel under the Sixth Amendment
and that defendants may be prejudiced by counsel’s deficient performance in recommending that a plea offer be rejected.107 Although
none of these cases are systemic reform cases (each is a claim for
post-conviction relief on behalf of a particular defendant); the last
two cases, Missouri v. Frye and Lafler v. Cooper, are not public defense cases at all. Each will have a salutary effect on racial justice, as
98. Arizona v. Lopez, No. CR-2007-1544, slip op. at 9-10 (Mohave Cnty. Sup. Ct.
Dec. 17, 2007).
99. Loisiana v. Edwards, No. 2007-K-639 (La. Ct. App. Aug. 16, 2007).
100. Petition for Writ of Certiorari, In re Pet. of Knox Cnty. Pub. Defender, No.
174552-C (Tenn. Knox Cnty. Ch. Ct. Mar. 9, 2009).
101. State ex rel. Mo. Pub. Defender Comm’n v. Pratte, 298 S.W.3d 870, 887 (2009).
102. H.B. 215, 97th Gen. Assemb., 1st Reg. Sess. (Mo. 2013).
103. Complaint for Declaratory and Injunctive Relief at *4, Wiebe & Fouser, P.A. v.
Idaho, No. CV09-4413-C, 2009 WL 3072841 (Dist. Ct. Apr. 24, 2009).
104. Wiebe & Fouser, P.A. v. Idaho, No. CV-2009-4413-C, 2009 WL 2980444, at *1
(Dist. Ct. July 31, 2009).
105. Kansas v. Ventris, 556 U.S. 586, 590 (2009).
106. Alabama v. Shelton, 535 U.S. 654, 658 (2002).
107. Lafler v. Cooper, 132 S. Ct. 1376, 1391 (2012); Missouri v. Frye, 132 S. Ct. 1399,
1408 (2012).
2014]
THE NEW RACIAL JUSTICE
857
reflected by a number of the amicus briefs submitted by various public interest organizations.108 Again, however, nowhere is the Equal
Protection Clause mentioned in the actual legal claims.
3. Beyond Public Defense Reform—Lessons for Other Claims
Each of the success stories in public defense reform has made a
real difference for the millions of people of color caught up in the
criminal justice system, with nary a mention of racial justice or equal
protection. Of the claims described in this Article, these are perhaps
closest in temperament to the “liberty-based” claims Yoshino describes and prescribes, as they seek to vindicate a right (to counsel)
functionally denied to some by relying on the universality of the
right’s guarantee to all.
The Bill of Rights has held great promise for other areas in which
notable racial disparity exists. For example, capital defense reform, a
close cousin of public defense reform, has also benefited enormously
from the tactics described above.109 The disparate racial impact of the
death penalty is clear. The current death row population is 41% black
and 43% white.110 Of the defendants executed in the United States
since 1976, 35% were black and 56% were white.111 And infamously,
the racial disparity when it comes to victims of crimes for which the
death penalty was inflicted is even starker: 15% of the victims were
black, compared to 77% who were white.112
Despite the racial disparity, perhaps nowhere are claims based on
disparate impact under the Equal Protection Clause more plainly
foreclosed. The Supreme Court dealt explicitly with the issue in
108. See, e.g., Brief for the Constitution Project as Amicus Curiae in Support of Respondents at 19, Lafler v. Cooper, 132 S. Ct. 1376 (2012) (No. 10-209), Missouri v. Frye,
132 S. Ct. 1399 (2012) (No. 10-444) (arguing the following as its second ground on appeal:
“Because Indigent Defendants Plead Guilty At An Even Higher Rate Than Others, Petitioners’ Proposed Standard For Prejudice Would Deprive Most Indigent Defendants Of Full
Protection Under The Sixth Amendment”); Brief of the National Association of Criminal
Defense Lawyers et al. as Amici Curiae in Support of Respondents, Lafler v. Cooper, 132 S.
Ct. 1376 (2012) (No. 10-209), Missouri v. Frye, 132 S. Ct. 1399 (2012) (No. 10-444).
109. Capital defense reform is understood to encompass an array of projects, including
the following: elimination of the death penalty, whether wholesale or piecemeal;
improvements in legal representation for those facing the death penalty, including the
implementation of standards governing who may represent those facing capital charges;
and direct representation.
110. Race of Death Row Inmates Executed Since 1976, DEATH PENALTY INFO. CTR.,
http://www.deathpenaltyinfo.org/race-death-row-inmates-executed-1976#defend (last visited June 29, 2014).
111. Id.
112. See id. Similarly, 19 white defendants were executed for murdering one or more
black victims, while 257 black defendants were executed for murdering one or more
white victims. Id.
858
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
McCleskey v. Kemp, finding that statistical evidence that the death
penalty was imposed far more frequently on black defendants who
killed white victims than on white defendants who killed black victims, even after for controlling for a number of other factors, was
insufficient to demonstrate a violation of the Equal Protection
Clause.113 The Court held that in order for McCleskey to prevail on
his claim, he “would have to prove that the Georgia Legislature enacted or maintained the death penalty statute because of an anticipated racially discriminatory effect.”114 Such evidence, of course,
would be virtually impossible to come by.115
McCleskey also raised an Eighth Amendment issue. But in contrast to the use of rights other than those associated with equal protection that this Article advocates, his claim was doomed because it
was premised on the same racial disparity he sought to address with
the equal protection claim: he argued that the statistical evidence of
racial disparity in the imposition of the death penalty made the
death penalty “cruel and unusual” as applied to him.116 The successful strategies this Article explores are based instead on assertions of
rights wholly independent of racial disparity challenges.
Death penalty abolitionists and racial justice advocates have succeeded in continually chipping away at the death penalty—and its
attendant racial disparities—through the Eighth Amendment. In
Atkins v. Virginia,117 the Court agreed that imposing the death penalty upon the intellectually disabled constitutes “cruel and unusual
punishment” for Eighth Amendment purposes.118 The Court premised
its holding on “the relative culpability of mentally retarded offenders,
and the relationship between mental retardation and the penological
purposes served by the death penalty” and noted also that “some
characteristics of mental retardation undermine the strength of the
procedural protections that our capital jurisprudence steadfastly
guards.”119 According to the Death Penalty Information Center, the
United States executed forty-four defendants with intellectual disa-
113. 481 U.S. 279, 297-99 (1987).
114. Id. at 298.
115. Cf. United States v. Armstrong, 517 U.S. 456, 458 (1996) (denying request for
discovery in selective prosecution claim based on disparate racial impact between prosecutions for crack cocaine and powder cocaine because respondents “failed to show that the
Government declined to prosecute similarly situated suspects of other races”).
116. McCleskey, 481 U.S. 297.
117. 536 U.S. 304 (2002).
118. Id. at 321.
119. Id. at 317.
2014]
THE NEW RACIAL JUSTICE
859
bilities between 1984 and 2002.120 Thirty-two percent of those were
white, 62% were black.121
Similarly, in Roper v. Simmons,122 the Court held that the Eighth
Amendment forbids “imposition of the death penalty on offenders
who were under the age of 18 when their crimes were committed.”123
In so holding, the Court noted the diminished culpability of juveniles
and stated “retribution is not proportional if the law’s most severe
penalty is imposed on one whose culpability or blameworthiness is
diminished, to a substantial degree, by reason of youth and immaturity.”124 According to the Death Penalty Information Center, as of the
end of 2004 (when Simmons was decided) there were seventy-one
people on death row for crimes committed as juveniles.125 Sixty-four
percent were people of color;126 41% were black, and 34% were
white.127 And, in a set of statistics that would be familiar to McCleskey, 71% of the victims whose race was known were white and 28%
were black.128
Reformers seeking to address racial disparity in public defense—
and the criminal justice system more generally—have found success
pursuing claims under other provisions of the Bill of Rights, as long
as those claims are founded in the “liberty” strand of doctrine described above, premised on rights held by all rather than the denial
of a right to some.129 Those who have made the greatest strides in
120. List of Defendants with Mental Retardation Executed in the United States, DEATH
PENALTY INFO. CTR., http://www.deathpenaltyinfo.org/list-defendants-mental-retardationexecuted-united-states (last visited June 29, 2012) (citing Denis Keyes et al., People with
Mental Retardation Are Dying, Legally: At Least 44 Have Been Executed, 40 MENTAL
RETARDATION 243, 243-44 (2002)).
121. Id.
122. 543 U.S. 551 (2005).
123. Id. at 578.
124. Id. at 571.
125. Juvenile Offenders Who Were on Death Row, DEATH PENALTY INFO. CTR.,
http://www.deathpenaltyinfo.org/juvenile-offenders-who-were-death-row#streiboverview (last
visited June 29, 2012) (citing VICTOR L. STREIB, THE JUVENILE DEATH PENALTY TODAY: DEATH
SENTENCES AND EXECUTIONS FOR JUVENILE CRIMES, JANUARY, 1, 1973 - APRIL 31, 2004 (2004),
available at http://www.internationaljusticeproject.org/pdfs/JuvDeathApril2004.pdf).
126. Id.
127. Id.
128. Id.
129. This reform work does not even include the on-going effort on behalf of the wrongfully convicted to pursue exonerations largely on the basis of DNA evidence via state statutory claims and federal and state habeas petitions. Poor lawyering at the trial level is
frequently implicated. The Innocence Project, perhaps the most well-known of the organizations pursuing exonerations on behalf of the wrongfully convicted, explicitly identifies
“bad lawyering” as one of the six potential causes of wrongful convictions, The Causes of
Wrongful Conviction, INNOCENCE PROJECT, http://www.innocenceproject.org/understand/
860
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
ensuring the provision of adequate counsel to the people of color most
gravely affected by shortcomings in our indigent defense system and
who have succeeded in protecting the most vulnerable from a death
penalty system riddled with disparate impact, have done so without
reference to race. There is every reason to believe that the reformers
tackling the next frontiers in this type of claim, such as the problem
of mass incarceration in our society, can and will do the same.130
B. School-to-Prison Pipeline, Statutory Claims, and Unconscious
Bias
The “school-to-prison-pipeline”131 is a term used by advocates,
scholars, and reformers to describe the phenomenon by which children are funneled out of the educational system and into the criminal
(last visited June 29, 2014), and notes, “Improving resources for public defense and ensuring the proper training and oversight of all defense lawyers can prevent wrongful convictions.” William Fleener, Staff Att’y, Cooley Innocence Project, Testimony to the Indigent
Defense Advisory Commission, MICH. CAMPAIGN FOR JUST. (Dec. 16, 2011),
http://www.michigancampaignforjustice.org/docs/Fleener%20Testimony%20to%20the%20Indi
gent%20Defense%20Advisory%20Commission%20final.doc; see also SAMUEL R. GROSS &
MICHAEL SHAFFER, NAT’L REGISTRY OF EXONERATIONS, EXONERATIONS IN THE UNITED
STATES, 1989–2012 42 (2012), available at http://www.law.umich.edu/special/exoneration/
Documents/exonerations_us_1989_2012_full_report.pdf (“For 104 exonerations, our information includes clear evidence of severely inadequate legal defense, but we believe that
many more of the exonerated defendants – perhaps a clear majority – would not have been
convicted in the first instance if their lawyers had done good work.”).
And the racial disparity being addressed is undeniable. According to the National
Registry of Exonerations, 1040 people had been exonerated as of the date of this Article.
About the Registry, NAT’L REGISTRY OF EXONERATIONS, http://www.law.umich.edu/special/
exoneration/Pages/about.aspx (last visited June 29, 2014). Of 802 crimes for which the race
of the defendant was known, 50% of the exonerees were black and 38% were white. See
GROSS, supra note 129, at 31 (noting “[i]t’s no surprise that black defendants are heavily
overrepresented among exonerees: they are heavily overrepresented among those arrested
and imprisoned for violent crimes and drug crimes. But the disproportions we see are
greater than what one would expect.”). The racial disparities were even greater for sexual
assault (63% black versus 32% white), attempted murder (59% black versus 12% white),
robbery (64% black versus 18% white), and drug crimes (60% black versus 10% white). See id.
130. This is not to say that claims based upon the Equal Protection Clause or disparate
impact are dead altogether. The recent challenges to the New York City Police Department’s “stop and frisk” policies and practices, for example, combine classic liberty claims
under the Fourth Amendment with equal protection (and an array of common law tort)
claims. Complaint at 4, 48-49, Ligon v. City of N.Y., 925 F. Supp. 2d 478 (S.D.N.Y. 2013)
(No. 12 Civ. 2274 (SAS)), 2012 WL 1031760; Complaint at 2, 40, 42, 50, Davis v. City of
New York, 902 F. Supp. 2d 405 (S.D.N.Y. 2012) (No. 10 Civ. 0699 (SAS)), 2010 WL
9937605; Second Amended Class Action Complaint for Declaratory and Injunctive Relief
and Individual Damages at 2, 34, 36, 42, Floyd v. City of New York, 283 F.R.D. 153
(S.D.N.Y. 2012) (No. 08 Civ. 01034 (SAS)); see also Opinion and Order at 6, Floyd v. City of
New York, 283 F.R.D. 153 (S.D.N.Y. 2012) (No. 08 Civ. 01034 (SAS)) (noting that 2.8 million people were stopped between 2004 and 2009 and that over 52% of those stops were of
blacks, 31% were of Latinos, and 10% were of whites).
131. This phenomenon is also sometimes referred to as “schoolhouse to the jailhouse.”
2014]
THE NEW RACIAL JUSTICE
861
justice system.132 The funneling effect of the pipeline can take place
at any number of junctures within the education system and via any
number of different administrative and educational policies and practices.133 Examples of the pipeline at work include overuse of the
school disciplinary system (sometimes through zero tolerance policies) that results in students being kept out of school,134 and referrals
to the criminal justice system for infractions traditionally handled
by schools.135
1. The Disparate Impact
The statistical evidence of the disparate impact the phenomenon
has on children of color is wide-ranging.136 In 2009-2010, the national
graduation rate for black male students was 52% and for Latino
males it was 58%; in contrast, white males graduated at a rate of
78%.137 A recent survey conducted by the Department of Education of
72,000 schools (covering 85% of students nationwide) found that
black students constitute 18% of the student body population, but
35% of the students suspended at least once, 46% of those suspended
more than once, and 39% of those expelled.138 In fact, black students
132. See, e.g., ADVANCEMENT
TO JAILHOUSE TRACK 11 (2005).
PROJECT, EDUCATION ON LOCKDOWN: THE SCHOOLHOUSE
133. See, e.g., Emily Chiang, No State Actor Left Behind: Rethinking Section 1983
Liability in the Context of Disciplinary Alternative Schools and Beyond, 60 BUFF. L. REV.
615, 621-22 & nn.23-29 (2012) (describing facets of the pipeline and providing citations);
Dean Hill Rivkin, Legal Advocacy and Education Reform: Litigating School Exclusion, 75
TENN. L. REV. 265, 268 (2008) (discussing schools’ uses of the juvenile justice system as a
disciplinary tool).
134. ADVANCEMENT PROJECT, supra note 132, at 7; THE C.R. PROJECT AT HARV.
UNIV. & ADVANCEMENT PROJECT, OPPORTUNITIES SUSPENDED: THE DEVASTATING
CONSEQUENCES OF ZERO TOLERANCE AND SCHOOL DISCIPLINE 1 (2000) [hereinafter HARV.
UNIV. C.R. PROJECT].
135. HARV. UNIV. C.R. PROJECT, supra note 134, at 15; CATHERINE Y. KIM ET AL., THE
SCHOOL-TO-PRISON PIPELINE: STRUCTURING LEGAL REFORM 3, 113-14 (2010).
136. See DANIEL J. LOSEN & JONATHAN GILLESPIE, THE C.R. PROJECT, OPPORTUNITIES
SUSPENDED: THE DISPARATE IMPACT OF DISCIPLINARY EXCLUSION FROM SCHOOL 6-7 (2012),
available at http://civilrightsproject.ucla.edu/resources/projects/center-for-civil-rights-rem
edies/school-to-prison-folder/federal-reports/upcoming-ccrr-research/losen-gillespie-opportu
nity-suspended-2012.pdf.
137. THE SCHOTT FOUND. FOR PUB. EDUC., THE URGENCY OF NOW: THE SCHOTT 50
STATE REPORT ON PUBLIC EDUCATION AND BLACK MALES 7 (2012), available at
http://blackboysreport.org/urgency-of-now.pdf.
138. ED Data Express: Data About Elementary & Secondary Schools in the U.S., DEP’T
OF EDUC., http://eddataexpress.ed.gov/state-tables-main.cfm (last visited June 29, 2014)
(select “All States” under section one; then select “Achievement Data” under section three;
then select “Graduation Rate Data”; then select “Display Report” at the bottom of the
page); Michael Harris, New National Data Shows Racial Disparities in School Discipline,
NAT’L CTR. FOR YOUTH L., http://www.youthlaw.org/publications/yln/2012/apr_jun_2012/
new_national_data_shows_racial_disparities_in_school_discipline/#sdfootnote1sym (last visit-
862
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
were more than three and a half times as likely to be suspended or
expelled as white students.139 More than 70% of students arrested in
school or referred to law enforcement were Hispanic or black.140 Studies also confirm that these disparities cannot be explained by worse
behavior or socioeconomic status.141
These disparities in school discipline rates have consequences beyond the disparity in graduation rates. In an analysis of the data
released by the Department of Education, the Center for American
Progress found:
Students who were suspended or expelled for even one discretionary violation in Texas were 2.85 times more likely than their peers
to be in contact with the juvenile justice system within the following year. Each subsequent violation exponentially increased [a]
student’s chances of juvenile justice involvement—nearly half (46
percent) of students with at least 11 disciplinary actions came into
contact with the juvenile justice system, compared to only 2.4 percent of students with no disciplinary violations.142
In 2009, the arrest rate for juveniles aged 10-17 per 100,000 people was 4,644.3 for whites but 10,096.3 for blacks.143 The Department
of Justice Office of Juvenile Justice and Delinquency Prevention
notes also that “between 1980 and 2010, the total juvenile arrest rate
decreased 54% for Asians, 51% for American Indians, and 30% for
whites, while the overall rate for black juveniles increased 8% during
this period.”144 Black juveniles constituted 16% of this age group but
51% of arrests for violent crimes and 33% of arrests for property
crimes.145 After they are arrested, black juveniles represent 31% of
ed June 29, 2014). The New York Times analysis of the data noted that in districts with
expulsions under zero tolerance policies, Hispanic and black students represented 45% of
students but 56% of those expelled. Tamar Lewin, Black Students Punished More, Data
Suggests, N.Y. TIMES, Mar. 6, 2012, at A11.
139. Harris, supra note 138.
140. Lewin, supra note 138.
141. See LOSEN & GILLESPIE, supra note 136, at 32.
142. Rachel Wilf, Disparities in School Discipline Move Students of Color Toward Prison, CTR. FOR AM. PROGRESS (Mar. 13, 2012), http://www.americanprogress.org/issues/race/
news/2012/03/13/11350/disparities-in-school-discipline-move-students-of-color-toward-prison/.
143. Juvenile Arrest Rate Trends, OFF. OF JUV. JUST. & DELINQ. PREVENTION,
http://ojjdp.gov/ojstatbb/crime/JAR_Display.asp?ID=qa05260&text=yes (last visited June
29, 2014).
144. Id.
145. CHARLES PUZZANCHERA & BENJAMIN ADAMS, OFF. OF JUV. JUST. & DELINQ.
PREVENTION, JUVENILE ARRESTS 2009 (2011), available at http://www.ojjdp.gov/pubs/
236477.pdf.
2014]
THE NEW RACIAL JUSTICE
863
referrals to juvenile court and 41% of waivers to adult court.146
In 2008, 37.2% of black men with less than a high school education
were incarcerated.147
As with public defense systems, it is difficult, if not impossible, to
demonstrate that educators and school administrators are intentionally discriminating against children of color.148 Some scholars have
advocated for education reform via the few remaining traditional
desegregation cases stemming from Brown and its progeny.149 While
this strategy would indeed provide plaintiffs seeking to shut down
some aspects of the school-to-prison pipeline with some measure of
relief under the Equal Protection Clause, these cases are exceptions
that prove the rule: few districts remain under desegregation orders
and their issuance depended upon the very discriminatory intent so
difficult to demonstrate today. As a result, reformers have evidenced
little desire to return to the old racial justice paradigm of equal protection litigation to combat the pipeline.
146. THE SENT’G PROJECT, REDUCING RACIAL DISPARITY IN THE CRIMINAL JUSTICE
SYSTEM 2 (2000), available at http://www.sentencingproject.org/doc/publications/rd_reduc
ingracialdisparity.pdf.
147. PETTIT, supra note 2, at 15.
148. See, e.g., Chauncee D. Smith, Note, Deconstructing the Pipeline: Evaluating
School-to-Prison Pipeline Equal Protection Cases Through a Structural Racism Framework,
36 FORDHAM URB. L.J. 1009, 1026-28 (2009). Smith notes:
[H]istorical and present-day actions that contribute to the pipeline can be categorized into three dimensions—criminalization, sorting, and economic policy.
Together, these dimensions form a structural racism framework that largely
encompasses the dynamic nature of disparate minority student pushout and incarceration. Thus, in contrast to a motive-centered approach, evaluating the
pipeline’s criminalization, sorting and economic dimensions reveals how fragmented inequities have a drastically unequal cumulative impact on students
of color.
Id. at 1026-27; see also Russell J. Skiba et al., African American Disproportionality in
School Discipline: The Divide Between Best Evidence and Legal Remedy, 54 N.Y. L. SCH. L.
REV. 1071, 1074 (2009/10) (“A similar analysis in the area of racial disparities in discipline
shows a distinct gap between the scientific knowledge base regarding racial disparities in
discipline and the absence of a legal strategy accepted by the courts to address such disparities. Analysis of case law reveals that this gap appears to be related to the court’s adherence to a colorblind interpretation of the Constitution.”).
149. See, e.g., Danielle Holley-Walker, A New Era for Desegregation, 28 GA. ST. U. L.
REV. 423, 426-27 (2012).
864
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
2. The Reform Response150
(a) Claims Under the Equal Protection Clause
There is surprisingly little difference in the outcomes of cases
challenging racial disparities under the Equal Protection Clause in
the imposition of school discipline before and after Washington v.
Davis: claims that succeed involve either (1) a flat-out admission of
racial discrimination or (2) evidence that white students were either
not disciplined at all for the same infractions or disciplined less
harshly. Neither of these elements is easy to come by and claims
based purely upon disparate impact have never had great success.
Both before and after Davis, courts have been satisfied by open
admissions of racial discrimination. In Hawkins v. Coleman, a 1974
case, the Northern District of Texas upheld a claim based upon statistical evidence demonstrating black students were disciplined more
frequently than white students.151 But the school district had been
segregated until three years before the court’s opinion was issued,152
and the superintendent of the school district testified in the case that
the high number of suspensions of black students was because “ ‘we
are a White controlled institution, institutional racism, [and] racism
among individuals.’ ”153 Similarly, in 1985, in Sherpell v. Humnoke
School District No. 5 of Lonoke County, Arkansas,154 the Eastern District of Arkansas ruled in favor of plaintiffs where there was evidence
teachers referred to black students as “niggers,” “blue-gums,” and
“coon.”155 In Mayorga Santamaria v. Dallas Independent School District,156 the Northern District of Texas found in favor of plaintiffs’
150. This Part provides only a brief overview of the reform response to the school-toprison pipeline problem to illustrate the direction reformers have taken away from claims
under the Equal Protection Clause and towards other alternatives. It also omits discussion
of claims for individual relief, e.g. claims brought on behalf of a single child to petition for
services under a statute or to challenge an individual instance of school discipline. For a
much more comprehensive discussion of the reform response, see generally KIM ET AL.,
supra note 135; Frances P. Solari & Julienne E.M. Balshaw, Outlawed and Exiled: Zero
Tolerance and Second Generation Race Discrimination in Public Schools, 29 N.C. CENT.
L.J. 147 (2007).
151. 376 F. Supp. 1330 (N.D. Tex. 1974).
152. Id. at 1331.
153. Id. at 1336.
154. 619 F. Supp. 670 (E.D. Ark. 1985).
155. Id. at 673. A teacher in Sherpell also testified she “personally witnessed the discipline of a black child by an administrative official which resulted in broken skin and blood;
that during [her] nine-year tenure, she had not witnessed any white child subjected to such
treatment.” Id. at 674; cf. Coleman v. Franklin Parish Sch. Bd., 702 F.2d 74, 77 (5th Cir.
1983) (permitting equal protection claim to proceed because “plaintiffs pleaded intent and
purpose to discriminate on the part of the defendants”).
156. No. Civ.A.3:06CV692-L, 2006 WL 3550194 (N.D. Tex. Nov. 16, 2006).
2014]
THE NEW RACIAL JUSTICE
865
equal protection claim where there was evidence that school administrators had intentionally and explicitly segregated students by race
and national origin.157
Both before and after Davis, courts have required plaintiffs lacking evidence of intentional discrimination to demonstrate not just
statistical disparity in the imposition of discipline, but evidence that
white students were not similarly disciplined for similar infractions.158 And even where some evidence is offered that white students
were not similarly disciplined, courts have been reluctant to grant
relief unless the students are nearly identically situated.159 These
claims are most likely to succeed when there is evidence that two
students involved in a fight were treated differently. For example, in
Payne v. Worthington Schools,160 the (black) plaintiff was given a oneday, in-school suspension and the school merely called the other
(white) child’s parents.161 Similarly, in Antoine v. Winner School District,162 plaintiffs were able to secure a consent decree in part
because they offered not only statistical evidence of disparate discipline but also evidence that Native American students were disci-
157. Id. at *32-39; see also People Who Care v. Rockford Bd. of Ed. Sch. Dist. #205, 851
F. Supp. 905, 933 (N.D. Ill. 1994) (upholding equal protection claim where school administrators intentionally segregated students).
158. See, e.g., Tasby v. Estes, 643 F.2d 1103, 1108 (5th Cir. 1981) (“[A]bsent a showing
of arbitrary disciplinary practices, undeserved or unreasonable punishment of black students, or failure to discipline white students for similar misconduct, the plaintiffs have not
satisfied their burden of proving that the disproportionate punishment of black students in
the [school district] is the product of a racially discriminatory purpose.”); Sweet v. Childs,
507 F.2d 675, 681 (5th Cir. 1975) (“There was no showing of arbitrary suspensions or expulsions of black students nor of a failure to suspend or expel white students for similar
conduct.”); Fuller v. Decatur Pub. Sch. Bd. of Ed. Sch. Dist. 61, 78 F. Supp. 2d 812, 815
(C.D. Ill. 2000) ([“Plaintiffs’] statistics failed to establish that any similarly situated Caucasian students were treated less harshly.”), aff’d on other grounds, 251 F.3d 662 (7th Cir.
2001); Collins v. Chichester Sch. Dist., No. CIV.A. 96-6039, 1998 WL 351718, at *6 (E.D.
Penn. June 29, 1998); Parker v. Trinity High Sch., 823 F. Supp. 511, 520 (N.D. Ill. 1993)
(“Plaintiffs must show that those who determined the punishment improperly considered
plaintiffs’ race. Stray remarks by nondecisionmakers or remarks unrelated to the disciplinary decision process do not satisfy this burden.”).
159. See, e.g., Tasby, 643 F.2d at 1107 n.1 (noting that “the statistics offered are based
upon a breakdown of offenses far too general to prove disproportionate severity in punishment . . . [and] do not reflect other relevant circumstances surrounding each individual
case of punishment . . . [such as] prior offenses”); Parker, 823 F. Supp. at 520 (refusing to
grant relief in part because evidence that white students were not disciplined for fighting
did “not represent a similar, repeated disregard for the authority of the teachers”).
160. Payne v. Worthington Sch., No. C2-99-830, 2001 WL 506509 (S.D. Ohio Apr. 25, 2001).
161. Id. at *8.
162. Consent Decree, Antoine v. Winner Sch. Dist. 59-2, No. Civ. 06-3007 (D.S.D.
Dec. 10, 2007).
866
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
plined more harshly than Caucasian students for participating in the
same fights.163
(b) Other Constitutional Claims
Absent the presence of these particularly egregious indicators of
racial discrimination, reformers have had to rely upon other claims
to combat the school-to-prison pipeline and its effects on children of
color. This Part will focus primarily on the use of statutory claims as
an alternative to the Equal Protection Clause, but as with indigent
defense reform, claims under other constitutional guarantees are
also effective.
Thus, in Antoine v. Winner School District, plaintiffs sought relief
under the Fifth Amendment and successfully obtained a settlement
agreement where school administrators routinely obtained confessions from students that were then used to prosecute them in juvenile court.164 Plaintiffs have similarly invoked the Due Process Clause
to challenge school discipline—most prominently in Goss v. Lopez,165
which resulted in a Supreme Court decision that students have due
process rights at school disciplinary proceedings166—and the Fourth
Amendment to challenge school searches.167
Plaintiffs have also attempted to seek relief under a variety of
common law tort claim theories, including negligence and intentional
infliction of emotional distress, but with only a handful of exceptions,
these claims have largely failed.168 Setting aside the practical difficul163. Complaint at 15, 25, 30, Antoine, No. Civ. 06-3007.
164. Consent Decree at 2-3, Antoine, No. Civ. 06-3007.
165. 419 U.S. 565, 581 (1975).
166. Id. at 574; see also, e.g., Ruiz v. Pedota, 321 F. Supp. 2d 538, 540-41, 543 (E.D.N.Y.
2004) (approving settlement agreement in suit challenging exclusions of students from
school in violation of Due Process Clause); D.C. v. Sch. Dist. of Phila., 879 A.2d 408, 419
(Pa. Commw. Ct. 2005) (holding students have right to opportunity to be heard prior to
being transferred to an alternative school and, incidentally, declining to address state
equal protection claim).
167. See, e.g., Safford Unified Sch. Dist. No. 1 v. Redding, 557 U.S. 364, 374-75 (2009)
(invalidating strip search of student); New Jersey v. T.L.O., 469 U.S. 325, 341-42 (1985)
(holding that the Fourth Amendment applies to students searched by school officials); Doe
v. Little Rock Sch. Dist., 380 F.3d, 349, 356 (8th Cir. 2004) (invalidating mass search of
entire student body); Atlanta Independent School System Officials Will Ensure Students’
Constitutional Rights Are Upheld After Settlement Of ACLU Lawsuit, AM. C.L. UNION (Dec.
16, 2009), http://www.aclu.org/racial-justice/atlanta-independent-school-system-officials-willensure-students-constitutional-right (describing settlement agreement that included end to
searches without reasonable suspicion).
168. See, e.g., Rembert v. Monroe Twp. Bd. of Educ., No. Civ. 95-4818 (JEI), 1997 WL
189318, at *7 (D.N.J. Apr. 14, 1997) (dismissing tort claims as being time-barred); Jackson
v. Katy Indep. Sch. Dist., 951 F. Supp. 1293, 1306 (S.D. Tex. 1996) (dismissing intentional
infliction of emotional distress claim because conduct was not sufficiently extreme or outrageous and negligence claims because of official immunity of defendants).
2014]
THE NEW RACIAL JUSTICE
867
ties in making out these claims, such as the need to overcome various
immunity doctrines, the very nature of tort relief is simply ineffective
at combating structural racial inequalities.169
(c) The Statutory Alternative
Reformers have had greater success with claims for statutory
relief, even (or particularly) where their statutory rights are not
explicitly premised on racial equality. Title VI of the 1964 Civil
Rights Act, for example, prohibits discrimination on the basis of
race, color, or national origin in programs or activities receiving federal funding.170 But because plaintiffs must still show discriminatory
intent to prevail on a Title VI disparate impact claim,171 such claims
can be nearly as difficult to make out as those under the Equal
Protection Clause.172
The Juvenile Justice and Delinquency Prevention Act of 2002,
which requires states participating in a federal funding program
to address disproportionate minority contact (DMC) within the juvenile justice system, has proven somewhat more helpful.173 For example, the Department of Justice and a Tennessee juvenile court (in a
jurisdiction where black children made up 97.8% of all juveniles referred to court) recently entered into an agreement to gather DMC
169. See, e.g., OWEN M. FISS, THE CIVIL RIGHTS INJUNCTION 90-91 (1978); Richard L. Abel,
A Critique of Torts, 37 UCLA L. REV. 785, 809 (1990); Chiang, supra note 133, at 680-82.
170. 42 U.S.C. § 2000d (2006); see also § 2000e-2(a) (prohibiting employment discrimination). Plaintiffs filing claims under Title VII are of course still free to allege disparate
impact as a basis for relief, see § 2000e-2(k), and a fair amount has been written about the
relationship between Title VII and the Equal Protection Clause. See, e.g., Mary C. Daly,
Some Runs, Some Hits, Some Errors – Keeping Score in the Affirmative Action Ballpark
from Weber to Johnson, 30 B.C. L. REV. 1, 6 (1988) (discussing relationship between Title
VII and equal protection in the context of affirmative action); Richard Primus, The Future
of Disparate Impact, 108 MICH. L. REV. 1341, 1342-43 (2010) (querying whether Title VII’s
disparate impact standard conflicts with equal protection doctrine); see also Ricci v. DeStefano, 557 U.S. 557, 595-96 (2009) (Scalia, J., concurring) (“[T]he war between disparate
impact and equal protection will be raised sooner or later, and it behooves us to begin
thinking about how—and on what terms—to make peace between them.”).
171. See Alexander v. Sandoval, 532 U.S. 275, 280 (2001) (noting that Title VI prohibits
only intentional discrimination).
172. See, e.g., Denney v. City of Albany, 247 F.3d 1172, 1186-87 (11th Cir. 2001) (dismissing plaintiff’s Title VII claim for failing to prove discriminatory intent in promotion
practices); Barnett v. Johnson City Sch. Dist., No. 3:04-CV-0763, 2006 WL 3423872, at *12
(N.D.N.Y. Nov. 28, 2006) (dismissing claim for failure to show disparate treatment); Brown
v. City of Grand Prairie, No. CIV.A. 3:01–CV–0139, 2002 WL 171728, at *5 (N.D. Tex. Jan.
29, 2002) (dismissing plaintiff’s equal protection claim for failure to prove discriminatory
intent); Jackson, 951 F. Supp. at 1300 (dismissing Title VI claim for failure to prove
discriminatory intent).
173. Juvenile Justice and Delinquency Prevention Act of 2002, 42 U.S.C. § 5633(a)(22)
(2006).
868
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
data to seek alternatives to juvenile detention, to train juvenile court
staff on racial bias recognition and reduction, and to form a plan to
reduce DMC.174
The No Child Left Behind Act (NCLB) contains a number of raceconscious accountability requirements and is explicitly directed at
“closing the achievement gap between high- and low-performing children, especially the achievement gaps between minority and nonminority students, and between disadvantaged children and their more
advantaged peers.”175 Its efficacy as a tool for reform, however, has
been largely limited to the data collection it enables.176 For the first
time, advocates are able to access education statistics disaggregated
by race; although the disparate impact those statistics indicate may
be insufficient for an equal protection claim, knowledge of its existence helps reformers to know they are targeting the right school
or schools.177
Other federal statutes that do not explicitly address race have
proven even more fruitful for litigators. Although there is no one
statute suitable for every school-to-prison pipeline claim, or even
most such claims, and although some claims cannot be addressed by
an existing statute, the statutory landscape is sufficiently rich to
provide a meaningful alternative to the Equal Protection Clause,
particularly when one considers the overlap between the groups targeted for statutory protection and children of color.
The Equal Educational Opportunities Act (EEOA) provides protections for children who do not speak English as a native language.178
Many if not all of these children are children of color, and many suffer from the effects of the school-to-prison pipeline.179 The EEOA re174. U.S. DEP’T OF JUST., C.R. DIV., MEMORANDUM OF AGREEMENT REGARDING THE
JUVENILE COURT OF MEMPHIS AND SHELBY COUNTY 21-23, 26 (2012), available at
http://www.justice.gov/crt/about/spl/documents/shelbycountyjuv_agreement_12-17-12.pdf. According to the Department of Justice, twenty-three states have full-time, state-level, DMC coordinators; thirty-one states have part-time or other state-level staff designated as DMC coordinators; thirty-four states have invested in targeted local DMC-reduction sites; and twelve states
have laws intended to reduce DMC. JEFF SLOWIKOWSKI, OFF. OF JUV. JUST. AND DELINQ.
PREVENTION, DISPROPORTIONATE MINORITY CONTACT 3 (Oct. 2009), available at
https://www.ncjrs.gov/pdffiles1/ojjdp/228306.pdf.
175. 20 U.S.C. § 6301(3) (2012).
176. Some have criticized NCLB for increasing school incentives to push lowperforming students of color out of schools. See, e.g., Daniel J. Losen, Challenging Racial
Disparities: The Promise and Pitfalls of the No Child Left Behind Act’s Race-Conscious
Accountability, 47 HOW. L.J. 243, 290-94 (2004).
177. See, e.g., No Child Left Behind Act of 2001, Pub. L. No. 107-110, § 1907, 115 Stat.
1425 (2002) (requiring states to collect school district data on annual school dropout rates
disaggregated by race and ethnicity).
178. 20 U.S.C. § 1703(f) (2012).
179. KIM ET AL., supra note 135, at 44.
2014]
THE NEW RACIAL JUSTICE
869
quires state and local educational agencies to help children with language barriers overcome those barriers and to provide them with an
adequate education.180 In particular, it mandates that English Language Learner students be provided with an education equal to that
provided to native English speakers and that they be instructed in
English.181 Reformers have filed cases both directly under the EEOA
and through the U.S. Department of Education, Civil Rights Division
to enforce these provisions.182
The Individuals with Disabilities Education Act (IDEA),183 section
504 of the Rehabilitation Act,184 and Title II provide an array of protections for students with disabilities.185 Black students are often
over-identified for certain types of disabilities and under-identified
for others.186 Disabled students of color are segregated from mainstream education more often than disabled white students.187 Disa180. § 1703(b), (f).
181. Castaneda v. Pickard, 648 F.2d 989, 1011 (5th Cir. 1981).
182. See, e.g., Gomez v. Ill. St. Bd. of Educ., 811 F.2d 1030, 1044-45 (7th Cir. 1987)
(dismissing equal protection claim for failure to allege intent to discriminate but permitting EEOA claim to proceed); Castaneda, 648 F.2d at 1015 (denying Title VI claim for lack
of intent to discriminate but remanding for further proceedings on EEOA claim); Flores v.
Arizona, 172 F. Supp. 2d 1225, 1239 (D. Ariz. 2000) (holding that state disbursement of 150
dollars per Limited English Proficient student violated the EEOA by providing insufficient
English instruction for students not proficient in English); U.S. DEP’T OF JUST.,
AGREEMENT BETWEEN THE UNITED STATES OF AMERICA AND THE LEWISTON, ME SCHOOL
DEPARTMENT 3-4 (2010), available at http://www.justice.gov/crt/about/edu/documents/Lew
istonAgree.pdf); Letter from March Roosevelt, Superintendent of Pittsburgh Pub. Schools,
to Michael Branigan, U.S. Dep’t of Educ. (May 15, 2006) (on file with author) (listing actions district will take to provide required services to Somali-speaking students); see also
Morales v. Shannon, 516 F.2d 411, 415 (5th Cir. 1975) (holding that failure to take appropriate actions to overcome language barriers is unlawful); Martin Luther King Jr. Elementary Sch. Child. v. Ann Arbor Sch. Dist. Bd., 473 F. Supp. 1371, 1390-91 (E.D. Mich. 1979)
(holding that school district is obligated to develop a program to assist teachers to take
home language into account in addressing children’s reading problems).
183. 20 U.S.C. §§ 1400–1410 (2012).
184. Pub. L. 93-112, § 504, 87 Stat. 355, 394 (1973) (codified at 29 U.S.C. § 794 (2006)).
185. 42 U.S.C. §§ 12131–12165 (2006).
186. See, e.g., U.S. DEP’T OF EDUC., 29TH ANNUAL REPORT TO CONGRESS ON THE
IMPLEMENTATION OF THE INDIVIDUALS WITH DISABILITIES EDUCATION ACT, 2007 123, 129
(2010) (noting that on average, across all fifty states and the District of Columbia, 1.81% of
black students ages six through twenty-one were labeled mentally retarded, and 1.37%
were labeled emotionally disturbed, whereas white students in that age range were given
those labels nearly three times less often).
187. See, e.g., KIM ET AL., supra note 135, at 54 (“Racial disparities are most pronounced among those students who are educated in regular schools but in settings that are
separate from their nondisabled peers for more than 60 percent of the school day.”). Systematic segregation from mainstream education can also occur when students of color with
disabilities are disproportionately punished with out-of-school suspension. LOSEN &
GILLESPIE, supra note 136, at 12-21 (noting that students of color with special needs “face
double the risk” for suspension compared with their non-disabled peers).
870
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
bled black students are disciplined more frequently and more severely than their white counterparts.188 And disabled black students receive supports and services that are inferior to those given to their
white counterparts.189
The IDEA guarantees children with disabilities a free and appropriate public education in the least restrictive environment and in
accordance with an individualized education plan;190 requires states
to make affirmative efforts to identify children with disabilities;191
prohibits schools from disciplining disabled students without first
establishing that the student’s conduct was not a manifestation of his
or her disability;192 and requires schools to put together a behavior
intervention plan when the misconduct at issue was in fact a manifestation of a disability.193 Section 504 of the Rehabilitation Act of
1973 prohibits recipients of federal funding from discriminating on
the basis of disability, and Title II prohibits such discrimination by
public schools and state departments of education regardless of the
receipt of federal funding.194 Advocates have had success filing challenges under a number of these provisions.195
188. See, e.g., KIM ET AL., supra note 135, at 170 n.22 (noting black students represented nearly half of all of the reported suspensions longer than ten days in 2005 for students
with disabilities); DANIEL J. LOSEN & TIA ELENA MARTINEZ, THE C.R. PROJECT, OUT OF
SCHOOL & OFF TRACK: THE OVERUSE OF SUSPENSIONS IN AMERICAN MIDDLE AND HIGH
SCHOOLS 11 (2013) (finding that 36% of all black male students with disabilities enrolled in
middle and high schools were suspended at least once in 2009-2010, compared to 17% for
white males with disabilities, and 6% for white females with disabilities).
189. David Osher et al., Schools Make a Difference: The Overrepresentation of African
American Youth in Special Education and the Juvenile Justice System, in RACIAL INEQUITY
IN SPECIAL EDUCATION 93, 93-116 (Daniel J. Losen & Gary Orfield eds., 2002).
190. 20 U.S.C. §1414 (2012).
191. § 1412(a)(3)(A).
192. 34 C.F.R. § 300.530(c), (e)(1)-(2) (2013).
193. Id. § 300.530(f)(1)(i)-(ii).
194. 42 U.S.C. §§ 12131–12165 (2006); Pub. L. 93-112, § 504, 87 Stat. 355, 394 (codified
at 29 U.S.C. § 794 (2006)).
195. See, e.g., Schmelzer v. New York, 363 F. Supp. 2d 453, 461 (E.D.N.Y. 2003) (denying summary judgment under Rehabilitation Act because plaintiffs failed to prove discrimination against disabled children but granting injunctive relief under IDEA); Corey H. v.
Bd. of Educ. of Chi., 995 F. Supp. 900, 918 (N.D. Ill. 1998) (granting injunctive relief under
IDEA for class of disabled children in Chicago); Order Preliminary Approving Settlement,
Directing Notice to the Class and Setting a Hearing on Proposed Settlement, Ray M. v. Bd.
of Educ. of N.Y.C., 884 F. Supp. 696 (E.D.N.Y. 1995) (No. 94 Civ. 1103 (EHN)/(JLC))
(providing preliminary approval of settlement in case involving disabled preschool students
who had not been evaluated for special education services in their native language); Settlement Agreement, Kina K. ex rel. Jamie S. v. Milwaukee Bd. of Sch. Dirs., No. 01-C-0928
(E.D. Wis. Feb. 27, 2008); Settlement Agreement at 1, P.J. v. Conn., Bd. of Educ., Civil
Action No. 291CV00180 (RNC) (D. Conn. filed May 31, 2002) (settling IDEA case that
covered all mentally retarded school-aged children in Connecticut).
2014]
THE NEW RACIAL JUSTICE
871
The McKinney Vento Act provides education protections for students who are homeless or in foster care, requiring that: students
who fall under its ambit be permitted to attend either their local
school or school of origin (and be provided transportation if they wish
to stay at their school of origin); such students be permitted to enroll
at a new school even if unable to produce documents such as proof of
residency or medical records; and such students receive services under Title I, Part A, of NCLB.196 Although statistics on the racial composition of homeless students are hard to come by, there is evidence
that children under eighteen form a disproportionate percentage of
the homeless population, which is in turn disproportionately of color.197 Reformers have had success filing claims under the McKinney
Vento Act, particularly in the aftermath of natural disasters such as
Hurricane Katrina.198
3. Beyond the School-to-Prison Pipeline—Lessons for Other
Claims
When statutory relief is available, it is a powerful resource for
advocates seeking to address racial disparities. Reformers in the
school-to-prison pipeline context have had the greatest success when
they have based their claims not on disparate impact, but rather
on statutory rights to which anyone would be entitled—a legislative
equivalent to the “liberty-based claim” referred to by Yoshino
and others.
The statutory and regulatory regime is complex and wide-ranging.
This Part covered only a handful of federal statutes relevant for pipeline purposes, but reformers seeking to address any sort of racial
disparity look to state statutes as well. And even where there is no
private right of action under a particular statute, the history of
school-to-prison pipeline reform demonstrates that relief may nevertheless be available through work with government agencies or
branches leveraging the statutory entitlements.
196. 42 U.S.C. § 11432(g)(1)(J)(iii)(I), (g)(3)(C)(i), (g)(4)(A)-(B) (2006).
197. See, e.g., Who Is Homeless?, NAT’L COAL. FOR THE HOMELESS (July 2009),
http://nationalhomeless.org/factsheets/who.html (noting that in 2003, children under eighteen comprised 39% of the homeless population and that a survey of twenty-five cities in
2006 indicated that the sheltered homeless population was 42% African-American).
198. See, e.g., Order at 1-2, Boisseau v. Picard, Civil Action No. 07-00565 (E.D. La. Jan.
7, 2008) (dismissing case following settlement of claims on behalf of children made homeless by Hurricane Katrina); see also Permanent Injunction at 1, Bullock v. Bd. of Educ. of
Montgomery Cnty., Civil Action No. DKC02CV798 (D. Md. Mar. 28, 2005) (enjoining school
district to permit homeless children to attend any school in the feeder system for the school
the student attended prior to homelessness, and to provide transportation services).
872
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
Where explicitly race-based statutory protections are available,
reformers do and should make use of those protections. For example,
the ACLU’s Racial Justice Project recently sued Morgan Stanley for
the disparate impact its sub-prime lending practices had on people of
color in Detroit, relying primarily on the federal Fair Housing Act
and the Equal Credit Opportunity Act, which prohibit racial discrimination in residential real estate and credit transactions.199
But the availability of federal statutory relief for claims based on
racial disparity in the absence of discriminatory intent appears to be
on the wane. Indeed, lower court skepticism about the substantive
merit of claims under section 5 of the Voting Rights Act (an Act
grounded in the Fifteenth Amendment and containing protections
explicitly designed to protect racial minorities),200 has recently given
way to a Supreme Court decision that holds section 4 of the Act unconstitutional.201 This Part catalogs some of the ways one group of
reformers has attempted to address this problem and urges those
who have not previously considered racially neutral statutory claims
to do so.
C. Immigration, Structural Arguments, and Racial Animus
Over the last decade, a number of states and localities have crafted legislation designed to discourage undocumented workers from
entering their jurisdictions and to encourage those already present to
leave. The bills passed typically contain some combination of the following provisions:202 prohibitions on transporting or concealing un199. Class Action Complaint at 1-3, Adkins v. Morgan Stanley, No. 12 CIV 7667(HB),
2013 WL 3835198 (S.D.N.Y. Oct. 15, 2012). The case survived a motion to dismiss. Adkins
v. Morgan Stanley, No. 12 Civ. 7667(HB), 2013 WL 3835198 (S.D.N.Y. July 25, 2013).
200. See, e.g., South Carolina v. United States, 898 F. Supp. 2d 30, 39 (D.C. Cir. 2012)
(upholding South Carolina Voter Identification Act under section 5 of the Voting Rights
Act); Florida v. United States, 885 F. Supp. 2d 299, 321-22, 353 (D.D.C. 2012) (denying
claim that change in voting procedures would have retrogressive effect on minority voters
in violation of Voting Rights Act and finding procedures were not enacted with discriminatory purpose); Perez v. Texas, 891 F. Supp. 2d 808, 834-35 (W.D. Tex. 2012) (denying claim
that congressional redistricting constituted impermissible racial gerrymander in violation
of Voting Rights Act). But see Texas v. Holder, 888 F. Supp. 2d 113, 143-44 (D.C.C. 2012)
(denying motion state request for declaratory judgment that voter identification law did
not violate the Voting Rights Act), vacated and remanded, 133 S. Ct. 2886 (2013).
201. Shelby Cty. v. Holder, No. 12-96, slip op. at 24 (U.S. June 25, 2013). In so holding,
the Court repeatedly emphasized its view that minority voters no longer suffer any disparate impact as a result of state and local voting laws. See id. at 4 (“Census Bureau data
indicate that African-American voter turnout has come to exceed white voter turnout in
five of the six States originally covered by §5, with a gap in the sixth State of less than one
half of one percent.”).
202. This list is by no means comprehensive. According to the National Conference on
State Legislatures, lawmakers in forty-six states and the District of Columbia introduced
948 bills and resolutions related to immigration, and enacted 114 of those bills and adopted
2014]
THE NEW RACIAL JUSTICE
873
documented workers,203 inducing an undocumented worker to enter
the state,204 and/or hiring an undocumented worker;205 requirements
related to documentation immigrants must produce or can use to
verify the lawfulness of their presence in the jurisdiction;206 providing
for additional state penalties for violating federal law;207 and/or authorizing local law enforcement to determine whether a person is
removable or detainable under federal law.208 This Part will explore
92 of those resolutions in the first half of 2012 alone. This flurry of activity actually marked
a decline from 2011. 2012 Immigration-Related Laws and Resolutions in the States (January 1 – June 30, 2012), NAT’L CONF. ON ST. LEGS. (Aug. 6, 2012), http://www.ncsl.org/issuesresearch/immig/2012-immigration-related-laws-and-resolutions.aspx.
203. See, e.g., Fremont, Neb., Ordinance 5165 (June 21, 2010) (prohibiting the harboring of an illegal alien); H.R. 87, § 7, 151st Gen. Assemb., 2011 Reg. Sess. (Ga. 2011) (codified at GA. CODE ANN. § 16-11-200 (West 2014)) (prohibiting the transport or movement of
an illegal alien “for the purpose of furthering the illegal presence of the alien in the United
States” and the concealment or harboring of an illegal alien); H.R. 56, § 13, 2011 Reg. Sess.
(Ala. 2011) (codified at ALA. CODE § 31-13-13(a)(1), (3) (2014)) (prohibiting the concealment, harboring or shielding from detection of any alien, as well as transporting an alien
“in furtherance of the unlawful presence of the alien”); S. 20, § 4, 119th Gen. Assemb., Sess.
(S.C. 2011) (codified at S.C. CODE ANN. § 16-9-460 (2014)) (making it a state crime for unlawfully present persons to shelter, harbor or transport themselves).
204. ALA. CODE § 31-13-13(a)(2) (2014); H.R. 87, § 7 (codified at GA. CODE ANN. § 16-11202(b) (West 2014)) (prohibiting people from inducing an illegal alien to enter into Georgia).
205. ALA. CODE § 31-13-11(a) (2014) (criminalizing an unauthorized alien’s application
for, solicitation of, or performance of work); ARIZ. REV. STAT. ANN. § 13-2928(C) (2014)
(making it a misdemeanor for an unauthorized alien to seek or engage in work in the
state); Fremont, Neb., Ordinance 5165, § 5(C) (June 21, 2010) (requiring employers to
execute an affidavit that they have not knowingly employed any unauthorized aliens); S.
20, § 2 (codified at S.C. CODE ANN. § 8-14-10(9) (2014)) (creating sanctions on employers
who knowingly hire unauthorized aliens).
206. ARIZ. REV. STAT. ANN. § 11-1051(B) (2014) (requiring officers in some circumstances to verify immigration status when conducting a stop, detention, or arrest); H.R. 87, § 8
(codified at GA. CODE ANN. § 17-5-100(b) (West 2014)) (authorizing but not requiring state
officials to conduct inquiry into immigration status); S. 590, § 18, 117th Gen. Assemb., First
Reg. Sess. (Ind. 2011) (codified at IND. CODE § 34-28-8.2-2 (2013)) (making it an infraction to
knowingly offer or accept a consular identification card as a valid form of identification).
207. ALA. CODE § 31-13-10(a) (2014) (also making failure to comply with federal alien
registration requirements a state misdemeanor); ARIZ. REV. STAT. ANN. § 13-1509 (2014)
(making failure to comply with federal alien registration requirements a state misdemeanor); S. 20, § 5 (codified at S.C. CODE ANN. § 16-17-740 (2014)) (also making failure to comply
with federal alien registration requirements a state misdemeanor).
208. ALA. CODE § 31-13-12(a) (2014) (requiring officers to determine immigration status when officer has reasonable suspicion a lawfully seized individual is unlawfully present); ARIZ. REV. STAT. ANN. § 13-3883(A)(5) (2014) (authorizing local law enforcement to
arrest without a warrant a person “the officer has probable cause to believe . . . has committed any public offense that makes the person removable from the United States”); GA.
CODE ANN. § 17-5-100(e) (West 2014) (authorizing local law enforcement to detain unauthorized aliens, to transport them to a detention facility, and/or to notify the Department of
Homeland Security); S. 590, § 19 (codified at IND. CODE § 35-33-1-1(a)(11), (12), (13) (2013))
(authorizing officials to make warrantless arrests of individuals when the officer has a
removal order issued by an immigration court, a detainer or notice of action issued by the
Department of Homeland Security, or the officer has probable cause to believe the person
874
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
the traditional equal protection claims available to combat any disparate racial impact these bills may have and identify the nontraditional legal challenges reformers have brought to bear in lieu of,
or in addition to, equal protection claims.
1. The Disparate Impact
Setting aside the facial classifications on the basis of citizenship or
alienage presented by these bills, this type of legislation also presents
the type of racial disparate impact problem in which this Article is
primarily interested. The reform response to these bills is animated
largely by a suspicion that their enforcement will disproportionately
target undocumented people of color, and more broadly, all persons
who look Latino, regardless of actual ethnic heritage or legal status.209
A number of these immigration bills contain provisions calculated
to inoculate the laws from traditional equal protection challenges by
specifically prohibiting consideration of race, color, or national origin
except as authorized by state and federal constitutions, and by requiring that implementation be consistent with federal laws governing civil rights.210 They explicitly disavow discriminatory intent and
thus seek to foreclose any cause of action via Washington v. Davis.
Although there may be some atmospheric evidence of discriminatory
intent, reformers seeking to combat the anticipated racial disparate
impact have had to think beyond the Equal Protection Clause.211
has been indicted for or convicted of one or more aggravated felonies); Fremont, Neb. Ordinance 5165, § 4(D) (June 21, 2010) (requiring city police to verify immigration status of
building occupants who have not declared themselves to be U.S. citizens and to revoke
occupancy licenses for unauthorized aliens); S. 20, § 6 (codified at S.C. CODE ANN. § 17-13170(A) (2014)) (requiring officers to determine immigration status when there is reasonable
suspicion a person being stopped or arrested is present unlawfully).
209. See, e.g., State Anti-Immigration Laws, AM. C.L. UNION, http://www.aclu.org/
immigrants-rights/state-anti-immigrant-laws (last visited June 29, 2014) (“Laws inspired
by Arizona’s SB 1070 invite rampant racial profiling against Latinos, Asian-Americans and
others presumed to be ‘foreign’ based on how they look or sound.”).
210. ARIZ. REV. STAT. ANN. § 11-1051(B), (L) (2014); H.R. 87, § 8 (codified at GA. CODE
ANN. § 17-5-100(d) (West 2014)), 2011 Ga. Laws 794, 805.
211. See, e.g., Cent. Ala. Fair Hous. Ctr. v. Magee, 835 F. Supp. 2d 1165, 1192-93 (M.D.
Ala. 2011) (stating that legislative history shows discriminatory intent behind Alabama
immigration laws).
2014]
THE NEW RACIAL JUSTICE
875
2. The Reform Response
(a) Facial Classification Based on Alienage
As a whole, immigration-related bills, like problems with state
public defense systems and the school-to-prison pipeline, have not
proven particularly amenable to equal protection challenges.212 The
bills largely target undocumented immigrants, who do not qualify as
a protected class and thus receive only rational basis review under
traditional equal protection analysis.213 To the extent that advocates
have attempted to make claims based upon disparate racial impact,
they have been foreclosed by the problem at the heart of this Article:
the requirement of discriminatory intent.214
The legislation has presented reformers seeking to combat racial
disparity with some low-hanging fruit, however. First, unlike the
previous two types of state action described, the immigration-related
bills sometimes present a facial classification based upon alienage
(citizenship), which is a protected class entitled to heightened scrutiny.215 Discrimination on the basis of citizenship or nationality is often
inextricably linked with discrimination on the basis of race. Second,
although state action involving the classification of undocumented
adult aliens receives only rational basis review, it still must survive
at least that minimal level of scrutiny.216 And finally, the undocumented children of undocumented aliens may qualify for some additional protections under the Fourteenth Amendment.217
At least some of the bills in question have proven vulnerable on
each of these fronts. It is worth noting that although these successful
claims rely on the Equal Protection Clause, they rely on a raceneutral aspect of the Clause. In Buquer v. Indianapolis,218 for example, the court found that a provision barring use of consular identification cards as valid identification did not satisfy rational basis review under equal protection analysis because it was “designed simply
212. See, e.g., Keller v. City of Fremont, 853 F. Supp. 2d 959, 975 (D. Neb. 2012) (rejecting equal protection challenge to various immigration related bills because city articulated
a rational basis for the different treatment afforded adults who are lawfully present and
those who are not).
213. Plyler v. Doe, 457 U.S. 202, 216 (1982).
214. See, e.g., Keller, 853 F. Supp. 2d at 982 (holding plaintiffs failed to state a claim
based upon disparate impact, because they failed to show discriminatory intent).
215. Graham v. Richardson, 403 U.S. 365, 371-72 (1971).
216. See Plyler, 457 U.S. at 223-25.
217. Id. at 216-17.
218. 797 F. Supp. 2d 905 (S.D. Ind. 2011).
876
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
to target foreign nationals.”219 Similarly, in Hispanic Interest Coalition of Alabama v. Governor of Alabama,220 the Eleventh Circuit
found that a state requirement that elementary and secondary
schools determine the immigration status of students violated Equal
Protection because the asserted state interest in collecting such information was insufficiently compelling and would impermissibly
burden the right of such children to obtain an education.221 In Ruiz v.
Robinson,222 the Southern District of Florida found unconstitutional a
state law that denied in-state tuition benefits to U.S. citizen students
who could not prove the federal immigration status of their parents.223 The court reasoned that the facial classification upon which
the statute rested (between U.S.-citizen students who could provide
immigration papers for their parents versus those who could not)
required heightened scrutiny because it punished citizen children for
the acts of their parents and that the classification failed to survive
that level of scrutiny.224
(b) Federal Preemption
The most successful across-the-board line of attack on these types
of bills has been a structural argument that these provisions are
preempted by federal law. The state legislation described follows a
long history of anti-immigrant legislation in this country, at least two
previous waves of which the Supreme Court has also dealt with on
federal preemption grounds.225 This time around, the Court ruled in
Arizona v. United States that three of the four key Arizona immigration law provisions were federally preempted.226 State and local laws
may be preempted in one of three ways: express preemption, in which
Congress explicitly withdraws specified powers from the states via
statute; field preemption, in which Congress determines that an entire area of the law must be regulated by the federal government exclusively; and conflict preemption, in which compliance with both the
219. Id. at 924.
220. 691 F.3d 1236 (11th Cir. 2012).
221. Id. at 1246-49.
222. 892 F. Supp. 2d 1321 (S.D. Fla. 2012).
223. Id. at 1333.
224. Id. at 1331.
225. See, e.g., Hines v. Davidowitz, 312 U.S. 52, 69 (1941) (finding state-imposed alien
registration requirements to be federally preempted and noting that “[o]ur Constitution
and our Civil Rights Act have guaranteed to aliens the equal protection of the laws which
is a pledge of the protection of equal laws” (quotation marks omitted)); Chy Lung v. Freeman, 92 U.S. 275, 280 (1875) (finding California statute that required bond payment by
certain foreign passengers to be federally preempted).
226. 132 S. Ct. 2492, 2510 (2012). Arizona, of course, rests firmly on the shoulders of
precedent like Hines. See generally Hines, 312 U.S. 52.
2014]
THE NEW RACIAL JUSTICE
877
state law and the federal law is impossible.227 In the context of immigration, the Court has often emphasized the traditional federal power
to determine immigration policy and the extensive way in which
Congress had already done so, occupying the field.228
In Arizona, the Court held that the state could not make “willful
failure to complete or carry an alien registration document” a state
misdemeanor because Congress had occupied the entire field of alien
registration.229 It found the state could not prohibit aliens from applying for or soliciting work because Congress had already enacted a
“comprehensive framework for ‘combating the employment of illegal
aliens.’ ”230 And it found that the state could not deputize state law
enforcement officials to arrest anyone who they had probable cause to
believe had committed an offense that would make them removable
from the United States because “[f]ederal law specifies limited circumstances in which state officers may perform the functions of an
immigration officer.”231
In the wake of the Court’s decision, lower courts have found a
spate of laws patterned upon those struck down in Arizona to be
preempted on similar grounds.232
(c) What Remains: Disparate Impact and a Return to the Bill of
Rights
The Court’s opinion in Arizona essentially deferred decision on the
fourth provision of the Arizona bill, which requires state officials to
make a reasonable attempt to determine the immigration status of
anyone stopped, detained, or arrested if there is reasonable suspicion
227. See generally Arizona, 132 S. Ct. at 2500-01 (summarizing basic preemption principles).
228. Id. at 2498-99 (“It is fundamental that foreign countries concerned about the
status, safety, and security of their nationals in the United States must be able to confer
and communicate on this subject with one national sovereign, not the 50 separate States.”).
229. Id. at 2501-02 (“[E]ven complementary state regulation is impermissible. Field
preemption reflects a congressional decision to foreclose any state regulation in the area,
even if it is parallel to federal standards.”).
230. Id. at 2504.
231. Id. at 2506. For discussion of the fourth provision of the Arizona legislation, which
the Court permitted to survive, see infra Part III.C.2.c.
232. See, e.g., Ga. Latino Alliance for Hum. Rts. v. Governor of Ga., 691 F.3d 1250,
1266-67 (11th Cir. 2012); United States v. Alabama, 691 F.3d 1269, 1292 (11th Cir. 2012);
Keller v. City of Fremont, 853 F. Supp. 2d 959, 973 (D. Neb. 2012). Of course, some lower
courts had already invalidated similar provisions as being preempted even before the
Court’s decision was announced. See, e.g., Villas at Parkside Partners v. City of Farmers
Branch, 675 F.3d 802, 817 (5th Cir. 2012); United States v. South Carolina, 840 F. Supp.
2d 898, 917-24 (D.S.C. 2011); Buquer v. City of Indianapolis, 797 F. Supp. 2d 905, 920 (S.D.
Ind. 2011).
878
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
that the person is an unlawfully present alien.233 Lower courts confronted by similar provisions have likewise withheld judgment.234 The
Court has also refused to find preempted state statutes that require
use of the federal E-Verify program to ascertain the work authorization status of employees.235
Advocates seeking to challenge these two types of state action—
immigration status checks conducted by state law enforcement officials and use of E-Verify to check work authorization status—must
look beyond preemption. Both of these types of legislation are facially
neutral but raise potential disparate impact concerns that workers of
certain races will be disproportionately subject to immigration or
work authorization checks. If the enforcement of these provisions is
sufficiently egregious with regard to racial impact, there is some
precedent for further challenge under equal protection, but such
claims are very difficult to make out.236
What remains, instead, is a return to the use of constitutional
protections other than the Equal Protection Clause to vindicate racial
inequality claims.237 Justice Alito acknowledges this eventuality in
Arizona: “If properly implemented, [the show-your-papers provision]
should not lead to federal constitutional violations, but there is no
denying that enforcement of [the provision] will multiply the occasions on which sensitive Fourth Amendment issues will crop up.”238
Some state provisions have in fact already been deemed unconstitutional on Fourth Amendment and due process grounds. For example, in Buquer v. City of Indianapolis,239 the Southern District of In233. Arizona v. United States, No. 11-182, slip op. at 23-24 (U.S. June 25, 2012). The
Court, it seems, did not see fit to take the invitation of a previous Court’s decision in Hines
to deem such provisions similarly preempted by an implicit Congressional desire “to leave
[aliens] free from the possibility of inquisitorial practices and police surveillance that might
not only affect our international relations but might also generate the very disloyalty
which the law has intended guarding against.” Hines v. Davidowitz, 312 U.S. 52, 74 (1941).
234. See, e.g., Ga. Latino Alliance, 691 F.3d at 1268; Alabama, 691 F.3d at 1292. But
see South Carolina, 840 F. Supp. 2d at 924 (finding state equivalent immigration status
check requirement preempted).
235. Chamber of Com. v. Whiting, 131 S. Ct. 1968, 1973 (2011) (upholding the Legal
Arizona Workers Act of 2007, which permits state courts to suspend or revoke business
licenses if an employer knowingly or intentionally employs an unauthorized alien and
requires use of the federal E-Verify program).
236. It appears the Court has not struck down a restriction purely on the grounds of
discriminatory enforcement since Yick Wo v. Hopkins, 118 U.S. 356 (1886).
237. See supra Part III.A.2.
238. Arizona v. United States, 132 S. Ct. 2492, 2529 (2012) (Alito, J., concurring in part
and dissenting in part); cf. Hines, 312 U.S. at 71 n.32 (“The requirement that cards be
carried and exhibited has always been regarded as one of the most objectionable features of
proposed registration systems, for it is thought to be a feature that best lends itself to
tyranny and intimidation.”).
239. 797 F. Supp. 2d 905 (S.D. Ind. 2011).
2014]
THE NEW RACIAL JUSTICE
879
diana granted a preliminary injunction that enjoined enforcement of
a state law provision authorizing state and local officials to conduct
warrantless arrests of individuals with outstanding federal removal
orders. In addition to the usual preemption problems the provision
presented,240 the court found that the provision violated the Fourth
Amendment because it authorized arrest for non-criminal offenses
and left a “deafening silence as to what happens to the arrestee post
his arrest.”241
3. Beyond Immigration—Lessons for Other Claims
The immigration cases demonstrate that even where the racial
animus is all but explicit, reformers still gain the most traction from
racially neutral claims. This observation holds true within the more
limited family of equal protection claims as well: even when the raceneutral claim is based on a factor linked to race (like citizenship),
race-neutral equal protection claims are more likely to succeed than
race-based claims. These cases may indicate that the more racially
inflammatory the underlying facts, the more important it is to rely
upon racially neutral legal claims. They also show, however, that
there is life yet in the Equal Protection Clause and that reformers
need not shy away from equal protection claims where they do exist,
as in facial classifications on the basis of alienage.242 And finally,
these cases serve as a powerful reminder that disparate racial impact
may sometimes productively be challenged on multiple fronts.
Reformers have made use of federal preemption in other contexts
as well. In a case related to the anti-immigration legislation, the Supreme Court recently struck down an Arizona proposition that required prospective voters to present documentary proof of citizenship
in order to register to vote.243 So-called “voter ID” laws like this one
have a disproportionate impact on people of color,244 who are less like240. Id. at 920.
241. Id. at 918; see also id. at 918-19 (“There is no mention of any requirement that
the arrested person be brought forthwith before a judge for consideration of detention or
release. There is in fact a complete void within the new statute regarding all other due
process protections.”).
242. Cf. Karl Manheim, State Immigration Laws and Federal Supremacy, 22 HASTINGS
CONST. L.Q. 939, 1012-17 (1995) (arguing that the Court’s equal protection alienage cases
are actually driven by preemption concerns).
243. Arizona v. Inter Tribal Council of Ariz., Inc., 133 S. Ct. 2247, 2260 (2013) (finding
proposition preempted under the National Voting Rights Act via the Elections Clause).
244. A number of laws seeking to restrict access to voting were passed or proposed in
the months leading up to the 2012 presidential election. See, e.g., ADVANCEMENT PROJECT,
VOTER PROTECTION PROGRAM, Segregating American Citizenship: Latino Voter Disenfranchisement in 2012 3-4 (2012), available at www.advancementproject.org/page/-/resources/
Latino%20Report%202012.pdf (summarizing laws).
880
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
ly than whites to have government issued identification.245 In a twist
on the use of federal preemption principles, the Court relied not
upon preemption under the Supremacy Clause, but rather under the
Elections Clause, which gives Congress the power to make or alter
state regulations governing the time, place, and manner of holding
federal elections.246
The Court’s decision is all the more interesting in light of the
claim that the Ninth Circuit denied, under Section 2 of the Voting
Rights Act, which provides explicitly race-based relief.247 The Ninth
Circuit noted the need for plaintiffs seeking such relief to demonstrate that the voting practice being challenged actually results in
race discrimination, with a causal connection between the practice
and the discriminatory result.248 It found that plaintiffs failed to provide evidence demonstrating even the disparate impact alleged.249
Here, racial justice was again better served by the race-neutral structural protections of federalism and federal preemption than by the
explicitly race-based protections conferred by federal statute.
Finally, courts across the country have dealt with a variety of cases involving disparate racial impact, immigration, and/or allegations
of racial animus in the post-9/11 cases. As in the other areas of law
examined by this Article, the disparate racial impact implicated by
these cases is relatively clear; the U.S. Government has focused almost exclusively on Muslims and predominantly Muslim countries in
an effort to prevent future terrorist attacks, and largely on Middle
245. See, e.g., NAT’L CONG. OF AM. INDIANS, VOTER ID LAWS & THE NATIVE VOTE: STATES OF
CONCERN 4, available at http://www.ncai.org/attachments/PolicyPaper_LaOnRbIuKyazjpiz
LTzOsFxvRiBpmwxxbGVFOeCNGMFwEuLZXwz_VoterIDs_NativeVote_States_of_Concern.pdf
(discussing the disparate impact voter ID laws have on Native Americans, noting, for example,
that many tribal communities do not have street addresses); JON C. ROGOWSKI & CATHY J.
COHEN, BLACK YOUTH PROJECT, TURNING BACK THE CLOCK ON VOTING RIGHTS: THE
IMPACT OF NEW PHOTO IDENTIFICATION REQUIREMENTS ON YOUNG PEOPLE OF COLOR 1
(2012), available at http://research.blackyouthproject.com/files/2012/09/Youth-of-Color-andPhoto-ID-Laws.pdf (focusing on effect of voter ID laws on young voters who are of color);
WENDY R. WEISER & LAWRENCE NORDEN, BRENNAN CTR. FOR JUST., VOTING LAW CHANGES
IN 2012 24 (2011), available at https://www.brennancenter.org/page/Democracy/VRE/
Brennan_Voting_Law_V10.pdf (noting that “as many as 25% of African-American voters do
not possess a current and valid form of government issued photo ID, compared to 11% of
voters of all races”).
246. Inter Tribal Council of Ariz., Inc., 133 S. Ct. at 2257.
247. 42 U.S.C. § 1973(a) (2006) (prohibiting states from imposing voting qualifications
that “result[] in a denial or abridgement of the right of any citizen of the United States to
vote on account of race or color”).
248. Gonzalez v. Arizona, 677 F.3d 383, 405 (9th Cir. 2012).
249. Id. at 407.
2014]
THE NEW RACIAL JUSTICE
881
Eastern citizens and countries.250 Those seeking to challenge directly
the racial justice implications of racial profiling or special registration requirements, however, have met with little success when relying on claims driven purely by equal protection.251 Race-neutral
claims, brought under the writ of habeas corpus, due process, and/or
federal statutes or treaties, have proven more fruitful.252
IV. CONCLUSION
The scenarios of racial inequality discussed above were neither
intentionally created by a racist government cabal nor the result of
bad state actors engaged in intentional discrimination. They arose,
instead, out of what some scholars have identified as structural racism: the development of social institutions over time, through myriad
government choices and actions.253 The resulting structural inequality is impossible to eradicate via the elimination of a single government policy or targeted firings of racist employees.
This new racial injustice is difficult, if not impossible, to combat
under the Court’s reading of the Equal Protection Clause, which requires the ill intent of a bad government actor, that is, the racist government employee. It requires, moreover, a structural solution. While
much of contemporary equal protection scholarship has focused on
how to extend the protections of the Equal Protection Clause to new
groups like lesbian women and gay men, this Article returns to the
paradigmatic application of the clause to prohibit classifications
based upon race.254 It offers an exploration of the many ways reform250. See, e.g., U.S. DEP’T OF ST., NATIONAL STRATEGY FOR COMBATING TERRORISM 2
(Sept. 2006), http://2001-2009.state.gov/s/ct/rls/wh/71803.htm (describing counter-terrorism
strategy and focusing on threat from radical Islamists in the Middle East).
251. See, e.g., Ashcroft v. Iqbal, 556 U.S. 662, 682-83 (2009) (dismissing complaint for
failing to allege facts showing petitioners “purposefully adopted a policy of classifying postSeptember-11 detainees as ‘of high interest’ because of their race, religion, or national
origin”); Rajah v. Mukasey, 544 F.3d 427, 438-39 (2d Cir. 2008) (denying equal protection
challenge to special registration requirements that applied only to adult male citizens of
Muslim majority states and North Korea because there was a rational national-security
basis for the program).
252. See, e.g., Boumediene v. Bush, 553 U.S. 723, 798 (2008) (holding that aliens held
as enemy combatants were entitled to habeas corpus hearings to challenge their detentions); Hamdan v. Rumsfeld, 548 U.S. 557, 567 (2006) (finding proposed military commissions violated Uniform Code of Military Justice and did not satisfy Geneva Conventions);
Hamdi v. Rumsfeld, 542 U.S. 507, 537-38 (2004) (requiring that citizen held as enemy
combatant be given meaningful opportunity to contest his detention pursuant to the Due
Process Clause).
253. See supra note 42.
254. This includes, of course, Yoshino’s article. Yoshino, supra note 9; see also Ball,
supra note 58, at 9-12 (discussing liberty and equality based claims in context of gay
rights); Tribe, supra note 48.
882
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:835
ers have found to fill the gaps wrought in the Equal Protection
Clause by Davis and its progeny, and it demonstrates that reformers
largely (and long ago) abandoned explicitly race-based claims, relying
instead on other guarantees contained in the Bill of Rights, statutory
claims, and federal structural arguments.
This Article is largely a descriptive project. It takes as a given
existing Court doctrine and recognizes—as reformers working on the
ground already have—that different paths can be taken to achieve
equality. But this description of today’s fight for racial equality helps
to illuminate what equality more generally might resemble in the
coming years, and whether and to what extent the Equal Protection
Clause will even be relevant.
Facially neutral policies with a disparate racial impact will likely
be a fact of life for some time. The continued vitality of the Equal
Protection Clause has perhaps never been in greater question, not
just for race, but for other groups seeking equality as well. The hardfought lessons racial justice reformers have learned over the years
may soon be relevant for others seeking to challenge disparate impact claims. This Article seeks to demonstrate that government policies creating such impact are not beyond the reach of the law. The
workarounds identified herein do not present a solution to every racial inequality—sometimes a political solution may be more appropriate. But where a judicial fix is feasible, these workarounds point
the way towards a new racial justice.
NATIONAL SECURITY RULEMAKING
ROBERT KNOWLES*
ABSTRACT
Agencies performing national security functions regulate citizens’ lives in increasingly
intimate ways. Yet national security rulemaking is a mystery to most Americans. Many
rules—like those implementing the National Security Agency’s vast surveillance schemes—
remain secret. Others are published, but the deliberations that led to them and the legal
justifications for them remain hidden.
Ordinarily, these rules would undergo the Administrative Procedure Act’s notice-andcomment process, which has earned wide, if not universal, praise for advancing democratic
values and enhancing agency effectiveness. But a national security exception from noticeand-comment in the APA itself, along with the overuse of classification authority, combine
to insulate most national security rulemaking from public scrutiny and meaningful judicial
review. The result is a national security administrative state that is insular and unaccountable to the public.
Some scholars find this exceptional treatment inevitable, while others have proposed
reforms. But no one has sought to provide a full accounting of national security rulemaking’s scope and historical origins. By doing so, this Article demonstrates that the APA exception is historically contingent—a response to the rise of totalitarian states and the Second
World War. As a product of its time rather than an essential attribute of all administrative
law systems, it is a relic in a globalized world in which the foreign and the domestic are
increasingly intertwined, and the line between national security and ordinary rulemaking
therefore begins to fade entirely.
This Article suggests reforms that would increase public deliberation in national security rulemaking, while accounting for the importance of secret-keeping when truly necessary.
Among these proposed reforms is a change to the current practice allowing national security
agencies to invoke the security exception to notice-and-comment after a rule is challenged in
court, rather than at the notice-and-comment stage itself. These reforms would improve the
current rulemaking practice, which undermines the transparency necessary for effective
democratic participation.
I. INTRODUCTION ..................................................................................................
II. OUR BIFURCATED ADMINISTRATIVE STATE .......................................................
A. The Importance of Notice-and-Comment Rulemaking ..............................
B. The Constitutional Dimension of Notice-and-Comment............................
C. Secrecy and the National Security Exception ............................................
III. THE SCHMITTIAN VIEW OF NATIONAL SECURITY RULEMAKING ........................
IV. THE HISTORICAL ROOTS OF NATIONAL SECURITY RULEMAKING ......................
V. REFORMING NATIONAL SECURITY RULEMAKING ..............................................
A. Eliminating the National Security Rulemaking Exception .......................
B. Encouraging Notice-and-Comment Rulemaking as a Best Practice .........
C. Establishing a Chenery-Type Rule for Invoking the National Security
Exception ....................................................................................................
VI. CONCLUSION .....................................................................................................
884
892
894
899
903
914
919
932
935
938
941
943
* Assistant Professor, Valparaiso University Law School. I must thank Anne
O’Connell, Barry Sullivan, Christopher Schmidt, Evan Criddle, David Herzig, Stuart Ford,
Rachel Levinson-Waldman, D.A. Jeremy Telman, Margaret Kwoka, Kati Kovacs, Anjali
Dalal, Catherine Deane, Neha Lall, Kavita Warrier, Marc Falkoff, and participants in
workshops at Loyola University Chicago School of Law, Chicago-Kent College of Law, and
the John Marshall Law School for helpful comments and advice. I also must thank Erin
Edwards for excellent research assistance and Scott Rafferty, Director of Research and
Policy at the Administrative Conference of the United States, for his valuable guidance.
884
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
“There are exceptions. If they cannot be explained, then the universal cannot be explained, either.”1
I. INTRODUCTION
In 2013, a series of leaks and disclosures revealed that the National Security Agency (NSA) conducts mass surveillance of Americans’ private electronic information on an unprecedented scale.2 The
content of e-mails, web searches, and phone calls of millions are regularly stored in databases, along with the metadata3 for such communications of hundreds of millions.4 Under certain rules—many still
secret—agencies such as the FBI and DOJ may access these records
to investigate unlawful activity by foreigners.5 But the difference between “foreign” and “American” often hangs on low-level NSA analysts’ judgment calls or search algorithms.6 And if these investigations
1. SØREN KIERKEGAARD, FEAR AND TREMBLING/REPETITION 227 (Howard V. Hong &
Edna H. Hong trans., Princeton University Press 1983) (1843).
2. The legal authorization for much of this surveillance was obtained by the FBI and
the Department of Justice. See Barton Gellman & Laura Poitras, U.S., British Intelligence
Mining Data from Nine U.S. Internet Companies in Broad Secret Program, WASH. POST
(June 7, 2013), http://www.washingtonpost.com/investigations/us-intelligence-mining-datafrom-nine-us-internet-companies-in-broad-secret-program/2013/06/06/3a0c0da8-cebf-11e28845-d970ccb04497_story.html.
3. Metadata is not the content of a record or communication but the information
about it, such as the time, date, and parties involved. See Jane Mayer, What’s the Matter
with Metadata?, NEW YORKER (June 6, 2013), http://www.newyorker.com/
online/blogs/newsdesk/2013/06/verizon-nsa-metadata-surveillance-problem.html. However,
metadata is often more useful for information-gathering than the content itself. See id.
4. See James Bamford, They Know Much More Than You Think, N.Y. REV. BOOKS
(Aug. 15, 2013), http://www.nybooks.com/articles/archives/2013/aug/15/nsa-they-knowmuch-more-you-think/?pagination=false (observing that metadata of telephone communications of “hundreds of millions of Americans” had been collected); Barton Gellman, U.S.
Surveillance Architecture Includes Collection of Revealing Internet, Phone Metadata, WASH.
POST (June 15, 2013), http://articles.washingtonpost.com/2013-06-15/news/39993852_
1_comey-national-intelligence-intelligence-collection
(describing
the
PRISM
and
NUCLEON Programs, which collect Internet and phone call content, respectively); Siobhan
Gorman, Evan Perez & Janet Hook, U.S. Collects Vast Data Trove, WALL ST. J. (June 7, 2013),
http://online.wsj.com/news/articles/SB10001424127887324299104578529112289298922
(describing the collection of phone and credit card transaction metadata); Glenn Greenwald, NSA Collecting Phone Records of Millions of Verizon Customers Daily, GUARDIAN
(London) (June 5, 2013), http://www.theguardian.com/world/2013/jun/06/nsa-phonerecords-verizon-court-order (same); Glenn Greenwald & Spencer Ackerman, How the NSA
is Still Harvesting Your Online Data, GUARDIAN (London) (June 27, 3013),
http://www.theguardian.com/world/2013/jun/27/nsa-online-metadata-collection (describing
the collection of Internet metadata).
5. See Gellman & Poitras, supra note 2.
6. See William C. Banks, Programmatic Surveillance and FISA: Of Needles in Haystacks, 88 TEX. L. REV. 1633, 1634 (2010) (“Instead of building toward an individual FISA
application by developing leads on individuals with some connection to an international
terrorist organization . . . officials now develop algorithms that search thousands or even
millions of collected e-mail messages and telephone calls for indications of suspicious activities.”); Glenn Greenwald & James Ball, The Top Secret Rules That Allow NSA to Use US
2014]
NATIONAL SECURITY RULEMAKING
885
uncover illegal activity by Americans—even inadvertently—the information is passed to other agencies for investigation and prosecution.7
This surveillance is far broader and deeper than previously understood.8 But it is also, to the surprise of many, arguably lawful.9 The
USA PATRIOT Act10 and the updated Foreign Intelligence Surveillance Act (FISA)11 give agencies wide discretion to obtain and search
private electronic files.12 The government claims that publicly availa-
Data
Without
a
Warrant,
GUARDIAN
(London)
(June
20,
2013),
http://www.theguardian.com/world/2013/jun/20/fisa-court-nsa-without-warrant.
7. See Greenwald & Ball, supra note 6. The DEA has a secretive Special Operations
Unit (SOP) that funnels information from intelligence databases to law enforcement agencies, which use it to launch investigations but do not disclose how these investigations
begin. See John Shiffman & Kristina Cooke, Exclusive: U.S. Directs Agents to Cover Up
Program Used to Investigate Americans, REUTERS (Aug. 5, 2013), http://www.reuters.com/
article/ 2013/08/05/us-dea-sod-idUSBRE97409R20130805.
8. For a history of the NSA’s surveillance of Americans in the first decade after the
September 11 attacks, see generally JAMES BAMFORD, THE SHADOW FACTORY: THE ULTRASECRET NSA FROM 9/11 TO THE EAVESDROPPING ON AMERICA (2008).
9. See, e.g., OFFICE OF THE DIR. OF NAT’L INTELLIGENCE, DNI STATEMENT ON RECENT
UNAUTHORIZED DISCLOSURES OF CLASSIFIED INFORMATION (2013), available at
http://www.dni.gov/index.php/newsroom/press-releases/191-press-releases-2013/868-dnistatement-on-recent-unauthorized-disclosures-of-classified-information
(defending
the
leaked surveillance program as having “been authorized by all three branches of
the Government”).
10. Uniting and Strengthening America by Providing Appropriate Tools Required to
Intercept and Obstruct Terrorism (USA PATRIOT Act) Act of 2001, Pub. L. No. 107-56, 115
Stat. 272.
11. Foreign Intelligence Surveillance Act of 1978, 50 U.S.C. §§ 1801–1855 (2006).
FISA was amended in 2008 to update and expand legal authority to search electronic information. FISA Amendments Act of 2008, Pub. L. No. 110-261, § 102, 122 Stat. 2436,
2459-60 (codified at 50 U.S.C. § 1812).
12. The government has not disclosed its complete legal justification for the NSA’s
programs, but some of the provisions that have been or may be relied on include USA
PATRIOT Act § 216, 18 U.S.C. § 3127(3) (2012) (expanding the availability of “pen register”
devices to encompass interception of internet metadata); USA PATRIOT Act § 214, 50
U.S.C. § 1842(c) (2006) (lowering the standard for obtaining internet metadata so that the
FBI need only certify to the Foreign Intelligence Surveillance Court (FISC) that the information likely to be obtained is “relevant to an ongoing investigation to protect against international terrorism or clandestine intelligence activities”); USA PATRIOT Act § 21, 50
U.S.C. § 1861 (2006) (the so-called “business records” provision, apparently interpreted by
the FISC, in a secret opinion, to authorize the FBI to obtain all of a carrier’s phone metadata, see David Cole, Can Privacy Be Saved?, N.Y. REV. BOOKS (Mar. 6, 2014),
http://www.nybooks.com/articles/archives/2014/mar/06/can-privacy-be-saved/?pagination=
false); and FISA Amendments Act § 702, 50 U.S.C. § 1881a (2006) (allowing the government to acquire foreign intelligence by obtaining the content of communications by nonU.S. persons “reasonably believed” to be outside U.S. borders, and interpreted to authorize
the collection of phone and Internet content of Americans in the process).
886
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
ble regulations13 and executive orders14 authorize parts of the controversial program. The Foreign Intelligence Surveillance Court (FISC),
through secret orders, approved some aspects.15 But most rules implementing the program were developed and approved internally by
agencies through secret rulemaking. Leaks or public pressure eventually forced disclosure of many rules, but secrecy still enshrouds the
processes that led to them and even the legal authority for them.16
Secret rulemaking makes up a growing portion of the federal government’s “legislative” rules—those that, among other things, prescribe rights or duties and fill statutory gaps—which have the force
of law.17 Yet under fundamental administrative law principles enu13. See NSA, UNITED STATES SIGNALS INTELLIGENCE DIRECTIVE 18: LEGAL
COMPLIANCE AND MINIMIZATION PROCEDURES (1993), available at http://www.gwu.edu/~ns
archiv/NSAEBB/NSAEBB23/07-01.htm (prescribing NSA safeguards for protecting constitutional rights of U.S. persons and regulating the collection, processing, and dissemination
of information concerning U.S. persons); U.S. DEP’T OF JUSTICE, THE ATTORNEY GENERAL’S
GUIDELINES
FOR
DOMESTIC
FBI
OPERATIONS
16
(2008),
available
at
http://www.justice.gov/ag/readingroom/guidelines.pdf (authorizing agents to engage in
proactive intelligence gathering in a manner “not limited to ‘investigation’ in a narrow
sense”); U.S. DEP’T OF DEF., DIRECTIVE 5240.1-R, PROCEDURES GOVERNING THE ACTIVITIES
OF DOD INTELLIGENCE COMPONENTS THAT AFFECT UNITED STATES PERSONS (1982), available at http://www.dtic.mil/whs/directives/ corres/pdf/524001r.pdf (regulating the role of DoD
personnel and resources in the NSA programs).
14. Executive Order 12,333 authorizes the NSA to collect “foreign intelligence or counterintelligence” information while not “acquiring information concerning the domestic activities of United States persons.” Exec. Order No. 12,333, 3 C.F.R. 200, 211 (1981). This
executive order was most recently amended by Executive Order 13,470, 3 C.F.R. 218, 229
(2008), which authorizes the NSA to “[p]rescribe . . . security regulations” consistent with
its authority.
15. See Eric Lichtblau, In Secret, Court Vastly Broadens Powers of N.S.A., N.Y. TIMES
(July 6, 2013), http://www.nytimes.com/2013/07/07/us/in-secret-court-vastly-broadenspowers-of-nsa.html?pagewanted=all&_r=0. FISA created the FISC to oversee requests for
surveillance warrants against suspected foreign intelligence agents inside the United
States by federal law enforcement agencies. See id. The FISC has been widely criticized for
its extreme deference to agencies and overly broad interpretations of statutory authority.
See, e.g., Letter from F. James Sensenbrenner, Jr., Congressman, U.S. House of Representatives, to Eric Holder, Attorney Gen., U.S. Dep’t of Justice (Sept. 6, 2013), available at
http://sensenbrenner.house.gov/uploadedfiles/sensenbrenner_letter_to_attorney_general_er
ic_holder.pdf (noting objections by the Patriot Act’s principal author to overly broad interpretations used to justify the NSA programs); infra Part II.C.
16. See, e.g., Samuel J. Rascoff, Domesticating Intelligence, 83 S. CAL. L. REV. 575, 632
(2010) (noting that the Attorney General Guidelines regulating domestic intelligencegathering by the FBI were traditionally kept secret but are now available online); Gellman
& Poitras, supra note 2 (describing secret rules instructing NSA analysts to enter search
terms “that are designed to produce at least 51 percent confidence in a target’s ‘foreignness’ ”);
Shiffman & Cooke, supra note 7 (describing secret documents regulating the DEA’s Special
Operations Division’s use of electronic information).
17. See Kristin E. Hickman, Unpacking the Force of Law, 66 VAND. L. REV. 465, 475
(2013); see also Hemp Indus. Ass’n v. DEA, 333 F.3d 1082, 1087 (9th Cir. 2003) (“Legislative rules . . . create rights, impose obligations, or effect a change in existing law pursuant
to authority delegated by Congress.”); United Techs. Corp. v. EPA, 821 F.2d 714, 719 (D.C.
Cir. 1987) (describing legislative rules as those “in which the agency sought to fill gaps and
inconsistencies left by the statutory scheme”); Jacob E. Gersen, Legislative Rules Revisited,
2014]
NATIONAL SECURITY RULEMAKING
887
merated in the Administrative Procedure Act (APA),18 these legislative rules must ordinarily be published and undergo the notice-andcomment process before they take effect.19 Notice-and-comment—
through which the public engages in dialogue with agencies, requiring
them to respond to its concerns—safeguards democratic values and enhances the quality of rulemaking.20 A prominent scholar in the field
called it “[o]ne of the [g]reatest [i]nventions of [m]odern [g]overnment.”21
Quite often, however, notice-and-comment is missing from national security rulemaking. Classification authority trumps other publication requirements,22 making notice-and-comment impossible.23 A
massive amount of government activity takes place entirely in secret.
By 2009, 1074 federal government organizations worked on programs
74 U. CHI. L. REV. 1705, 1708-09 (2007) (describing the difficulty courts sometimes encounter distinguishing legislative or “substantive” rules from merely “interpretive” rules).
18. 5 U.S.C. §§ 500–596 (2012); see also infra Part II.A (describing the APA’s fundamental importance in American law).
19. See 5 U.S.C. § 553(b) (This provision states that “[g]eneral notice of proposed rule
making shall be published in the Federal Register” unless there is “actual notice,” and
requires that the notice include “(1) a statement of the time, place, and nature of public
rule making proceedings; (2) reference to the legal authority under which the rule is proposed; and (3) either the terms or substance of the proposed rule or a description of the
subjects and issues involved.”); id. § 553(c) (requiring that agencies provide opportunity for
public and interested parties to respond to the agency notice of rulemaking by tendering
written opinions, information, or statements); id. § 553(d) (requiring a minimum thirty-day
grace period between the announcement of a rule and its effective date); Gersen, supra
note 17, at 1709-11 (explaining that legislative rules, also known as “substantive” rules,
may only be promulgated through notice and comment unless a statutory exception applies).
20. See infra Part II.A.
21. Kenneth Culp Davis, Administrative Law Treatise § 6.15, at 283 (Supp. 1970).
22. See Arthur Earl Bonfield, Military and Foreign Affairs Function Rule-Making
Under the APA, 71 MICH. L. REV. 221, 238-40 (1972). The APA’s publication requirements
originally included an exception for “any function of the United States requiring secrecy in
the public interest.” Administrative Procedure Act, Pub. L. No. 79-404, § 3, 60 Stat. 237,
238 (1946). When the APA was amended by the Freedom of Information Act (FOIA), Pub.
L. No. 89-487, 80 Stat. 250 (1966) (codified at 5 U.S.C. § 552), Congress included two exceptions related to national security—Exception 1, which authorizes withholding information
classified “by an Executive order to be kept secret in the interest of national defense or
foreign policy,” 5 U.S.C. § 552(b)(1), and Exception 3, which authorizes withholding information “exempted from disclosure by statute,” id. § 552(b)(3).
23. Because the existence of many national security-related agencies and programs is
secret, it is impossible to know how many legislative rules are developed in secret. See
infra Part II.C. Examples of statutes that authorize secret rulemaking include 50 U.S.C.A.
§§ 831–832 (West 2012) (providing limitations and guidelines on who has access to classified information at the NSA); id. § 3024(g) (holding the Director of National Intelligence
accountable for safeguarding intelligence information from disclosure); id. § 3161 (governing the process of classifying information and accessing classified information); id. § 3365
(limiting the dissemination of privileged information); id. § 3121 (punishing individuals
who reveal the identity of undercover agents and classified information); and id. § 3142
(allowing operational files of the National Geospatial-Intelligence Agency to be kept secret
from the public). See also Sudha Setty, The Rise of National Security Secrets, 44 CONN. L.
REV. 1563, 1583 (2012) (discussing the dangers of allowing “invocations of secrecy to
go unchecked”).
888
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
at the top-secret level alone.24 The number of agencies and employees
working on merely “secret” level programs is surely much larger.25
Amid these agencies’ secret rulemaking lie the great “unknown unknowns” of the administrative state.26
We can learn a great deal about dysfunctions in national security
rulemaking when leaks reveal programs like the NSA’s. But we can
learn even more from national security rulemaking that the public
knows about but cannot participate in. I call this opaque rulemaking.27 It accounts for a larger share of national security rulemaking
than secret rulemaking and is much broader in scope. It includes, for
example, regulations implementing treaties28 and altering the legal
rights of immigrants.29
Opaque rulemaking occurs when an agency makes legislative
rules available to the public through some means, but refuses to conduct notice-and-comment or explain why.30 The authority to bypass
notice-and-comment this way comes from a little-understood but
broadly interpreted exception in the APA for “foreign affairs or military functions.”31 I call this the national security exception.32
24. See Dana Priest & William M. Arkin, Top Secret America: The Rise of the New
American Security State 86 (2011).
25. See id. at 86-87.
26. See Donald H. Rumsfeld, Sec’y of Def., Department of Defense News Briefing (Feb.
12, 2002), available at http://www.defense.gov/transcripts/transcript.aspx?transcript
id=2636. As David Posen has explained, Rumsfeld’s (infamous) taxonomy aptly describes
“deep secrets”—those about which Americans “are in the dark about the fact that they are
being kept in the dark.” David E. Pozen, Deep Secrecy, 62 STAN. L. REV. 257, 260 (2010).
27. Opaque rulemaking creates what David Pozen calls “shallow” secrets: the public
knows about the secret’s existence only from the publicly available final rule. See Pozen,
supra note 26, at 260.
28. See Int’l Bhd. of Teamsters v. Peña, 17 F.3d 1478, 1486-87 (D.C. Cir. 1994) (holding that regulations made pursuant to NAFTA exempting Mexican truck drivers from Federal Highway Administration (FWHA) licensing guidelines could be promulgated without
notice-and-comment).
29. See Rajah v. Mukasey, 544 F.3d 427, 436-38 (2d Cir. 2008) (upholding, under the
APA exception, the DOJ’s failure to use notice-and-comment to determine the countries
whose citizens would be required to report to the FBI under a post-September 11 registration
program, and who could be detained or deported in secret for immigration law violations).
30. Such rules can be published in the Code of Federal Regulations, the Federal Register, or simply online. Secret rulemaking becomes opaque when the content of the rules is
made available, through leaks or deliberate disclosure, to the public.
31. See 5 U.S.C. § 553(a)(1) (2012) (stating that notice-and-comment requirements
apply “except to the extent that there is involved a military or foreign affairs function of
the United States”).
32. See infra Part II.C. The term “national security” lacks a precise definition, but it
seems to be an ever-expanding concept. The Department of Defense recently defined it as
“[a] collective term encompassing both national defense and foreign relations of the United
States.” DEP’T OF DEF., JOINT PUBLICATION 1-02, DEPARTMENT OF DEFENSE DICTIONARY OF
MILITARY AND ASSOCIATED TERMS 182 (2010) (as amended through Jan. 15, 2014). I use the
term here in an even broader sense—to include these subjects and any action that may fall
under the exception in 5 U.S.C. § 553(a)(1).
2014]
NATIONAL SECURITY RULEMAKING
889
While some scholars regard the exception as inevitable,33 and others have proposed reforms to increase transparency,34 the full scope
and historical origins of national security rulemaking—both secret
and opaque—remain largely unexplored. Secrecy makes a complete
accounting impossible, but this Article uncovers two unique features
that expose the constitutionally problematic way this rulemaking is
conducted and point the way toward necessary reforms.
First, this Article considers the entire corpus of cases discussing
the APA exception, concluding that courts generally apply it even
when it was not invoked by the agency until the rule was challenged
in court.35 This makes it difficult to estimate how often agencies actually rely on the exception to avoid notice-and-comment.36 It also
makes court oversight of national security rulemaking far weaker.
Under the Chenery rule, courts will uphold agency rules only on the
grounds articulated by the agency when the rule was developed.37
Chenery I enables courts to perform their constitutionally critical role
of ensuring that agencies do not exceed their delegated authority.38
33. See Major Thomas R. Folk, The Administrative Procedure Act and the Military
Departments, 108 MIL. L. REV. 135, 142 (1985) (observing with approval that “[s]everal
court decisions have . . . given the term ‘military function’ its broadest possible definition”);
Adrian Vermeule, Our Schmittian Administrative Law, 122 HARV. L. REV. 1095, 1096
(2009) (briefly discussing the exception as one of many “black holes” in U.S. administrative
law and noting that courts have generally construed it broadly).
34. Bonfield, supra note 22, at 238 (conceding that, while a narrow construction would
be preferable, the language of the “military or foreign affairs function” exception is nevertheless “very broad”); Evan J. Criddle, When Delegation Begets Domination: Due Process of
Administrative Lawmaking, 46 GA. L. REV. 117, 192-93, 204 (2011) (proposing that the
exception be eliminated and that agencies use the APA’s “good cause” exception when notice-and-comment is inappropriate); Oona A. Hathaway, Presidential Power over International Law: Restoring the Balance, 119 YALE L.J. 140, 223-24, 262-63 (2009) (noting that
the exception relieves the President of the responsibility for conducting notice-andcomment for international agreements, and proposing reforms); C. Jeffrey Tibbels, Delineating the Foreign Affairs Function in the Age of Globalization, 23 SUFFOLK TRANSNAT’L L.
REV. 389 (1999) (contending that agencies use the exception too often for economic regulation, and proposing that the courts rein in its use); see also William D. Araiza, Note, Noticeand-Comment Rights for Administrative Decisions Affecting International Trade: Heightened Need, No Response, 99 YALE L.J. 669, 671 (1989) (proposing a statutory grant of notice-and-comment for rulemaking on the classification of imported goods, which is subject
to the exception).
35. See infra notes 161-56 and accompanying text.
36. Over the decades, the exception has been invoked by agencies across the government—including the Departments of Defense, Homeland Security, State, Commerce,
Treasury, Energy, Transportation, and Agriculture; the Federal Communications Commission; the Food and Drug Administration; the Nuclear Regulatory Commission; and even
the Postal Service. See Bonfield, supra note 22, at 232 n.38; see also id. at 232-34 (describing various agencies’ responses to a survey, stating they had relied on the exception and
concluding that “[i]n practice . . . most agencies do not usually exercise their discretion to
follow the [notice-and-comment] requirements . . . when they are not bound to do so because” the exception applies); infra Part II.C.
37. See SEC v. Chenery Corp. (Chenery I), 318 U.S. 80, 94-95 (1943); infra Part II.B.
38. See, e.g., Kevin M. Stack, The Constitutional Foundations of Chenery, 116 YALE
L.J. 952, 958-59 (2007); infra Part II.B.
890
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
When the agency does not use notice-and-comment, however—or offer any justification for departing from it—there are no legal or factual justifications for the court to review.
Second, this Article examines, for the first time, the exception’s
development through numerous drafts of APA predecessor legislation
over eighteen years and its historical and political context.39 This
analysis reveals that the national security exception was not a simple
inevitability; instead, it emerged from a fierce debate during the 1930s
and 1940s about how the United States could best compete with fascist
states without succumbing to fascism itself.40 When the APA exception
is seen as a product of its time rather than an essential attribute of all
administrative law systems, its purpose becomes clearer.
The Article proceeds in four parts. I focus on the national security
exception in Part II, revealing the general operation of the national
security administrative state. Notice-and-comment rulemaking is
“the dominant mode of administrative action.”41 Because notice-andcomment rulemaking is such an important means by which agencies
create rules with the force of law, exceptions from it deserve special
attention. By carving out the exception, the APA creates a distinct
and constitutionally suspect administrative law regime for national
security rulemaking.
As Part III explains, the limited scholarly debate over the APA
exception has occurred without exploring its origins. Professor Adrian Vermeule, adopting the theories of German political theorist Carl
Schmitt, labeled the exception one of the “black holes” permitting untrammeled executive discretion, which are “integral” and “inevitable”
in “a massive and massively diverse administrative state.”42 Vermuele’s critics have argued that such black holes are not necessary
and have proposed eliminating them.43 But Vermeule’s invocation of
the APA exception fits neatly into discussion about emergency governance—quite often grappling with Schmitt’s perspective—that has
dominated national security law scholarship at least since September
39. See infra Part IV. Scholars discussing the exception have examined the legislative
history of the bill that became the APA, but not the rich history of earlier reform proposals
or the historical context. See, e.g., Tibbels, supra note 34, at 395-96 (discussing the APA
legislative history and noting its sparseness on the subject of the national security exception).
40. See infra Part IV. For a history of the New Deal in the context of the rise of totalitarian states around the world, see IRA KATZNELSON, FEAR ITSELF: THE NEW DEAL AND THE
ORIGINS OF OUR TIME (2013).
41. Nina A. Mendelson, Rulemaking, Democracy, and Torrents of E-Mail, 79 GEO.
WASH. L. REV. 1343, 1343 (2011).
42. Vermeule, supra note 33, at 1149.
43. For criticism of Vermeule’s observations, see, for example, Criddle, supra note 34,
at 192, 193, 204. See also infra Part III.
2014]
NATIONAL SECURITY RULEMAKING
891
11, 2001.44 This Article, in Part IV, re-orients that discussion by focusing on the specific geopolitical concerns and theories of governmental organization animating the exception’s inclusion in the APA.
In Part V, this Article explores why national security rulemaking
must be reformed and how to do so. Quite simply, the world has
changed in fundamental ways since the APA was enacted. The distinction between the foreign and the domestic—and between what is
and is not “national security”—has faded. Because people and products cross boundaries as never before, national security concerns continue to expand to new areas of government policymaking.45 And the
threats America faces have changed dramatically as well. Our most
dangerous enemies are no longer nation states, but terrorist organizations or lone wolves.46 And in an era of high-tech global surveillance where a mere search algorithm can determine the difference
between what is foreign and domestic, the legitimating and deliberation-enhancing qualities of notice-and-comment rulemaking are as
necessary for national security agency action as for any other.47
Part V concludes by proposing reforms to an area of regulation
badly in need of a balance between secrecy and greater scrutiny and
public participation. If eliminating the APA exception entirely is not
feasible, the President should issue an executive order requiring
agencies to use notice-and-comment whenever possible. In addition,
the courts should be directed to impose a Chenery-type rule requiring
agencies to invoke the exception specifically when they issue a rule
without notice-and-comment and to specifically articulate their reasons for relying on the exception.48 Even if these justifications cannot
be disclosed to the public for some time, or ever, the requirement will
nonetheless discipline and improve agency decisionmaking.
44. See MARY L. DUDZIAK, WAR TIME: AN IDEA, ITS HISTORY, ITS CONSEQUENCES 11516 (2012) (discussing the surge of interest in, and citation to, Schmitt following September
11). Even before September 11, scholarly attention to the jurisprudence of emergency and
exception—and the work of German political scientist Carl Schmitt in particular—was
steadily increasing. See Oren Gross, The Normless and Exceptionless Exception: Carl
Schmitt’s Theory of Emergency Powers and the “Norm-Exception” Dichotomy, 21 CARDOZO L.
REV. 1825, 1825-26 (2000) (noting the revival of interest in Schmitt’s work in the late 1990s).
45. See, e.g., William C. Banks, Programmatic Surveillance and FISA: Of Needles in
Haystacks, 88 TEX. L. REV. 1633, 1634 (2010) (“[M]ore Americans than ever are engaged in
international communications, and there is far greater intelligence interest in communications to and from Americans. Both circumstances increase the likelihood that the government will be intercepting communications of innocent Americans . . . .”).
46. See infra note 315 and accompanying text.
47. See infra note 314 and accompanying text.
48. See SEC v. Chenery Corp. (Chenery I), 318 U.S. 80, 94-95 (1943); infra notes 109115 and accompanying text.
892
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
II. OUR BIFURCATED ADMINISTRATIVE STATE
The United States government has a bifurcated administrative
state. There is an ordinary administrative state, in which agencies
must solicit and consider public comments before issuing rules with
the force of law. And there is a national security administrative state,
in which agencies may choose to issue the same sort of rules without
first publishing them and without soliciting or receiving public comments, while some rules may be kept entirely secret. The two administrative states co-exist within most agencies, but the national security administrative state is more pervasive in the Departments of State
and Defense and in the array of intelligence agencies that operate
largely away from the public eye.49
Sometimes the two administrative states will co-exist within the
same rulemaking process. Inevitably, however, in these instances it
is the national security administrative state that governs the most
important aspects of the rulemaking. After the September 11 attacks,
the Department of Justice (DOJ) established a Special Call-In Registration Program “requir[ing] non-immigrant alien males over the age
of 16 from designated countries to appear for registration and fingerprinting.”50 The DOJ followed the APA’s notice-and-comment requirements by publishing the proposed rule establishing the program
and accepting public comments before issuing the final rule two
months later.51 But the DOJ did not publish for public comment a
crucial portion of the rule, which designates the countries whose citizens must report.52 Nor did the DOJ explain why it had decided not
to consider comments on this portion.53
49. Secret rulemaking also accounts for a significant, and troubling, portion of the
national security administrative state. See PRIEST & ARKIN, supra note 24, at 86-87; supra
notes 22-26 and accompanying text.
50. Rajah v. Mukasey, 544 F.3d 427, 432 (2d Cir. 2008).
51. See Registration and Monitoring of Certain Nonimmigrants, 67 Fed. Reg. 40,581
(proposed June 13, 2002) (to be codified at 8 C.F.R. pts. 214, 264). The proposed rule received only fourteen comments, many of which were similar. See Registration and Monitoring of Certain Nonimmigrants, 67 Fed. Reg. 52,584 (Aug. 12, 2002) (to be codified at 8
C.F.R. pts. 214, 264).
52. See Registration and Monitoring of Certain Nonimmigrants, 67 Fed. Reg. at
40,581 (stating that the selected countries whose citizens must report will be specified separately in notices published in the Federal Register); Registration and Monitoring of Certain Nonimmigrants, 67 Fed. Reg. at 52,584 (affirming this procedure in the final rule).
53. See Registration and Monitoring of Certain Nonimmigrants, 67 Fed. Reg. 52,584,
52,589 (Aug. 12, 2002) (stating, in response to a comment that the future designation of
countries “would be antithetical to the relationship between the United States and that
country and its citizens,” that “[t]he listing of countries from which nonimmigrant aliens
will be subject to special registration is determined by the Attorney General in consultation
with the Secretary of State, thereby ensuring that foreign policy implications will be considered when evaluating the possible designation of any specific country” and that the
“comment is outside the scope of this final rule”).
2014]
NATIONAL SECURITY RULEMAKING
893
Three months later, the DOJ began publishing a series of notices
in the Federal Register announcing the twenty-five designated countries—which included almost exclusively Muslim-majority nations
such as Morocco, Iran, Iraq, Libya, Bahrain, and Tunisia.54 Thousands of nonimmigrant alien visitors who reported to the FBI under
this program were interrogated and deported, many of them in secret, and hundreds were detained for significant periods of time.55
The Second Circuit upheld the program as lawful and explained
why the DOJ was justified in invoking the national security exception. Providing justification where the DOJ had offered almost
none,56 the court explained that, in the process of responding to comments, the DOJ might have been required to reveal sensitive national security information in justifying its choice of countries, which
would damage relations with those countries, slow down the designation process, and diminish the nation’s ability to collect intelligence
and prevent terrorist attacks.57
These were legitimate concerns—at least in the abstract. But in
practice, the registration program is widely considered to have been a
failure.58 Like many of the post-September 11 detention and removal
policies, it had troubling implications for due process and equal protection.59 Moreover, in targeting particular religious and ethnic com54. E.g., id. (designating Iran, Iraq, Libya, Sudan, and Syria as countries whose citizens meeting the conditions must report); Registration of Certain Nonimmigrant Aliens
from Designated Countries, 67 Fed. Reg. 70,526 (Nov. 22, 2002) (designating Afghanistan,
Algeria, Bahrain, Eritrea, Lebanon, Morocco, North Korea, Oman, Qatar, Somalia, Tunisia,
United Arab Emirates, and Yemen). In a yet later notice, three more countries—Armenia,
Pakistan, and Saudi Arabia—were added. Registration of Certain Nonimmigrant Aliens
from Designated Countries, 67 Fed. Reg. 77,136 (Dec. 16, 2002). A subsequent notice removed Armenia from the list and modified the registration requirements. See Registration of
Certain Nonimmigrant Aliens from Designated Countries, 67 Fed. Reg. 77,642 (Dec. 18, 2002).
55. See, e.g., Kevin R. Johnson & Bernard Trujillo, Immigration Reform, National
Security After September 11, and the Future of North American Integration, 91 MINN. L.
REV. 1369, 1383-84 (2007); Sadiq Reza, A Trap for Middle Eastern Visitors, WASH. POST,
Jan. 10, 2003, at A21. The program was only part of the constitutionally questionable detention of thousands of Muslims in the United States following the September 11 attacks.
See DAVID COLE, ENEMY ALIENS: DOUBLE STANDARDS AND CONSTITUTIONAL FREEDOMS IN
THE WAR ON TERRORISM 22-46(2003); Johnson & Trujillo, supra, at 1382-84.
56. When publishing the notice listing designated countries, the DOJ stated that
these determinations were justified “[i]n light of recent events, and based on intelligence
information available to the Attorney General.” Registration of Certain Nonimmigrant
Aliens from Designated Countries, 67 Fed. Reg. at 67,766. On appeal, the DOJ argued that
“[p]ublic commentary . . . would serve no other purpose, and achieve no other end, than
potentially inflaming relations” with the designated countries. Brief for Respondents at 37,
Rajah v. Mukasey, 544 F.3d 449 (2d Cir. 2008) (No. 06-3493), 2007 WL 6449014.
57. Rajah v. Mukasey, 544 F.3d 427, 437 (2d Cir. 2008).
58. Johnson & Trujillo, supra note 55, at 1384.
59. See, e.g., David Cole, Enemy Aliens, 54 STAN. L. REV. 953, 957 (2002); cf. Hiroshi
Motomura, Immigration and We the People After September 11, 66 ALB. L. REV. 413, 42021 (2003) (disagreeing with Cole’s substantive arguments but agreeing that there are due
process and equal protection concerns); see also Ty S. Wahab Twibell, The Road to Intern-
894
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
munities and casting too broad a net, such policies alienated the very
people who were most likely to provide the government with useful
information for preventing future terrorist attacks.60 The resulting
arrests, detentions, and deportations prompted protests.61 The DOJ
never claimed that the program thwarted any terrorist activity, and
it was abandoned a year later.62
The DOJ’s detention program demonstrates why the failure to involve the public in rulemaking through notice-and-comment increases the likelihood of impracticable, ill-conceived agency action that
results in inefficiency, abuses, and backlash. Notice-and-comment
was designed to avoid such harmful effects and to root policy-making
in the constitutional values of public participation, transparency, legitimacy, and governmental effectiveness.
A. The Importance of Notice-and-Comment Rulemaking
The difference between our two administrative regimes in the way
they are permitted to undertake rulemaking is not merely procedural; it is fundamental. Notice-and-comment rulemaking is a pillar of
the administrative law regime that arose after World War II. The
APA, the “constitution of the administrative state,” established and
codified the notice-and-comment procedure for “legislative” or “substantive” rules—those that are legally binding on agencies and the
public.63 With the explosive growth of the federal government during
the 1930s and 1940s, lawmakers realized that “Congress cannot
ment: Special Registration and Other Human Rights Violations of Arabs and Muslims in
the United States, 29 VT. L. REV. 407, 411-19 (2005).
60. Cf. Tom R. Tyler, Stephen Schulhofer & Aziz Z. Huq, Legitimacy and Deterrence
Effects in Counterterrorism Policing: A Study of Muslim Americans, 44 LAW & SOC’Y REV.
365 (2010) (examining social science data regarding the Muslim community in New York
City and concluding that perceptions of governmental legitimacy are the most important
incentive to cooperate with counterterrorism efforts).
61. Johnson & Trujillo, supra note 55, at 1384 (citing Emily Bazar, New Battle on
Civil Rights Front—The Registration of Some Immigrants Is Denounced as Racial Profiling, SACRAMENTO BEE, Jan. 20, 2003, at A1; Wyatt Buchanan, Hundreds Protest INS Registration/Men from 13 Countries Sign In, S.F. CHRON., Jan. 11, 2003, at A13).
62. Id.; see Suspending the 30-Day and Annual Interview Requirements from the
Special Registration Process for Certain Nonimmigrants, 68 Fed. Reg. 67,578 (Dec. 2,
2003); see also Removing Designated Countries from the National Security Entry-Exit
Registration System (NSEERS), 76 Fed. Reg. 23,830 (Apr. 28, 2011) (formally finalizing
the abandonment of the program). Under US-VISIT, most visitors are subject to the collection of biometric data upon entry at air- and seaports. See Implementation of the United
States Visitor and Immigrant Status Indicator Technology Program (“US-VISIT”); Biometric Requirements, 69 Fed. Reg. 468 (Jan. 5, 2004).
63. See Hickman, supra note 17, at 481-82 (discussing the factors courts use to determine whether a rule is legislative). Because agencies have incentives to avoid notice-andcomment by labeling rules as merely “interpretive,” even if they really are legislative,
many courts do not accept agencies’ characterizations at face value. See id.; William Funk,
Legislating for Nonlegislative Rules, 56 ADMIN. L. REV. 1023, 1029-30 (2004).
2014]
NATIONAL SECURITY RULEMAKING
895
manage the regulatory demands of the modern [administrative] state
on its own.”64 Notice-and-comment rulemaking was one compromise
under which Congress could delegate to agencies the power to issue
legally binding rules without abandoning oversight of those agencies.
The APA also technically requires more robust, “formal,” trial-like
rulemaking procedures in some circumstances when the agency’s enabling act or another statute requires rulemaking “to be made on the
record after opportunity for an agency hearing.”65 However, courts
have generally given agencies broad discretion in their choice of procedures and have rarely interpreted statutes as requiring agencies to
engage in formal rulemaking.66 By the 1970s, with the courts’ blessing, notice-and-comment became the norm for agency rulemaking.67
In the twenty-first century, it remains the primary means for the
resolution of policy matters by agencies and continues to enjoy the
strong endorsement of administrative law scholars.68
Under its notice-and-comment procedures, the APA requires
agencies to publish a proposed rule in the Federal Register and provide the public with an opportunity to respond during a fixed period,
which may be as long as 180 days, depending on the rule’s complexity.69 When issuing a final rule that implements a congressional mandate, agencies must provide “a concise general statement” of the proposed regulation’s “basis and purpose.”70 The process requires agencies to “specifically explain their policy choices, their consideration of
important aspects of the problem, and their reasons for not pursuing
viable alternatives.”71
Publication of a proposed rule begins a dialogue between policymakers and those affected by a regulation; interested parties must
64. See Hathaway, supra note 34, at 243.
65. 5 U.S.C. § 553(c) (2012).
66. SEC v. Chenery Corp. (Chenery II), 332 U.S. 194, 203 (1947) (“The choice made
between proceeding by general rule or by individual, ad hoc litigation is one that lies primarily in the informed discretion of the administrative agency.”); Edward Rubin, It’s Time
to Make the Administrative Procedure Act Administrative, 89 CORNELL L. REV. 95, 107
(2003) (“Because the impracticalities of formal rulemaking are well known, Congress rarely
requires this technique, and courts avoid interpreting statutes to require it, even in the
rare cases where the statute seems to do so.”).
67. See Hickman, supra note 17, at 476-77; Thomas W. Merrill & Kathryn Tongue
Watts, Agency Rules with the Force of Law: The Original Convention, 116 HARV. L. REV.
467, 546-59 (2002).
68. See Mariano-Florentino Cuéllar, Rethinking Regulatory Democracy, 57 ADMIN. L.
REV. 411, 428-29 (2005) (“Notice and comment deserves close attention in a study of regulatory democracy because the bulk of regulation is crafted through that procedure today . . . .”);
Mendelson, supra note 41, at 1380. But see, e.g., Richard J. Pierce, Jr., Response, Presidential Control Is Better Than the Alternatives, 88 TEX. L. REV. SEE ALSO 113, 116-17 (2010)
(describing notice-and-comment as too burdensome).
69. 5 U.S.C. § 553(b).
70. Id. § 553(c).
71. Stack, supra note 38, at 972.
896
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
have the opportunity to comment on the proposed regulation, and the
agency must respond to the comments.72 Courts enforce these requirements through “hard look” review, invalidating regulations that
they find to be “arbitrary and capricious,” and upholding regulations
that are well reasoned and well supported by facts.73
The notice-and-comment process has been widely praised for improving agency decisionmaking from a number of different perspectives, many of which are staples of administrative law. No matter
which of the major theories of agency decisionmaking one adopts, the
notice-and-comment process arguably adds value. Even from a purely
“expertocratic” perspective,74 in which the public’s contribution to decisionmaking is considered less important, the notice requirement
arguably results in better policy because it disciplines agency decisionmaking from the very beginning. It provides incentives for agency employees to create an accurate record, to communicate with one
another clearly, to involve the entire agency in the process, and to
identify and attempt to resolve critical issues before issuing notice of
the proposed rule.75
For those who believe that the involvement of the public is crucial
for rulemaking to result in good policies, the benefits of notice-andcomment are more obvious. A public choice or pluralist view regards
72. See 5 U.S.C. § 553(c) (enumerating the notice-and-comment requirements); United
States v. Nova Scotia Food Prods. Corp., 568 F.2d 240, 253 (2d Cir. 1977) (requiring response to significant comments); Portland Cement Ass’n v. Ruckelshaus, 486 F.2d 375,
392-93 (D.C. Cir. 1973) (requiring disclosure of “technical data or studies” in time to allow
for meaningful comment).
73. Judicial review of agency action is a cornerstone of the APA and necessary to enforce the notice-and-comment requirements. The APA requires reviewing courts to “hold
unlawful and set aside agency action, findings, and conclusions found to be . . . arbitrary,
capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. §
706(2). The Supreme Court has interpreted “arbitrary and capricious” review as requiring
courts to review the record and “satisfy[] themselves that the agency has made a reasoned
decision based on its evaluation of the significance—or lack of significance—of . . . information.” Marsh v. Or. Natural Res. Council, 490 U.S. 360, 378 (1989). “Hard look” review
describes the way the courts enforce the “arbitrary and capricious” standard, so the terms
are usually considered interchangeable. See Kathryn A. Watts, Proposing a Place for Politics in Arbitrary and Capricious Review, 119 YALE L.J. 2, 5 (2009).
74. Jim Rossi, Participation Run Amok: The Costs of Mass Participation for Deliberative Agency Decisionmaking, 92 NW. U. L. REV. 173, 179 (1997) (“Expertocratic models view
the decision process primarily as an exercise in scientific validation by virtue of the method
and culture of the agency experts’ profession.”).
75. See William F. Pedersen, Jr., Formal Records and Informal Rulemaking, 85 YALE
L.J. 38, 59-60 (1975); Mark Seidenfeld, Demystifying Deossification: Rethinking Recent
Proposals to Modify Judicial Review of Notice and Comment Rulemaking, 75 TEX. L. REV.
483, 506-10 (1997); Matthew C. Stephenson, A Costly Signaling Theory of “Hard Look”
Judicial Review, 58 ADMIN. L. REV. 753, 755 (2006) (contending that judicial review of
agency action disciplines and improves the rulemaking process because “the expert government decisionmaker’s willingness to produce a high-quality explanation” for a rule “signals that the government believes the benefits of the proposed policy are high,” making
judicial approval more likely).
2014]
NATIONAL SECURITY RULEMAKING
897
regulation as the resolution of conflicts among competing interests,
while a public interest view holds that regulators can discern and
pursue the general good.76 But from either view, active participation
of the public matters immensely because rulemaking is an important
means by which the government resolves not just “technical” or “scientific” questions, but also questions of values.77 Selecting among different potential gains is a value judgment, as Cass Sunstein observed, and one that should be “made publicly and exposed to democratic view.”78 Moreover, public deliberation and participation can be
regarded as irreducible democratic values in themselves.79
Notice-and-comment rulemaking also strengthens legitimacy.
Agencies must be held accountable if their actions are to be seen as
legitimate. Accountability flows from meaningful constraints on
agency authority.80 Notice-and-comment is a powerful constraint: it
ensures that agency action is seen as legitimate because it provides
direct accountability to the public, which means that its choices will
more closely reflect the popular will. This can be true whether or not
the process is viewed from a public choice or public interest perspective. For public choice theorists, if the agency hears from and considers a wide range of interests, its decisions will be seen as more democratic: the agency serves as a broker among interest groups, just as
Congress serves as a broker for interest groups during the enactment
of legislation.81 From the (more optimistic) public interest perspec76. See Rossi, supra note 74, at 198-99.
77. See Emily Hammond Meazell, Super Deference, the Science Obsession, and Judicial Review as Translation of Agency Science, 109 MICH. L. REV. 733, 735-36 (2011) (contending that “agency science . . . is laced with policy decisions at numerous levels” and that
“interested parties and agencies alike are incentivized to cloak their policy choices in the
seemingly unassailable mantle of science”).
78. Cass R. Sunstein, Is the Clean Air Act Unconstitutional?, 98 MICH. L. REV. 303,
364 (1999).
79. See Robert W. Bennett, Democracy as Meaningful Conversation, 14 CONST.
COMMENT. 481, 481-82 (1997) (advancing a “conversational” model of democracy, “under
which the citizenry is engaged by ongoing public conversation about public policy,” and
arguing that “it is this engagement that is the stabilizing force in the system”); Rossi,
supra note 74, at 179.
80. See Criddle, supra note 34, at 159 (“[W]hen Congress delegates lawmaking authority to administrative agencies, structural due process requires that agency lawmakers be
subject to meaningful political accountability and that persons adversely affected by agency action have an opportunity to test the constitutional adequacy of Congress’s delegation
through judicial review.”); Edward Rubin, The Myth of Accountability and the AntiAdministrative Impulse, 103 MICH. L. REV. 2073, 2075 (2005) (“[T]rue accountability, in the
realm of law and politics, involves many of the features that are central to the administrative state and that people find so unattractive about it—hierarchy, monitoring, reporting,
internal rules, investigations, and job evaluations.”).
81. Richard B. Stewart, The Reformation of American Administrative Law, 88
HARV. L. REV. 1667, 1683 (1975); id. at 1670 (“Increasingly, the function of administrative law is not the protection of private autonomy but the provision of a surrogate political process to ensure the fair representation of a wide range of affected interests in the
process of administrative decision.”).
898
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
tive, the agency rulemaking process is a forum for democratic deliberation, in which agencies and citizens alike will change their views
in response to reasoning of others.82 During the notice-and-comment
rulemaking process, both citizens (including interest groups) and the
agency engage in dialogue. Even if the dialogue is limited by the
brevity of the comment period, the agency staff can act as agents of
constituent stakeholders.83
In addition to its role in facilitating direct accountability to the
public, notice-and-comment rulemaking can support and enhance
congressional oversight of agency action. It is true that, if an agency
thwarts the public interest, the public can contact their congressional
representatives and exert pressure through the vote. But this process
is a lengthy and frustrating one: Congress’s attention span is notoriously short, and ordinary congressional oversight through hearings,
confirmation decisions, and appropriations is limited, infrequent, and
ad hoc.84 This sort of congressional “walking the beat” is labeled “police-patrol” oversight by Mathew McCubbins and Thomas Schwartz,
who criticize it as overlooking many important agency problems.85
Notice-and-comment rulemaking can help address this gap by enabling other informal, but complementary, oversight processes. If a
proposed regulation provokes significant controversy, Congress is
likely to re-examine the delegation of authority to the agency and intervene in a more timely fashion. In other words, notice-and-comment
rulemaking permits the public to engage in what McCubbins and
Schwartz call “fire alarm” oversight.86 Notice-and-comment rulemaking—and subsequent judicial enforcement—“allows Congress to harness the power of private actors to enhance its oversight capacity.”87
82. Mark Seidenfeld, The Quixotic Quest for a “Unified” Theory of the Administrative
State, ISSUES IN LEGAL SCHOLARSHIP, Mar. 2005, at 1, 4-5, available at http://www.deg
ruyter.com/view/j/ils.2005.5.issue-1/ils.2005.5.1.1056/ils.2005.5.1.1056.xml?format=INT.
83. See id.
84. See, e.g., THOMAS MANN & NORMAN ORNSTEIN, THE BROKEN BRANCH: HOW
CONGRESS IS FAILING AMERICA AND HOW TO GET IT BACK ON TRACK 170 (2006) (lamenting
the infrequency of congressional oversight); Brian D. Feinstein, Avoiding Oversight: Legislator Preferences and Congressional Monitoring of the Administrative State, 8 J.L. ECON. &
POL’Y 23, 40-41 (2011) (concluding that members of Congress generally assign a low value
to service on oversight subcomittees); David E. Lewis & Terry M. Moe, The Presidency and
the Bureaucracy: The Levers of Presidential Control, in THE PRESIDENCY AND THE
POLITICAL SYSTEM 374, 382-85 (Michael Nelson ed., 10th ed. 2014) (arguing that reelection-oriented legislators have scant incentive to participate in oversight activities).
85. See Mathew D. McCubbins & Thomas Schwartz, Congressional Oversight Overlooked: Police Patrols Versus Fire Alarms, 28 AM. J. POL. SCI. 165, 166 (1984).
86. See id.
87. Hathaway, supra note 34, at 247 n.320; see also Kenneth A. Bamberger, Regulation as Delegation: Private Firms, Decisionmaking, and Accountability in the Administrative State, 56 DUKE L.J. 377, 399-408 (2006) (discussing the “administrative accountability
paradigm”); Lisa Schultz Bressman, Procedures as Politics in Administrative Law, 107
2014]
NATIONAL SECURITY RULEMAKING
899
Although it enjoys broad support, notice-and-comment is certainly
not without its critics. Notice-and-comment is a time and resourceintensive process. Indeed, the substantial literature on “agency ossification” largely blames notice-and-comment requirements for agency
sluggishness and resistance to change.88 Richard Pierce summarizes
this critique:
[I]t takes a long time to issue a rule; agencies never issue many of
the rules that would be beneficial to the public; agencies maximize
their use of procedural alternatives that are inferior to rulemaking
to avoid the delay and cost of the notice-and-comment process; and
agencies often decline to amend or to rescind rules that have
become obsolete.89
These problems may account for a noticeable trend by Congress toward forcing agencies to act without notice-and-comment, either by
mandating the issuance of particular rules or by imposing deadlines
that make the procedure impossible.90
Nonetheless, notice-and-comment remains the benchmark. And
despite its drawbacks, it is the least-worst means for achieving better
policy outcomes and keeping agencies accountable.
B. The Constitutional Dimension of Notice-and-Comment
The benefits of notice-and-comment rulemaking in advancing
democratic values are often celebrated. But it has a constitutional dimension as well. Congress designed it to be a means through which
agencies are constrained from exceeding their delegated authority and
thereby violating separation of powers principles or individual rights.
The Constitution, by vesting “[a]ll legislative [p]owers herein
granted” in Congress, limits Congress’s ability to delegate that power
to the executive branch or independent agencies.91 The nondelegation doctrine requires Congress to “ ‘lay down by legislative act
an intelligible principle to which the [agency] is directed to conform.’ ”92
COLUM. L. REV. 1749, 1770 (2007) (“[C]ourts force agencies to comply with the procedures
that facilitate fire-alarm oversight.”).
88. See Pierce, supra note 68, at 116-17 (summarizing the literature on ossification in
agency rulemaking); see also STEPHEN BREYER, BREAKING THE VICIOUS CIRCLE: TOWARD
EFFECTIVE RISK REGULATION 49 (1993); Thomas O. McGarity, Response, The Courts and
the Ossification of Rulemaking: A Response to Professor Seidenfeld, 75 TEX. L. REV. 525,
525-29 (1997).
89. Pierce, supra note 68, at 117.
90. See infra notes 151-153 and accompanying text (discussing use of the “good
cause” exception).
91. U.S. CONST. art. I, § 1.
92. Whitman v. Am. Trucking Ass’ns, 531 U.S. 457, 472 (2001) (quoting J.W. Hampton, Jr., & Co. v. United States, 276 U.S. 394, 409 (1928)). Under the Supreme Court’s current approach to the non-delegation doctrine, Congress cannot delegate any amount of its
own actual legislative power but may nonetheless delegate vast “rulemaking” authority to
900
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
Failure to do so violates constitutional separation of powers principles. In addition, the agency itself lacks the power to create an “intelligible principle” when Congress has failed to provide one.93
It is extremely rare for courts to invoke the non-delegation doctrine to strike down legislation as unconstitutional.94 Since 1935,
courts have bowed to the exigencies of the massive modern administrative state by permitting Congress to assign vast rulemaking authority to agencies.95 Instead, the courts have trod a different path
toward regularly enforcing constitutional separation of powers principles in administrative law. Through judicial review of agency action, courts play a critical role in ensuring that agencies act only
within the bounds of their constitutionally-valid delegated authority.96
In the APA, Congress codified this role for courts, and scholars
have come to recognize that role’s constitutional dimensions.97 But it
is a role that courts have been playing for as long as the administrative state has existed.98 The APA empowers courts to set aside agency
action that is “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.”99 Courts have further underscored
the fundamental importance, if not explicitly the constitutional dimensions of, APA review by gradually developing more stringent
standards for reviewing agency action than for review of most statutes.100 Congress originally understood the APA’s arbitrary and caagencies. See id. at 488-89 (Stevens, J., concurring) (disagreeing with the majority that the
properly delegated authority is not “legislative”).
93. Id. at 472 (majority opinion).
94. See, e.g., United States v. Whaley, 577 F.3d 254, 263 (5th Cir. 2009) (observing
that “the limits on delegation are frequently stated, but rarely invoked,” and that “the Supreme Court has not struck down a statute on nondelegation grounds since 1935” (citing
A.L.A. Schechter Poultry Corp. v. United States, 295 U.S. 495 (1935); Pan. Ref. Co. v.
Ryan, 293 U.S. 388 (1935)).
95. See Stack, supra note 38, at 956.
96. See id. at 958-59; see also Criddle, supra note 34, at 132. But cf. Eric A. Posner &
Adrian Vermeule, Interring the Nondelegation Doctrine, 69 U. CHI. L. REV. 1721, 1721
(2002) (“A statutory grant of authority to the executive branch or other agents never effects
a delegation of legislative power.”).
97. See generally Metzger, supra note 100 (describing the “deeply embedded practice”
of judges’ infus[ing] constitutional values into their development of administrative law
doctrines or readings of ordinary administrative requirements.”); Stack, supra note 38.
98. See, e.g., Mahler v. Eby, 264 U.S. 32, 44-45 (1924) (citing precedent that, if not
presented on the record, express findings by an agency could not “be supplied by implication” or by reference to litigation documents before the agency and concluding that such a
defect “goes to the existence of the power on which the proceeding rests” (citing Wichita
R.R. & Light Co. v. Pub. Utils. Comm’n, 260 U.S. 48, 59 (1922)); see also Stack, supra note
38, at 983-89 (discussing Mahler, Wichita Railroad, and similar contemporaneous decisions
holding that the limitation derived from constitutional concerns).
99. Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 41
(1983) (quoting Administrative Procedure Act, 5 U.S.C. § 706(2)(A)).
100. Gillian E. Metzger, Ordinary Administrative Law as Constitutional Common Law,
110 COLUM. L. REV. 479, 491 (2010).
2014]
NATIONAL SECURITY RULEMAKING
901
pricious standard as equivalent to the “minimal scrutiny [of] constitutional rationality review.”101 Over the years, however, the courts’
interpretations of agency action have diverged from their interpretation of statutes. The APA’s arbitrary and capricious standard became
gradually more rigorous, evolving into today’s “hard look” review.102
As Gillian Metzger has recounted, this divergence was driven largely
by constitutional concerns raised by the expansion of regulatory authority during the 1960s and 1970s with fears of agency capture by
industry.103 With separation of powers principles at stake, courts
came to view agency action with a more skeptical eye.
Of course, some may reject the idea that hard look review is a fullblown constitutional requirement.104 It may be merely a prudential
doctrine or occupy some quasi-constitutional twilight zone.105 But it is
difficult to imagine that our massive modern administrative state
could function in a constitutionally sound manner without the sort of
judicial scrutiny that hard look review provides. As courts have come
to interpret it, the APA is, therefore, not merely a constitution for the
administrative state;106 it provides essential constitutional validity
for the administrative state.
However, unless an agency is required to state its reasoning for
acting while it takes action, and unless it actually does so, reviewing
courts cannot perform their constitutionally critical duty to discern
whether the agency acted for lawful reasons.107 If hard look review
has constitutional dimensions, the requirement of contemporaneous
101. Id.
102. Id.; Watts, supra note 73, at 15-16.
103. Metzger, supra note 100, at 491.
104. See, e.g., Note, Rationalizing Hard Look Review After the Fact, 122 HARV. L. REV.
1909, 1921 (2009) (arguing that eliminating the Chenery rule and permitting courts to
consider post-hoc rationales for agency action would be constitutional and would help ameliorate ossification of agency decisionmaking).
105. Scholars and jurists have not only likened the APA to a constitution, but they also
see it as having quasi-constitutional status within the framework of American law. See,
e.g., Bruce Ackerman, The Emergency Constitution, 113 YALE L.J. 1029, 1077 (2004) (“Although it is packaged as a statute, the APA is the product of constitutional thought, and the
courts have given quasi-constitutional status to its provisions.”); Metzger, supra note 100,
at 484-85 (describing judicial review of administrative action as a form of constitutional
common law subject to congressional revision).
106. Steven P. Croley, The Administrative Procedure Act and Regulatory Reform: A
Reconciliation, 10 ADMIN. L.J. AM. U. 35, 35 (1996) (describing the APA as “the constitution
of the modern regulatory state”); see also Abbott Labs. v. Gardner, 387 U.S. 136, 140 (1967)
(characterizing the APA as a “seminal act” governing the relationship between courts
and agencies).
107. See, e.g., Olenhouse v. Commodity Credit Corp., 42 F.3d 1560, 1565 (10th Cir.
1994) (setting aside an agency decision because the court could not “ascertain” from a thorough review of the administrative record the basis and support for the decision). The court
explained that “the grounds upon which the agency acted must be clearly disclosed in, and
sustained by, the record,” and that “[t]he agency must make plain its course of inquiry, its
analysis and its reasoning.” Id. at 1575 (citation omitted).
902
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
reason-giving by the agency would therefore have constitutional
dimensions as well.108
Recognizing the fundamental importance of this requirement even
prior to the APA’s enactment, the Supreme Court established the
Chenery rule to test the validity of any agency’s exercise of discretion.109 In Chenery I, the Court held that discretionary administrative
action will only be upheld on grounds articulated by the agency in the
record.110 This is in marked contrast to judicial review of statutes:
unless strict scrutiny is required, courts may uphold legislation on
any constitutional basis—even one Congress never articulated.111 In
establishing the Chenery rule, therefore, the Court identified a key
difference between the way courts should evaluate legislation and the
way they should evaluate agency action.
The Chenery rule delimits the responsibilities and powers of agencies and courts in the judicial review process in a manner animated
by, and sensitive to, separation of powers concerns.112 It preserves in
agencies the formal authority to exercise the discretion delegated by
Congress, preventing courts from substituting their own policy for
that of the agency. At the same time, however, the rule limits agency
power to make policy away from public scrutiny.
The Chenery rule also helps promote integrated and cohesive decisionmaking within agencies. Knowing that courts will privilege those
agency rationales proffered at the time of the agency decision and in
the record—rather than often ad-hoc justifications by staff or counsel
off the record, before the agency process officially begins, or during
judicial review—senior agency officials have incentives to exercise
control over staff and counsel, focusing the agency’s attention on the
most persuasive, rational, and legally supportable justifications for
the action. Chenery I has helped make “explicit reason-giving a major
part of the industry of the administrative state.”113
For courts, in turn, “an agency’s contemporaneous explanation for
its decisions remains one of the most common grounds for judicial
reversal and remand.”114 As Kevin Stack has observed, courts treat
an agency’s failure to articulate the right reasons for its decision as a
108. Metzger, supra note 100, at 491-93; Stack, supra note 38, at 1003-04. But see Note,
Rationalizing Hard Look Review After the Fact, supra note 104, at 1921.
109. See SEC v. Chenery Corp. (Chenery I), 318 U.S. 80 (1943).
110. Id. at 87-88.
111. See, e.g., id. at 88; Helvering v. Gowran, 302 U.S. 238, 245 (1937) (stating the wellsettled principle that “if the decision below is correct, it must be affirmed, although the
lower court relied upon a wrong ground or gave a wrong reason”).
112. Stack, supra note 38, at 957-58.
113. Id. at 957.
114. Id.
2014]
NATIONAL SECURITY RULEMAKING
903
constitutional error.115 The Chenery rule helps enforce the nondelegation doctrine.
The notice-and-comment requirements work hand-in-hand with
Chenery I to ensure that agency rulemaking remains within constitutionally permissible bounds. That is why exceptions to notice-andcomment requirements are constitutionally problematic. Without a
requirement that agencies articulate and test their rationales during
the decisionmaking process, the courts may have difficulty discerning
which rationales the agency actually relied upon in making its decision. This significantly weakens the courts’ ability to enforce the
Chenery rule and with it, the non-delegation doctrine. Just as the notice-and-comment procedures have constitutional roots and a constitutional dimension, the exceptions to those procedures, too, must
have some other constitutional justification.
C. Secrecy and the National Security Exception
As I will discuss further below, the APA provides several exceptions enabling agencies to disregard the notice-and-comment requirements, despite their importance. But the APA establishes only
two substantive exceptions from notice-and-comment requirements.116
The first is an exception for “matter[s] relating to agency management or personnel or to public property, loans, grants, benefits, or
contracts.”117 The first portion of this exception, respecting internal
agency matters, is relatively easy to justify because these matters
affect the public only indirectly.118 But the second portion is controversial in its own right: rules regarding public property, loans,
grants, benefits, and contracts undoubtedly have the potential to
substantially affect millions of Americans directly. Courts and com-
115. Id. at 983-84.
116. Not all departures from, or additions to, notice-and-comment rulemaking standards are found in the APA. An agency’s organic statute or other procedural statute may
provide for more or fewer requirements. See, e.g., National Environmental Policy Act of
1969 (NEPA), 42 U.S.C. §§ 4321–4370h (2006) (mandating thorough scientific and systematic analysis of environmental problems in rulemaking and imposing procedural requirements to generate information for the agency and the public).
117. 5 U.S.C. § 553(a)(2) (2012).
118. Compare Tunik v. Merit Sys. Prot. Bd., 407 F.3d 1326, 1343-44 (Fed. Cir. 2005)
(holding that the exception did not apply to the purported repeal by Merit Systems Protection Board (MSPB) of a regulation governing removal of administrative law judges because
the rule implicated a broader interest of the public in having private rights adjudicated by
persons who had some independence from the agency that opposed them), with Favreau v.
United States, 317 F.3d 1346, 1359 (Fed. Cir. 2002) (holding that the exception applied to
memoranda detailing when the United States could seek recoupment of prepaid bonuses
from armed forces personnel).
904
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
mentators expressed puzzlement,119 and the Administrative Conference of the United States recommended that this portion of the exception be eliminated.120 Subsequent agency-specific statutes and
regulations have considerably constrained its scope.121
The second substantive exception is the national security exception, which, unlike its ill-fated twin, is alive and well. It is an expression of, if not the foundation of, the national security administrative
state. If the proposed rule involves a “foreign affairs” or “military”
function of the United States, the agency may disregard the APA’s
requirements for notice, comment, and delay in rule implementation.122
This is not the only way in which the framework of U.S. administrative law provides national security exceptions. Classification authority enables agencies to avoid notice-and-comment by keeping a
portion, or the entire existence, of rulemaking secret.123 And the APA
itself includes other exceptions for national security agency action.
Importantly, agencies are also exempt from the adjudication requirements when conducting a foreign affairs or military function.124
And although the APA usually applies to military departments and
agencies, it does not apply at all to military commissions, courts martial, or to “military authority exercised in the field in time of war or
in occupied territory.”125
Moreover, Executive Order 12,866, which establishes, over all U.S.
government agency rulemaking, centralized review by the Office of
119. Nat’l Wildlife Fed’n v. Snow, 561 F.2d 227, 231 (D.C. Cir. 1976) (“[T]he APA [contracts exception] does create a serious gap in the procedural protections the APA was
enacted to provide.”).
120. Arthur Earl Bonfield, Public Participation in Federal Rulemaking Relating to
Public Property, Loans, Grants, Benefits, or Contracts, 118 U. PA. L. REV. 540, 611 (1970).
121. See, e.g., Elimination of Certain Exceptions from the APA Rulemaking Requirements, 1 C.F.R. § 305.69-8 (1974) (detailing 5 U.S.C. § 553(a)(2) rules pertaining to “ ‘public
property, loans, grants, benefits, or contracts’ ”); see also United States v. AEY, Inc., 603 F.
Supp. 2d 1363, 1376 (S.D. Fla. 2009); Pierce, supra note 68, at 116. The exception is still
invoked, however, for a small percentage of rules. See U.S. GOV’T ACCOUNTABILITY OFFICE,
GAO-13-21, FEDERAL RULEMAKING: AGENCIES COULD TAKE ADDITIONAL STEPS TO RESPOND
TO PUBLIC COMMENTS 8, 15 (2012) [hereinafter GAO REPORT].
122. 5 U.S.C. § 553(a)(1).
123. For a list of the key statutes providing authority to classify information, see supra
note 12. The President also engages in national security rulemaking directly through a
variety of different types of executive orders, which are usually published in the Federal
Register, and National Security Decision Directives (which go by many different names),
which are usually not. See STEPHEN DYCUS, ET. AL, NATIONAL SECURITY LAW 40-41 (5th
ed. 2011).
124. See 5 U.S.C. § 554(a)(4). Discussion of the adjudication exception is beyond the scope of
this Article, but it raises many, if not most, of the same problems as the rulemaking exception.
125. 5 U.S.C. § 551(1)(F)–(G). For an analysis of the military authority exception and a
compelling argument that it should be narrowly construed, see Kathryn E. Kovacs, A History of the Military Authority Exception in the Administrative Procedure Act, 62 ADMIN. L.
REV. 673 (2010).
2014]
NATIONAL SECURITY RULEMAKING
905
Information and Regulatory Affairs (OIRA) in the Office of Management and Budget (OMB), echoes the APA by exempting “[r]egulations
or rules that pertain to a military or foreign affairs function of the
United States, other than procurement regulations and regulations
involving the import or export of non-defense articles and services.”126
By virtue of the exceptions in the APA and Executive Order 12,866,
national security rulemaking is not subject to other statutory and
regulatory requirements, the goal of which is to keep agencies accountable to the public and policymakers. For example, many national security rules are not included in the publication of the “Unified
Agenda,” which is intended to be a central database of current agency
rulemaking throughout the U.S. government.127 Nor are agencies required to conduct periodic review of existing national security regulations under the Regulatory Flexibility Act or conduct cost-benefit
analyses when they engage in national security rulemaking.128
But in our administrative law’s web of national security exceptionalism, the APA’s notice-and-comment rulemaking exception
stands out because it applies to every agency, in times of war and
peace. And it does the most work to strip national security rulemaking of the key features that are believed to ensure democratic accountability, transparency, and legitimacy.
A lack of available data makes it difficult to establish how often
the national security exception has actually been used by agencies to
disregard the notice-and-comment rulemaking requirements. A 1969
survey of agencies by the Administrative Conference of the United
States revealed that “[i]n practice, . . . most agencies do not usually
exercise their discretion to follow the [notice-and-comment] requirements . . . when they are not bound to do so because the rule-making
involves a ‘military or foreign affairs function.’ ”129 Most revealing
was the response by the Department of Defense (DoD), which seemed
126. Exec. Order No. 12,866, 3 C.F.R. § 638 (1993), reprinted as amended in 5 U.S.C. §
601. The OIRA review process was established by this Executive Order under the Clinton
Administration. The Order was revised under the Bush Administration, but the Obama
Administration restored the original language. Exec. Order No. 13,497, 74 Fed. Reg. 6113
(Feb. 4, 2009).
127. The Unified Agenda is maintained by the Regulatory Information Center in the
OIRA. See Introduction to the Unified Agenda of Federal Regulatory and Deregulatory
Actions, 77 Fed. Reg. 7664, 7665 (Feb. 13, 2012) (“Executive Order 12866 does not require
agencies to include [in the Unified Agenda] regulations concerning military or foreign affairs
functions or regulations related to agency organization, management, or personnel matters.”).
The Unified Agenda is available at http://www.reginfo.gov/public/do/eAgendaMain.
128. These requirements are only triggered by the notice-and-comment process. See, e.g.,
Airports of Entry or Departure for Flights to and from Cuba, CUSTOMS BULL. & DECISIONS,
Feb. 2011, at 1, 5 (observing that, because the national security exception relieved the
Commerce Department from conducting notice-and-comment, the “Department does not
consider this document to be subject to the provisions of the Regulatory Flexibility Act”).
129. Bonfield, supra note 22, at 232.
906
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
to have concluded that it was exempt entirely from notice-andcomment rulemaking requirements. When asked how often it relied
on the specific exception for “military functions,” and how often it relied on the other substantive exceptions for “foreign affairs,” “agency
management or personnel,” “public property,” or “contracts,” the Department replied that “ ‘[i]n a fundamental sense all regulations and
directives of the Department are incident to its essentially military
function of national defense.’ ”130 The State Department estimated
that forty percent of its rules fit within the exception.131
However, during the mid-1970s, following the Vietnam War and
Watergate, when Congress and the public heavily scrutinized the activities of military and intelligence agencies and proposals for sweeping legislative reform were in the air, DoD changed its tune.132 In response to proposals by the Administrative Conference of the United
States that the national security exception be eliminated,133 DoD in
1975 issued regulations to “voluntarily adopt procedures for public
participation in rulemaking having direct and substantial public impact.”134 Through 2005, the announced Department policy was to use
notice-and-comment procedures for such rules “unless it [was] determined by the DoD Component as a matter within its sole and exclusive prerogative that the employment of the exception or exemption
[was] appropriate to satisfy a significant and legitimate interest of
the DoD Component or the public.”135 This language, which was published in the Code of Federal Regulations, obviously still left the Department plenty of flexibility to avoid notice-and-comment.
In 2006, the Department undertook further reforms designed to
regularize the use of notice-and-comment. It removed the rule from
the Code of Federal Regulations and adopted a new Administrative
Instruction (AI 102) generally requiring the Department “Components” to comply with all rulemaking requirements that would apply
130. Id. at 239 (quoting House Comm. on Gov’t Operations, 85th Cong., Survey and
Study of Administration, Organization, Procedure and Practice in the Federal Agencies 278
(Comm. Print 1957)).
131. See id. at 261.
132. See Anjali S. Dalal, Shadow Administrative Constitutionalism and the Creation of
59,
78,
available
at
Surveillance
Culture,
2014
MICH. ST. L. REV.
http://digitalcommons.law.msu.edu/lr/vol2014/iss1/3 (observing that revelations in the early 1970s about widespread national security-related abuses of power forced the nation “into
a national dialogue about the constitutional boundaries of executive power . . . and the
appropriateness of domestic intelligence gathering”).
133. Elimination of the “Military or Foreign Affairs Function” Exemption from APA
Rulemaking Requirements, 1 C.F.R. § 305.73-5 (1975).
134. 40 Fed. Reg. 4911 (Feb. 3, 1975). The most recent version of this rule, valid
through 2005, is codified at 32 C.F.R. § 336.3.
135. 32 C.F.R. § 336.3. The regulation was superseded by U.S. Dep’t of Defense Administrative Instruction 102 and removed from the Federal Register in 2006. See Removal of
Parts, 71 Fed. Reg. 12,280, 12,281 (Mar. 10, 2006).
2014]
NATIONAL SECURITY RULEMAKING
907
to non-military or foreign affairs functions when the rules would
have an impact on the public in certain ways.136 Specifically, AI 102
requires the Department to follow notice-and-comment procedures
when a rule “[i]s presently in the CFR”; “[g]rants a right or privilege
to the public or has a direct or substantial impact on any significant
portion of the public (e.g., visitors allowed to tour the Pentagon)”;
“[p]rescribes a course of conduct that must be followed by persons
outside the Government to avoid a penalty or secure a right or privilege (e.g., behavior when visiting the Pentagon)”; “[c]onstitutes authority for persons outside the Government to act or secure immunity
from the consequence of not acting (e.g., security officers)”; “[i]mposes
an obligation on the general public or members of a class of persons
outside the Government (e.g., charges to reside at a dwelling owned
by DoD)”; “[d]escribes where the public may obtain information, instructions, and forms; make submittals or requests; take examinations; or obtain decisions (e.g., Web site information)”; or “[d]escribes
procedures by which a DoD Component conducts its business with
the public (e.g., financial institutions on DoD installations).”137
These changes reduced the need for DoD to rely on the national
security exception. But they did not necessarily make DoD more
transparent. In imposing these requirements, DoD was careful to exclude rules that “pertain[] to a military or foreign affairs function of
the United States determined to require a security classification in
the interests of national defense or foreign policy under the criteria of
an E.O. or statute (e.g., foreign military sales).”138
Even though DoD’s changes increased transparency with respect
to certain types of rulemaking, the trend toward increased secrecy
pushed other types of rulemaking further into the shadows. Overclassification of national security information has been a serious
problem for decades, despite attempts at reform. In the decade after
September 11, the number of government employees and contractors
with security clearances and the number of documents—including
agency rules—that were subject to some form of classification grew
dramatically.139 Moreover, similar voluntary notice-and-comment
procedures were not adopted by other agencies, such as the National
Security Agency or the Departments of State, Treasury, or Homeland
Security, that engage in a significant amount of national security
rulemaking directly affecting American citizens.
136. WASH. HEADQUARTERS SERVS., U.S. DEP’T OF DEF., Administrative Instruction 102
(2006), available at http://www.dtic.mil/whs/directives/corres/pdf/a102p.pdf.
137. Id. §§ E3.2.1–E.3.2.1.7.
138. Id. § E3.2.2.4.
139. See PRIEST & ARKIN, supra note 24, at 86-87.
908
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
In fact, the Federal Register contains relatively few specific invocations of the national security exception by agencies.140 The most
commonly found examples are from the Commerce and Treasury Departments, which seem to invoke the exception rather regularly when
issuing regulations regarding trade matters.141 For reasons I discuss
below, however, agencies often rely on the exception without indicating that they are doing so.
It is not uncommon for agencies to issue rules without notice-andcomment for a number of reasons. For one thing, the rules may be
merely interpretive and non-legally binding “policy statements.”142
The APA also exempts “rules of agency organization, procedure, or
practice.”143 Even for substantive rules, moreover, the APA contains a
key exception from notice-and-comment requirements for when there
is “good cause” on the ground that such requirements would be “impracticable, unnecessary, or contrary to the public interest.”144 The
“good cause” exception is by far the one most commonly invoked by
agencies during the rulemaking process to justify disregarding the
notice-and-comment procedures for legislative rules.145
A dearth of accessible data or empirical studies makes it difficult
to get an accurate picture of agency rulemaking in general.146 But secrecy and opacity make national security rulemaking especially
tricky to track. Anne Joseph O’Connell constructed the first extensive
database of rulemaking activities reported by agencies between 1983
and 2003.147 Analyzing this data, O’Connell concluded that the agen140. A search of the Federal Register yielded 2370 results suggesting an explicit invocation of the national security exception by agencies during the rulemaking process since
1981. The following query was used: (((“military” /6 “function”) or (“foreign affairs” /6 “function”)) /p (“rulemaking” or “rule”)). However, this is a very rough measure of the number of
rules for which the exception was invoked. The number may be much lower because agencies typically issue more than one publication regarding the same rule. See Anne Joseph
O’Connell, Political Cycles of Rulemaking: An Empirical Portrait of the Modern Administrative State, 94 VA. L. REV. 889, 925 n.104 (2008).
141. See, e.g., Amendments to the Export Administration Regulations: Implementation
of Limited Syria Waiver for Reconstruction Assistance, 78 Fed. Reg. 43,972 (July 23, 2013)
(to be codified at 15 C.F.R. pts. 736, 746).
142. See, e.g., Elec. Privacy Info. Ctr. v. U.S. Dep’t of Homeland Sec., 653 F.3d 1, 7
(D.C. Cir. 2011).
143. 5 U.S.C. § 553(b)(3)(A) (2012).
144. Id. § 553(b)(3)(B).
145. See GAO REPORT, supra note 121, at 19.
146. For key exceptions, see generally Kristin E. Hickman, Coloring Outside the Lines:
Examining Treasury’s (Lack of) Compliance with Administrative Procedure Act Rulemaking Requirements, 82 NOTRE DAME L. REV. 1727, 1730 (2007) (reporting results by the author of the first empirical study of Treasury’s APA compliance); O’Connell, supra note 140,
at 895 (analyzing trends using the first extensive database on agency rulemaking activities
between 1983–2003, constructed by the author from agencies’ semi-annual reports).
147. The construction of this database was a major accomplishment because it required
identifying and consolidating numerous agency documents concerning each rule.
2014]
NATIONAL SECURITY RULEMAKING
909
cies issuing the greatest percentage of their rules without notice-andcomment were, in order, the Department of State (49.4%), the Department of Homeland Security (46.2%), the Office of Personnel
Management (OPM) (42.3%), the Department of Defense (37.6%),
NASA (36.5%), and the Department of Justice (34.2%).148 With the
exception of the OPM (whose rules are likely to fall under other APA
exceptions), these agencies happen to be the most likely to engage in
opaque national security rulemaking.149
In December 2012, the General Accounting Office (GAO) reported
that it had examined a sample of published rules issued by agencies
between 2003 and 2010, and that “agencies published about 35 percent of major rules and about 44 percent of nonmajor rules” without
the Notice of Proposed Rulemaking (NPRM) required by the APA.150
The lack of an NPRM generally indicates that the agency issued the
rule without following the notice-and-comment requirements, either by
failing to solicit or consider comments at all or by soliciting comments
only after the rule had taken effect (an “interim-final” rule). Approximately seventy-seven percent of the time this happened for major rules,
the GAO reported, the agency relied on the “good cause” exception.151
The frequency with which the good cause exception is used may
seem to suggest that agencies need not rely on the national security
exception if they wish to avoid notice-and-comment. Indeed, observing recent use of the “good cause” exception, one scholar has argued
that it is being exercised by agencies “with increasing and troubling
frequency that indicates a casual disregard of public participation.”152
However, the GAO report demonstrates that something else is at
work. In fact, by far the most common reason why agencies rely on
the “good cause” exception is because they have no choice: Congress
has either mandated issuance of the rule or imposed a strict deadline
that does not give the agency enough time to conduct notice-andcomment rulemaking. The most common reasons cited by agencies
for invoking this exception were that (1) another statute or court order required the agency to act before it had time to conduct the notice-and-comment procedure; (2) another statute directly prescribed
the content of the rule; or (3) the agency was responding to an emer148. O’Connell, supra note 121, at 934.
149. Many other agencies, like the NSA, will have a larger percentage of their rules
falling into the national security exception, but most such rules remain secret and would
not be reflected in O’Connell’s study results. See supra notes 2-17 and accompanying text.
150. See GAO REPORT, supra note 121, at 8. A “major” rule is “one that, among other
things, has resulted in or is likely to result in an annual effect on the economy of $100 million or more.” See id. at 7.
151. See id. at 15.
152. Babette E.L. Boliek, Agencies in Crisis? An Examination of State and Federal
Agency Emergency Powers, 81 FORDHAM L. REV. 3339, 3343 (2013).
910
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
gency.153 In contrast, of the 123 rules issued without an NPRM that
the GAO Report examined, only two, by the Departments of Homeland Security and Commerce, specifically invoked the national security exception.154
While the GAO Report provides a useful overall picture of the frequency with which agencies avoid notice-and-comment rulemaking
for many types of rules, it is much less useful for revealing how often
agencies avoid the procedures on the ground that the rules substantively fall within the national security exception. For one thing, given
the vast number of agencies that conduct national security-related
regulation, it is quite probable that a great deal of substantive rulemaking by federal agencies happens in secret, justified on legal
grounds that never see the light of day or by the statutes that permit
important government information—even rulemaking—to be kept
secret.155 Therefore, with the explosion in the use of this classification
by agencies in the past decade, there could be thousands of unpublished rules—which still have the force of law—that will not turn
up in a search of the Federal Register and would not have been captured in the GAO Report.156
One prominent example of secret substantive rulemaking is the
development and implementation of Standard Operating Procedures
(SOPs) for airport passenger screening by the Transportation Security Administration (TSA), an agency in the Department of Homeland
Security. Although most members of the public are directly affected
by these rules, and they implicate constitutional rights, they are generally not available to the public.157 After significant public criticism
and a lawsuit by privacy advocates, the D.C. Circuit held in 2011
that the TSA must use notice-and-comment rulemaking in promulgating rules for the use of advanced imaging technology (AIT) at air153. GAO REPORT, supra note 121, at 16.
154. Changes to the Visa Waiver Program to Implement the Electronic System for
Travel Authorization (ESTA) Program, 73 Fed. Reg. 32,440, 32,444 (June 9, 2008); Revisions and Technical Correction to the Export Administration Regulations, 72 Fed. Reg.
20,221, 20,222 (Apr. 24, 2007) (making several corrections to the Export Administration
Regulations regarding Libya and terrorist-supporting countries).
155. For examples of statutes that authorize secret rulemaking, see supra note 23.
156. See supra notes 23-26 and accompanying text (discussing the problem of rampant
over-classification). “Unpublished” rules are not necessarily secret or unavailable to the
public. They are often a form of opaque rulemaking signaled in various forms in agency
communications, such as policy manuals, that may eventually be released.
157. Elec. Privacy Info. Ctr. v. U.S. Dep’t of Homeland Sec., 653 F.3d 1, 3 (D.C. Cir.
2011) (“Congress generally has left it to the agency to prescribe the details of the screening
process, which the TSA has documented in a set of Standard Operating Procedures not
available to the public.”). Some of the SOPs were inadvertently published online in 2009.
See OFFICE OF INSPECTOR GEN., U.S. DEP’T OF HOMELAND SEC., OIG-10-37, TSA’S BREACH
OF SENSITIVE SECURITY INFORMATION (2010), available at http://www.oig.dhs.gov/
assets/Mgmt/OIG_10-37_Jan10.pdf.
2014]
NATIONAL SECURITY RULEMAKING
911
ports, although it permitted the current rules to remain in place.158
Although the D.C. Circuit rejected the TSA’s arguments that the
rules were exempt, the agency did not rely on the national security
exception.159 In March 2013, the TSA issued a Notice of Proposed
Rulemaking on AIT, beginning the notice-and-comment process.160
But the most important factor that makes national security rulemaking difficult to track is that the APA’s national security exception
is a unique card that can be played by an agency at any point. In
stark contrast to the use of the “good cause” exception, courts simply
have not required agencies eschewing the notice-and-comment process to invoke the national security exception as their reason for doing so contemporaneously with the issuance of the rule. In other
words, an agency may fail to follow notice-and-comment rulemaking
procedures without explaining why, but it may subsequently rely on
the national security exception when the rule is challenged in
court.161 Even when agencies do invoke the exception when issuing a
rule, they have not been required by courts to provide any justification for why the exception applies.162
Since the APA was enacted, federal courts have discussed the national security exception in approximately one hundred cases.163 In
approximately seventy-six of those cases, the court weighed the applicability of the exception, even though the agency had not invoked
the exception during the rulemaking process and was apparently relying on the exception for the first time. In about sixty-two of those
cases, the court concluded that the exception applied; in about fourteen cases, the court held that the exception did not apply and that
the agency had therefore violated the APA. In stark contrast, if an
agency does not invoke the “good cause” exception during the rule158. Elec. Privacy Info. Ctr., 653 F.3d at 6 (holding that, because an AIT scanner produces
“an image of the unclothed passenger,” it “intrudes upon his or her personal privacy in a way a
magnetometer does not,” and that this “change substantively affects the public to a degree
sufficient to implicate the policy interests animating notice-and-comment rulemaking”).
159. See id. at 5.
160. See Passenger Screening Using Advanced Imaging Technology, 78 Fed. Reg.
18,287 (Mar. 26, 2013).
161. See, e.g., City of New York v. Permanent Mission of India to United Nations, 618
F.3d 172, 175 (2d Cir. 2010) (holding that the exception applied to the State Department’s
regulation, despite the Department’s failure to invoke the exception during rulemaking or
in the final rule); Designation and Determination Under the Foreign Missions Act, 74 Fed.
Reg. 31,788 (July 2, 2009).
162. See, e.g., Rajah v. Mukasey, 544 F.3d 427, 437 (2d Cir. 2008) (rejecting petitioners’
assertion that “the foreign affairs exception is inapplicable because the regulation itself did
not contain a statement of the undesirable international consequences flowing from the
application of notice and comment review” and concluding that there is “no requirement
that the rule itself state the undesirable consequences”).
163. These numbers, calculated from my searches on Westlaw and Lexis, are approximate because it is not always clear whether the national security exception was necessary
for the court’s decision in a particular case.
912
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
making process and provide justification for invoking it, in most cases the agency will not be permitted to rely on the “good cause” exception in court.164 The courts’ practice of permitting agencies to rely on
the national security exception in this manner is plainly inconsistent
with the Chenery rule, under which courts will uphold agency action
only on the grounds upon which the agency relied when it acted.165
Given that much of national security rulemaking happens “in the
shadows” and that the agencies themselves are not normally providing contemporaneous reasons why the national security exception
applies, courts have failed to offer clear guidance about the appropriate scope of the exception.166 Anything beyond a superficial discussion
can be found in just a handful of cases.
What is most well established is that the exception applies to international executive agreements that have the force of law.167 Similarly, regulations implementing treaties are almost always regarded
as falling within the exception.168 In one of the earliest cases applying
it, the Second Circuit determined that the FCC acted lawfully when
it eschewed notice-and-comment procedure in promulgating, pursuant to a treaty with Canada, a regulation that reduced the rights of
U.S. AM radio broadcasters.169 In 1994, the Department of Transportation relied on the national security exception to defeat a legal challenge by the Teamsters Union to a rule, enacted pursuant to the
North American Free Trade Agreement (NAFTA) without notice-andcomment, exempting Mexican truckers operating in the United
States from complying with U.S. motor vehicle laws.170
In the 1980s, the Court of International Trade attempted to discern the limits of the exception by analyzing the APA’s legislative
history and the AG Manual.171 Looking to the purpose of the rules
under review rather than the agency issuing them, the court concluded that the exception applied to rules “ ‘clearly and directly in164. See, e.g., Carlisle Tire & Rubber Co., Div. of Carlisle Corp. v. United States, 10 Ct.
Int’l Trade 301, 306 (1986).
165. See supra notes 109-115 and accompanying text.
166. See Vermeule, supra note 33, at 1112.
167. Hathaway, supra note 34, at 221 (noting that, because of the national security
exception, “international agreements are not subject to the same notice and comment rulemaking procedures that apply to nearly every other administrative rule and regulation issued
by the U.S. government” and that “no alternative oversight mechanism stands in its place”).
168. See, e.g., Int’l Bhd. of Teamsters v. Peña, 17 F.3d 1478, 1486 (D.C. Cir. 1994); Am.
Ass’n of Exps. & Imps.-Textile & Apparel Grp. v. United States, 751 F.2d 1239, 1249 (Fed.
Cir. 1985); Earth Island Inst. v. Christopher, 913 F. Supp. 559, 574 (Ct. Int’l Trade 1995);
Am. Inst. for Imp. Steel, Inc. v. United States, 600 F. Supp. 204, 211 (Ct. Int’l Trade 1984).
169. See WBEN, Inc. v. United States, 396 F.2d 601 (2d Cir. 1968).
170. See Int’l Bhd. of Teamsters, 17 F.3d at 1486.
171. Mast Indus., Inc. v. Regan, 8 Ct. Int’l Trade 214, 229-32 (1984) (noting that courts
in prior cases had not attempted to identify a limiting principle for the exception).
2014]
NATIONAL SECURITY RULEMAKING
913
volved’ in a ‘foreign affairs function,’ ”172 which extended to rules by
the U.S. Customs Service negotiated with foreign governments,173 but
not to rules by the Commerce Department establishing when a
dumping margin was regarded as “de minimis” for determining countervailing duties in dumping investigations.174
But it has not been necessary for the agency to be regulating pursuant to an international agreement for the exception to apply. Instead, it has been enough that a court perceive that the rulemaking
process would interfere with international negotiations or result in
retaliation against the United States by other countries. For example, the Second Circuit held in 2010 that the exception applied to a
State Department decision to abruptly cut short a long-running dispute between New York City and two nations’ foreign missions to the
United Nations.175 The city had maintained that, while the mission
offices were not taxable, the adjacent residences of mission employees
were.176 When two of the missions brought suit to challenge the tax,
the federal district court had agreed that, under relevant treaty provisions, the residences were taxable.177 While the missions appealed,
the State Department acted swiftly, issuing a regulation, without notice or comment, declaring that such residences were not taxable and
preempting all local and state laws.178 The Second Circuit upheld the
regulation and agreed that the national security exception applied,
even though the State Department had not invoked it when issuing
the rule.179
Apart from the context of international agreements, courts are
more likely to hold that the national security exception applies when
the court perceives that notice-and-comment would cause “definitely
undesirable international consequences,”180 the rulemaking is a response to a crisis with foreign affairs implications, or military operations are involved. None of these factors were present when a court
held that the exception did not apply to Energy Department regulations governing civilian contractors at facilities where nuclear explosives were manufactured.181 Similarly, some courts have rejected the
172. Id. at 231 (quoting H.R. REP. NO. 79-1980, at 257 (1946)).
173. Id. at 228.
174. See Carlisle Tire & Rubber Co., Div. of Carlisle Corp. v. United States, 10 Ct. Int’l
Trade 301, 305-06 (1986).
175. See City of New York v. Permanent Mission of India to United Nations, 618 F.3d
172, 201 (2d Cir. 2010).
176. Id. at 174-75.
177. Id. at 176.
178. Id. at 178.
179. Id. at 201.
180. See Rajah v. Mukasey, 544 F.3d 427, 437 (2d Cir. 2008).
181. Indep. Guard Ass’n, Local No. 1 v. O’Leary, 57 F.3d 766, 770 (9th Cir. 1995).
914
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
applicability of the exception in a few immigration contexts. For example, one court held that rulemaking regarding administrative detention of Haitian refugees did not fall within the exception, although
the INS was responding to a crisis, because nothing in the record
supported a finding that notice-and-comment rulemaking would have
resulted in “undesirable international consequences.”182 However,
multiple courts held that the national security exception applied to
rules tightening restrictions on Iranian nationals in the wake of the
Iranian Hostage Crisis in 1980.183 And courts have held that the exception applies to the designation by the military of “temporary security
zone[s]” such as the weapons testing site near Vieques in Puerto Rico.184
Although the cases addressing the exception are rare, they do
suggest that its applicability, if not always its application, is broad.
Due to the secrecy and opacity of agency decisionmaking in these areas, it is difficult to tell whether agencies are, sub silentio, relying on
the exception, or whether they are simply following a past practice of
failing to comply with the APA. Either way, national security rulemaking takes place within an administrative culture that eschews
notice-and-comment and does not value public participation or accountability. In the remainder of this Article, I discuss whether national security rulemaking should be different and why.
III. THE SCHMITTIAN VIEW OF NATIONAL SECURITY RULEMAKING
As Part II explained, the national security exception is not the only way in which agencies are authorized—by the APA itself or other
statutes—to depart from notice-and-comment procedures when issuing rules that have the force of law. What is special about the exception is the way in which it categorically separates one substantive
area of rulemaking from all others.185 If one examined only the legislative history of the bill that became the APA, one could conclude
that the decision to carve out this exception was simply based on accepted, if unexpressed, assumptions about the exceptional nature of
national security rulemaking.186
182. Jean v. Nelson, 711 F.2d 1455, 1477-78 (11th Cir. 1983), vacated and rev’d on
other grounds en banc, 727 F.2d 957 (11th Cir. 1984), aff’d, 472 U.S. 846 (1985).
183. See Nademi v. INS, 679 F.2d 811, 814 (10th Cir. 1982); Malek-Marzban v. INS, 653
F.2d 113, 115-16 (4th Cir. 1981); Yassini v. Crosland, 618 F.2d 1356, 1360 (9th Cir. 1980).
184. United States v. Ventura-Melendez, 321 F.3d 230, 232 (1st Cir. 2003).
185. The other substantive exception is for government grants and contracts, and it has
been narrowed significantly by subsequent statutes and regulations. See supra notes 119121 and accompanying text.
186. See, e.g., STAFF OF S. COMM. ON THE JUDICIARY, 79TH CONG. (Comm. Print 1945),
reprinted in ADMINISTRATIVE PROCEDURE ACT: LEGISLATIVE HISTORY, 79TH CONGRESS,
1944–46, at 15, 17 (1946) (referring to the exception as “self-explanatory”); infra notes 218224 and accompanying text.
2014]
NATIONAL SECURITY RULEMAKING
915
Indeed, this is probably the accepted view. Professor Adrian Vermeule has recently argued that the entire corpus of American administrative law, from authorizing statutes to court decisions, allows
agencies to transcend the bounds of administrative procedure and
meaningful judicial review in emergencies.187 Vermeule identifies the
national security exception as one example of a number of such
“black holes” created by the APA and other statutes—areas where
the executive may operate free of legal constraints.188 In addition,
Vermeule contends that administrative law contains numerous “grey
holes”—ambiguous legal provisions that provide courts opportunities
to indulge their natural tendency to give heightened deference to
agencies during national crises.189 Within these grey holes, courts can
preserve the pretense of the rule of law while deferring entirely to
the executive.190
In articulating this thesis, Vermeule embraced the controversial
theories of German political scientist Carl Schmitt, “ ‘the outstanding
legal theorist of the . . . exception.’ ”191 Schmitt’s theories have descriptive and normative aspects. He is well known for his assertion
that the “[s]overeign is he who decides on the exception.”192 However,
Schmitt not only argued that this phenomenon is inevitable; he concluded that it is necessary for the state’s survival.193 Moreover, as
Schmitt developed his theories, the exception took on greater and
greater importance, and Schmitt ultimately argued against limiting
the sovereign’s discretion at all.194 Because “[t]he precise details of an
emergency cannot be anticipated” in advance, Schmitt argued, the
“sovereign” must enjoy absolute freedom both to determine “whether
there is an . . . emergency” and “what must be done to eliminate it.”195
Vermeule takes from Schmitt that an administrative legal regime
replete with black and grey holes is inevitable. It is “hopelessly utopian,” Vermeule argues, to attempt to apply meaningful administra187. Vermeule, supra note 33, at 1096-97; see also Evan J. Criddle, Mending Holes in
the Rule of (Administrative) Law, 104 NW. U. L. REV. 1271, 1271 (2010).
188. Vermeule, supra note 33, at 1096.
189. Id.
190. See id. at 1114. Vermeule was using and responding to the concept of grey holes
and black holes used by David Dyzenhaus. See DAVID DYZENHAUS, THE CONSTITUTION OF
LAW: LEGALITY IN A TIME OF EMERGENCY 3 (2006).
191. Gross, supra note 44, at 1826 (emphasis omitted) (quoting HANS KELSEN & CARL
SCHMITT, A JUXTAPOSITION 10 (Dan Diner & Michael Stolleis eds., 1999)).
192. Carl Schmitt, Political Theology: Four Chapters on the Concept of Sovereignty 5
(George Schwab trans., MIT Press 1985) (1922).
193. See CARL SCHMITT, THE CRISIS OF PARLIAMENTARY DEMOCRACY 12 (Ellen Kennedy
trans., MIT Press 1985) (1923); David Dyzenhaus, Emergency, Liberalism, and the State, 9
PERSP. ON POL. 69, 72 (2011).
194. See Gross, supra note 44, at 1841.
195. SCHMITT, supra note 192, at 6-7.
916
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
tive law constraints during emergencies: “The exception cannot, realistically, be banished from administrative law; exceptions are necessarily built into its fabric.”196 Vermeule is less clear about whether
this state of affairs is a desirable one.197
While Vermeule is almost certainly correct that exceptions cannot
be banished entirely from administrative law, critics have disagreed
with his contention that courts in fact do practice the sort of absolute
deference he identifies.198 They have also argued that administrative
law can be reformed to reduce the ability of the executive to act without constraint during emergencies—for example, by altering the nature of the exceptions,199 changing the culture of agencies and courts
to encourage consideration of “public-regarding factors,”200 or creating
a more detailed legal framework for emergency administration.201
It is no coincidence that interest in Schmitt, on the part of legal
scholars, has exploded in the decade following the September 11 attacks and the resulting global war against Al Qaeda and the Taliban.202 Schmitt’s core observation that states of exception must exist
in liberal democracies is an intuitively powerful one. Although
Schmitt’s views have recently found more sympathy among conservative scholars, their influence—even if Schmitt himself is not mentioned—is something to be reckoned with. Nearly every theorist seriously addressing emergency or war governance after September 11
has found it necessary to grapple with the Schmittian perspective.203
It is also no coincidence that Schmitt developed his theories during a period of acute crisis in his native Germany during the 1920s
196. Vermeule, supra note 33, at 1097, 1104.
197. Philip Alston, The CIA and Targeted Killings Beyond Borders, 2 HARV. NAT’L
SECURITY J. 283, 429 (2011) (observing that a “problem with Vermeule’s approach is the
extent to which it blurs empirical conclusions with normative arguments”).
198. See, e.g., Criddle, supra note 187, at 1274, 1275 & n.20.
199. See id. at 1272; David Dyzenhaus, Schmitt v. Dicey: Are States of Emergency Inside or Outside the Legal Order?, 27 CARDOZO L. REV. 2005, 2026 (2006).
200. Criddle, supra note 187, at 1272.
201. See Ackerman, supra note 105, at 1030-31, 1044 (proposing “a newly fashioned
emergency regime” that permits “short-term emergency measures but draws the line
against permanent restrictions” to “rescue the concept” of emergency power “from fascist
thinkers like Carl Schmitt, who used it as a battering ram against liberal democracy”).
202. See DUDZIAK, supra note 44, at 116 (observing that “[c]itations to Schmitt in legal
scholarship steadily increased” after September 11, with “twenty-four citations to Schmitt
in 2001, twenty-nine in 2002, fifty-one 2003, and eighty-six in 2009”); David Luban, Carl
Schmitt and the Critique of Lawfare, 43 CASE W. RES. J. INT’L L. 457, 463 (2010). Interest
in Schmitt had already been gaining steam prior to September 11. See D.A. Jeremy Telman, Should We Read Carl Schmitt Today?, 19 BERKELEY J. INT’L L. 127, 129 (2001)
(book review).
203. See DUDZIAK, supra note 44, at 116-18 (discussing responses to Schmitt among
legal scholars).
2014]
NATIONAL SECURITY RULEMAKING
917
and 1930s.204 Schmitt witnessed the perceived weakness of the democratic Weimar government and saw it collapse with Hitler’s rise to
power.205 Schmitt himself became a Nazi Party member, and the extent to which he justified the exploitation of legal loopholes that enabled the rise of the Third Reich is the subject of intense debate.206 For
many contemporary American legal scholars like Vermeule, however,
the history is beside the point; what is important is not the historical
context in which Schmitt arrived at his understanding of governance,
but what he understood and explained about the essential nature of
administration during war and emergency, which is true for every
democratic government in every era.207
This elevation of theory over context is problematic. It is entirely
possible that rules created in times of emergency and war are intended to address the current emergency and were never intended to be
appropriate for future wars or emergencies. Yet path dependency and
legislative and regulatory ossification leave legal frameworks in place
long after the problems they were intended to address no longer exist.208
Moreover, where Vermeule’s description of black holes in American administrative law fits uneasily with Schmitt’s approach is in the
concept of emergency. Although the core of Schmitt’s work explored
states of “exception” (Ausnahmezustand), his later writings made
clear that he meant for the term to encompass much more than what
we commonly understand to be emergencies—temporary periods of
extreme peril, such as war or natural disasters.209 In fact, Schmitt
ultimately concluded that, to limit the exception to emergencies—or
to impose any limitations at all on the breadth of the concept—would
204. See Telman, supra note 202, at 130-32 (discussing the influence of World War I
and the collapse of the pre-war order on Schmitt’s philosophy). For biographical information about Schmitt, see, for example, JOSEPH W. BENDERSKY, CARL SCHMITT: THEORIST
FOR THE REICH (1983).
205. See Telman, supra note 202, at 136 (“Having struggled, along with others, to provide the fledgling Weimar Republic with a legal theory that could guarantee its stability,
Schmitt arrived at the conclusion that the Sovereign needed, in certain situations, to become the exception that cannot be bound by law.”).
206. Ellen Kennedy, Carl Schmitt and the Frankfurt School, TELOS, Fall 1987, at 37,
37-39; Telman, supra note 202, at 131-32.
207. Writing with Eric Posner, Vermeule described his approach as “extract[ing] the
marrow from Schmitt and then throw[ing] away the bones for the professional exegetes to
gnaw.” ERIC A. POSNER & ADRIAN VERMEULE, TERROR IN THE BALANCE: SECURITY,
LIBERTY, AND THE COURTS 38 (2007).
208. Kim Lane Scheppele, Exceptions That Prove the Rule: Embedding Emergency Government in Everyday Constitutional Life, in THE LIMITS OF CONSTITUTIONAL DEMOCRACY
124, 124-34 (Jeffrey K. Tulis & Stephen Macedo eds., Princeton University Press 2010)
(detailing examples of foreign and domestic legislative acts that tend to “constitutionalize
emergency exceptions to normal government” (emphasis omitted)); cf. GIORGIO AGAMBEN,
STATE OF EXCEPTION 2 (Kevin Attell trans., Univ. of Chi. Press 2005) (contending that
states of emergency tend to become permanent).
209. See Gross, supra note 44, at 1826-28.
918
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
endanger the state by constraining the executive’s ability to determine when a state of exception exists and to respond appropriately.210
Therefore, our administrative law is both more and less Schmittian than Vermeule describes. It is more Schmittian because many of
the black holes that he identifies are regularly used by agencies in
times of peace and non-emergency. This is especially true of the national security exception. There was, for example, no war or emergency concerning Canada that required the FCC to issue regulations
limiting the power of AM radio stations without notice-andcomment.211 The same could be said of most regulations to which the
courts have concluded the exception applies.212 At the same time, the
national security exception is perhaps less Schmittian because, although it may be broadly construed, it is arguable whether it truly
constitutes a “suspension of the entire existing order.”213 The only
way to make the actual operation of the exception fit with Vermeule’s
use of the word “emergency” would be to acknowledge that we live in
a permanent state of emergency.214
The problem with Vermeule’s Schmittian approach to our administrative law, then, is that, like Schmitt’s own theories, it has little to
offer concerning the limits of the state of exception. A Schmittian/Vermeulian approach to the national security exception would be
an acknowledgement that it has no practical limits. The exception is
capable of swallowing the general rule.
Yet if we wish to impose some limits on the exception, how can we
do so? Although there are many ways to approach this question, the
historical context in which the APA’s national security exception was
created reveals a great deal about why the exception was believed to
be necessary at the time and whether it continues to be necessary
today. The next Part offers the first explanation for the exception
based on this historical context.
210. See William E. Scheuerman, Legal Indeterminacy and the Origins of Nazi Legal
Thought: The Case of Carl Schmitt, 17 HIST. POL. THOUGHT 571, 589 (1996) (“If legal indeterminacy is a truly ubiquitous facet of legal experience, then dictatorship similarly must
take something close to an omnipresent, even permanent form.”).
211. See WBEN, Inc. v. United States, 396 F.2d 601 (2d Cir. 1968).
212. See supra notes 166-184 and accompanying text.
213. SCHMITT, supra note 192, at 12.
214. This is in fact exactly what Italian political theorist Giorgio Agamben has argued,
using Schmitt’s theories and the Guantánamo Bay detention camp as points of departure.
See AGAMBEN, supra note 208, at 2 (“[T]he voluntary creation of a permanent state of
emergency . . . has become one of the essential practices of contemporary states, including
so-called democratic ones.”).
2014]
NATIONAL SECURITY RULEMAKING
919
IV. THE HISTORICAL ROOTS OF NATIONAL SECURITY RULEMAKING
The conventional wisdom assumes that the fundamental exigencies driving national security rulemaking are basically static.215 Defenders of national security exceptionalism argue that the nature of
national security policymaking is fundamentally different from ordinary policymaking, while critics argue that exceptional treatment of
national security policymaking threatens democratic values.216 The
debate is an essentialist one. However, it is entirely possible that, as
the world changes and threats to America’s security change with it,
the architecture of the administrative state ought to change in response. If this is so, it is vitally important to understand the historical
context in which current administrative law doctrine was developed.
This Part departs from the conventional wisdom and advances an
alternative interpretation of the APA’s national security exception
that is rooted in the history surrounding its enactment. Despite the
dearth of specific legislative history explaining the exception with
respect to the final APA itself, it is possible to reconstruct the exception’s purpose by reference to the historical context, the debates surrounding the predecessor legislation to the APA, and influential conceptions of organizational theory at the time of the APA’s enactment.217
An examination of this context reveals the striking degree to
which the national security exceptionalism embedded in our current
administrative law regime was a creature of the unique geopolitical
and domestic political concerns of the 1930s and 1940s. Pearl Harbor,
the rise of fascism, and the war-related failures of federal govern215. See supra Part III. For an argument that, although the conventional wisdom assumes that national security considerations are timeless, they are in fact rooted in outdated views of international relations theory, see Robert Knowles, American Hegemony and
the Foreign Affairs Constitution, 41 ARIZ. ST. L.J. 87, 93 (2009).
216. Compare Jide Nzelibe, The Uniqueness of Foreign Affairs, 89 IOWA L. REV. 941,
943-44 (2004) (arguing that foreign affairs (and national security) matters deserve unique
treatment by courts and in constitutional law), and supra notes 197-201, with Jeremy
Waldron, Security and Liberty: The Image of Balance, 11 J. POL. PHIL. 191, 206 (2003) (“We
have to worry that the very means given to the government to combat our enemies will be
used by the government against its enemies . . . .”).
217. The following is drawn from primary sources and invaluable historians’ and political scientists’ accounts of the APA and its era—in particular, those that describe the deep
connection between New Deal politics and international relations and those addressing
specifically the debate surrounding administrative law reform. Especially helpful were
DUDZIAK, supra note 44; KATZNELSON, supra note 40; IDO OREN, OUR ENEMIES AND US:
AMERICA’S RIVALRIES AND THE MAKING OF POLITICAL SCIENCE (2003); DOUGLAS T. STUART,
CREATING THE NATIONAL SECURITY STATE: A HISTORY OF THE LAW THAT TRANSFORMED
AMERICA (2008) (offering a definitive history of the National Security Act); Kovacs, supra
note 125; McNollgast, The Political Origins of the Administrative Procedure Act, 15 J.L.
ECON. & ORG. 180 (1999); Reuel E. Schiller, Reining in the Administrative State: World
War II and the Decline of Expert Administration, in TOTAL WAR AND THE LAW: THE
AMERICAN HOME FRONT IN WORLD WAR II 185 (Daniel R. Ernst & Victor Jew eds., Praeger
Publishers 2002); and George B. Shepherd, Fierce Compromise: The Administrative Procedure Act Emerges from New Deal Politics, 90 NW. U. L. REV. 1557 (1996).
920
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
ment agencies were constantly on the minds of those contemplating
and constructing the modern administrative state.
These contemporaneous concerns infused, even distorted, the debate concerning administrative law reform to such a degree that
there is, ironically, very little discussion of national security rulemaking, as such, in the legislative history of the bill that became the
APA.218 Nonetheless, through a decades-long, often-vitriolic debate on
reforming the administrative state that culminated in the APA, a
story emerges about the origins of national security rulemaking:
against the backdrop of a world in crisis, with the rise of fascist powers—and eventually, a world war against those powers—Congress
decided to carve out greater and greater space for a distinct regime of
national security rulemaking, even as it created a regime for ordinary
rulemaking intended to rein in the excesses of the administrative state.
The APA emerged as the ultimate compromise after a long struggle during the 1930s and ’40s by New Deal opponents to tame the
growing administrative state.219 With respect to core administrative
law issues such as judicial review of agency action, it reflected the
settlement of “ ‘long-continued and hard-fought contentions, and enact[ed] a formula upon which opposing social and political forces have
come to rest.’ ”220
Yet by the time Congress debated the bill that became the APA,
the question of how to treat national security rulemaking appears to
have been largely settled. The presence of the national security exception was little remarked upon,221 and it was often referred to as
“self-explanatory.”222 The explanations that were offered seem to
suggest that some in Congress believed it should be narrowly construed. One Congressman, in reference to the exception, said the following: “The exemption of military and naval functions needs no explanation here. The exempted foreign affairs are those diplomatic
functions of high importance which do not lend themselves to public
procedures and with which the general public is ordinarily not directly concerned.”223 The House and Senate Report cautioned against
overbroad interpretation:
218. See Tibbels, supra note 34, at 395 n.27 (noting that the legislative history on the
exception is limited to a paragraph).
219. See generally Shepherd, supra note 217.
220. Vt. Yankee Nuclear Power Corp. v. Natural Res. Def. Council, Inc., 435 U.S. 519,
523 (1978) (quoting Wong Yang Sung v. McGrath, 339 U.S. 33, 40 (1950)).
221. Hathaway, supra note 34, at 243 n.312.
222. See, e.g., STAFF OF S. COMM. ON THE JUDICIARY, 79TH CONG. (Comm. Print 1945),
reprinted in ADMINISTRATIVE PROCEDURE ACT: LEGISLATIVE HISTORY, 79TH CONGRESS,
1944–46, at 15, 17 (1946); see also Hathaway, supra note 34, at 171-72.
223. 92 CONG. REC. 5650 (1946) (statement of Rep. Walter).
2014]
NATIONAL SECURITY RULEMAKING
921
The phrase “foreign affairs functions,” used here and in some other
provisions of the bill, is not to be loosely interpreted to mean any
function extending beyond the borders of the United States but only those “affairs” which so affect relations with other governments
that, for example, public rule making provisions would clearly provoke definitely undesirable international consequences.224
Although some in Congress apparently believed that the national
security administrative state should be small, the Executive Branch,
after the APA became law, seized on the provision’s language to imagine a vast regime of national security rulemaking. Indeed, as the first
scholar to analyze the exception observed, “the language of the . . . exemption is very broad” and the “functions excluded are written in
terms easily susceptible to wide application.”225 The influential 1947
Interpretive Guide to the APA issued by the office of the Attorney
General and future Supreme Court Justice Tom Clark took what one
commentator has called an “exceptionally sweeping” view of the national security exception’s scope.226 The Attorney General’s Manual
is, for courts, a “key document” for interpreting the APA, contains the
“most authoritative account” of the history of its passage,227 and has
been “given some deference . . . because of the role played by the Department of Justice in drafting the legislation.”228
The reference to “military and naval functions” in the national security exception, the Manual concluded, was “not limited to activities
of the War and Navy Departments but covers all military and naval
functions exercised by any agency.”229 For example, the Manual offered, “the exemption applies to the defense functions of the Coast
Guard and to the function of the Federal Power Commission under
section 202(c) of the Federal Power Act.”230 By referring to “foreign
affairs functions,” according to the Manual, the provision was “applicable to most functions of the State Department and to the foreign
affairs functions of any other agency.”231
The Manual’s broad interpretation of the national security exception was the culmination of a trend that began with that era’s first
224. S. REP. NO. 79-752, at 13, 16 (1945); see also H.R. REP. NO. 79-1980, at 2-3, 23,
27 (1946).
225. Bonfield, supra note 22, at 238.
226. Tibbels, supra note 34, at 396; see generally U.S. DEP’T OF JUSTICE, ATTORNEY
GENERAL’S MANUAL ON THE ADMINISTRATIVE PROCEDURE ACT 26-27 (1947)
[hereinafter MANUAL].
227. Citizens Awareness Network, Inc. v. United States, 391 F.3d 338, 362 (1st Cir.
2004) (citations and internal quotation marks omitted).
228. Vt. Yankee Nuclear Power Corp. v. Natural Res. Def. Council, Inc., 435 U.S. 519,
546 (1978).
229. MANUAL, supra note 226, at 26.
230. Id.
231. Id. at 27.
922
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
efforts at administrative law reform. From the first proposed legislation intended to restrain administrative agencies in 1929, to the final
passage of the APA in 1946, with some exceptions, each subsequent
draft of legislation recognized a broader national security administrative state than the one before it. The earliest proposals, which saw
no action in Congress, did not mention the military or foreign affairs
at all.232 By 1937, however, the world had changed and domestic politics along with it. Much of the public had begun to perceive agencies
in the Roosevelt Administration as overreaching. Amid a recession
and backlash against Roosevelt’s court-packing plan, and with a newly-quiescent Supreme Court no longer willing to strike down New
Deal programs, support for administrative reform as a counterbalance to the New Deal grew.233 The anti-New Deal “conservative
coalition” of Republicans and mostly-Southern Democrats took effective control in Congress.234
Fears of federal agencies run amuck became intertwined with increasing awareness by Americans of totalitarianism in Europe and
Asia. Through the 1930s, the United States had closer ties with Fascist Italy than the Communist USSR, Nazi Germany, or Imperial Japan.235 For many Americans, however, the entanglement of business
and government in fascist states bore a disturbing similarity to the
way U.S. government agencies were closely involved with businesses
in controlling wages and prices and the detailed regulation of industries.236 The floor debates on APA predecessor legislation during the
1930s and ’40s reflect these concerns: they are “ ‘riddled with comparisons of the administrative state to fascist and communist govern232. These early proposals, which began before FDR took office, sought simply to vest
review of administrative decisions in the courts, which were considered conservative and
skeptical of regulation. See Kovacs, supra note 125, at 681-83; Shepherd, supra note 217, at
1566-67. The first proposal introduced by Senator George Norris in 1929, see S. 5154, 70th
Cong. (1929), went nowhere because even conservatives thought it unnecessary. See Kovacs, supra note 125, at 682-83.
233. See KATZNELSON, supra note 40, at 256-75 (describing the court-packing plan and
recession); Shepherd, supra note 217, at 1586-93 (describing gathering momentum for
administrative reform).
234. See McNollgast, supra note 217, at 204.
235. See KATZNELSON, supra note 40, at 92-93 (“[C]ore policymakers . . . were drawn to
Mussolini’s Italy, which self-identified as a country that had saved capitalism.”); id. at 6368 (describing the rapturous public reception Fascist Italy’s Air Force Marshall and Mussolini heir-apparent Italo Balbo received during a visit to the United States).
236. See, e.g., James Q. Whitman, Of Corporatism, Fascism, and the First New Deal, 39
AM. J. COMP. L. 747, 748 (1991) (“To supporters and critics alike, [the National Recovery
Act], a vast scheme for delegating governmental authority to private cartels, seemed akin
to the ‘corporativism’ of Italian Fascism.”); Hoover Warns New Deal Leads U.S. to Fascism,
WASH. POST, May 6, 1938, at X2 (summarizing and quoting a speech by FDR’s defeated
predecessor, Herbert Hoover, broadcast to millions on the radio). For a recent revisionist
history echoing these criticisms and roundly condemning the New Deal’s economic reforms
and alleged autocratic tendencies, see AMITY SHLAES, THE FORGOTTEN MAN: A NEW
HISTORY OF THE GREAT DEPRESSION (2007).
2014]
NATIONAL SECURITY RULEMAKING
923
ments’ ” and accusations that administrative agencies were being
used to advance FDR’s totalitarian ambitions.237
In 1937, the influential (and decidedly anti-New Deal) ABA Special Committee on Administrative Law, led by the “dyspeptic” former
Harvard Law Dean, Roscoe Pound,238 drafted legislation that would
have imposed tighter control on agencies than earlier proposals. It
required notice and public hearings prior to regulation, formal administrative hearings, and judicial review.239 At the same time,
however, perhaps recognizing the increasing instability in geopolitics, the Committee proposed to exempt from the bill “foreign affairs” and “the conduct of military or naval operations in time of
war or civil insurrection.”240
Two years later, after Hitler had invaded Poland and World War
II had begun,241 a bill similar to the APA’s proposal was introduced in
Congress as the Walter-Logan Bill.242 The national security exception
now was slightly broader: it excluded altogether “any matter concerning or relating to the conduct of foreign affairs” and “the conduct of
military or naval operations in time of war or civil insurrection.”243
Despite the presence of these exceptions, they were strongly criticized as too narrow. As the nation sought to absorb the news of the
devastating strength of Hitler’s military and the relative weakness of
America’s,244 the War Department savaged the Walter-Logan bill, arguing that it would be “gravely subversive of military discipline in all
components of the Army, destructive of efficiency in the performance
237. See Kovacs, supra note 125, at 685 (quoting Reuel E. Schiller, Free Speech and
Expertise: Administrative Censorship and the Birth of the Modern First Amendment, 86 VA.
L. REV. 1, 86 (2000)).
238. Schiller, supra note 217, at 197 (describing Pound as “dyspeptic”).
239. Shepherd, supra note 217, at 1582-83; Report of the Special Committee on Administrative Law, 62 A.B.A. ANN. REP. 846-50 (1937) [hereinafter Special Committee].
240. Special Committee, supra note 239, at 789. For various reasons—having mostly to
do with pre-existing judicial review provisions—the Committee also exempted the Federal
Reserve Board, the Comptroller of the Currency, and the Federal Deposit Insurance Corporation, in addition to exempting “any case arising under the internal revenue, customs,
patent . . . or Longshoremen's & Harbor Workers' Laws; or any case in which the aggrieved
person was denied a loan or is dissatisfied with a grading service rendered by the United
States in connection with the purchase or sale of agricultural products.” Id. at 850; see also
Shepherd, supra note 217, at 1600.
241. See KATZNELSON, supra note 40, at 302 (concluding that “Germany’s lightning
attack on Poland transformed legislative possibilities” for ending the embargo on arms
sales to Britain and France).
242. S. 915, 76th Cong. (1939); H.R. 4236, 76th Cong. (1939); H.R. 6324, 76th Cong.
(1939); Schiller, supra note 217, at 197.
243. S. 915, 76th Cong. § 6(b) (1939).
244. See KATZNELSON, supra note 40, at 306-08; GEORGE F. KENNAN, AMERICAN
DIPLOMACY 1900–1950, at 66 (1951) (observing that, in 1939, “the overwhelming portion of
the world’s armed strength in land forces and air forces” belonged to Germany, Japan, and
the USSR, and that Western democracies “had become militarily outclassed”).
924
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
of the functions of the War Department, both military and nonmilitary, obstructive to progress in preparedness for national defense, and generally disastrous from the viewpoint of the public interest.”245 The bill would, according to the Department, enable military personnel to challenge orders “on any occasion except in time of
war or insurrection.”246 The Department suggested that “all matters
concerning or relating to the operations of the War Department and
the Army” be exempted.247
In response to these objections, one of the bill’s co-sponsors, Senator M.M. Logan, amended the bill in a draft that defined “agency” for
the first time, and he broadened the military exception so that it applied “to the conduct of military or naval operations” at all times.248
However, the new draft did not include an exception for foreign affairs. Although the House Judiciary Committee reported a similar
bill favorably, an influential critique emerged from Congressman Cellar, a co-sponsor of the earlier, even more restrictive ABA bill. Congressman Cellar, who had a change of heart, now argued that “it
would be manifestly inappropriate to require the War Department to
conduct hearings on Army regulations.”249 As the House debated the
bill just days after Hitler invaded Denmark, other congressmen also
complained that the military exception was too narrow.250 A lastminute amendment by Congressman Walter purported to further expand the exception, re-defining “military ‘operations’ to include ‘strictly
military and naval activities of the War and Navy Departments.’ ”251
The Senate grappled with similar concerns. Two months later, in
summer 1940, Hitler’s armies marched into Paris and the Luftwaffe
245. Letter from Harry H. Woodring, Sec’y of War, U.S. War Dep’t, to Rep. Hatton W.
Sumners, Chairman, Judiciary Comm., U.S. House of Representatives (May 6, 1939), reprinted in Bills to Provide for the More Expeditious Settlement of Disputes with the United
States, and for Other Purposes: Hearings on H.R. 4236, H.R. 6198, and H.R. 6324 Before
Subcomm. No. 4 of the H. Comm. on the Judiciary, 76th Cong. 102 (1939).
246. Id. at 103.
247. Id.
248. S. REP. NO. 76-442, at 1, 5 (1939).
249. H.R. REP. NO. 76-1149, pt. 2, at 6 (1940) (Minority Rep.); see also id. at 5 (contending that the exception should encompass “other activities” of the Departments of War and
the Navy “which highly affect public interest and the national defense, such as river and
harbor improvements, and purchase of munitions and supplies”).
250. Id. at 688; see 86 CONG. REC. 4653 (1940) (statement of Rep. McGranery) (contending that the exemption would enable military personnel to challenge promotion decisions in
the courts of appeals and substitute the court’s judgment for that of armed services). But
see id. at 4649 (statement of Rep. Gwynne) (arguing, in defense of the existing language,
that “the actual conduct of the armies and the navies is an executive function and Congress
and the courts have very little, if anything, to do with it”).
251. Kovacs, supra note 125, at 688 (quoting 86 CONG. REC. 4725 (1940) (statement of
Rep. Walter)). Congressmen disagreed about whether this actually narrowed or broadened
the national security exception, but the language was added and the bill passed. See id.
at 689.
2014]
NATIONAL SECURITY RULEMAKING
925
began the London Blitz while “[a] remarkable national consensus developed among political leaders and the mass populace to build
American strength,”252 and Congress was debating what would become the first peacetime conscription bill in U.S. history.253 The Senate Judiciary Committee, working from the same language as the
House, expanded the exception further, omitting the word “conduct”
and adding the phrase “any other agency or authority hereafter created to expedite military and naval defense.”254 Congressman
Sumners, a proponent of the bill, concluded that this language “covers everything that may be done by any agency concerning or relating to the Military and Naval Establishments.”255 Congressman
Cochran thought the language was still not broad enough and would
not cover civilian agencies “performing functions which are indispensable to the workings of our defense program.”256 In any event, the
full Senate adopted this language, and the House concurred in
December 1940.257
Yet despite the support of large majorities from both parties in
Congress, the Walter-Logan bill would not become law. President
Roosevelt vetoed it on December 18, 1940, unhappy with the constraints it imposed on administrative agencies, but also concerned
about its effect on preparations for war.258 In his veto statement, Roosevelt articulated the need for a broader national security administrative state. He noted that Walter-Logan would impose new regulatory burdens on agencies, such as the Maritime Commission and the
Departments of Commerce and Treasury, “whose activities have an
important collateral effect on the defense program,” and which had
“pointed out serious delays and uncertainties which would be caused
by the present bill.”259 An attempted veto override in the House failed
that same day.260
Although his veto was sustained, Roosevelt had recognized the
momentum behind the reformers and already ordered the creation of
252. KATZNELSON, supra note 40, at 307.
253. See 86 CONG. REC. 11,489 (1940); KATZNELSON, supra note 40, at 312-13 (describing the conscription bill as “revolutionary”).
254. The language read as follows: “Nothing contained in this act shall apply to or affect any matter concerning or relating to the Military or Naval Establishments . . . and any
other agency or authority hereafter created to expedite military and naval defense.” 86
CONG. REC. 13,746-47 (1940).
255. 86 CONG. REC. 13,811 (1940) (statement of Rep. Sumners).
256. Id. at 13,810 (statement of Rep. Cochran).
257. Id. at 13,815-16; Shepherd, supra note 217, at 1625.
258. Id. at 13,942-43; Kovacs, supra note 125, at 690.
259. Franklin D. Roosevelt, U.S. President, The President Vetoes the Bill Regulating
Administrative Agencies (Dec. 18, 1940), in 1940 THE PUBLIC PAPERS AND ADDRESSES OF
FRANKLIN D. ROOSEVELT 616, 620 (Samuel I. Rosenman ed., 1941).
260. 86 CONG. REC. 13,953 (1940).
926
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
an Attorney General’s Committee on Administrative Procedure,
which submitted a report on January 22, 1941.261 The report contained two proposals for legislation—one from the liberal majority
and one from the conservative minority.262 Both bills gave more flexibility to agencies than had the ill-fated Walter-Logan bill, but their
approaches varied on both ordinary rulemaking and national security
rulemaking. The liberal bill “imposed little restraint on agencies,”
while the conservative bill “would have controlled agencies substantially.”263 The liberal bill, however, did not exempt national security
rulemaking at all, while the conservative bill carved out national security exceptions that began to closely resemble those that would be
in the APA. The conservative bill authorized the President to temporarily suspend any of the act’s provisions under certain circumstances.264 And with respect to notice-and-comment rulemaking, it provided that “[w]henever expressly found by an agency to be contrary to
the public interest, the provisions of this title, in whole or part, shall
not apply to . . . the conduct of military, naval, or national-defense
functions, or the selection or procurement of men or materials for the
armed forces of the United States . . . .”265
Both bills were introduced in Congress, which debated them during the summer of 1941, as Nazi Germany began its brutal invasion
of the Soviet Union, and Japan had already “conquered the Philippines, Burma, Hong Kong, Malaya, Singapore, and the Dutch East
Indies.”266 The War Department objected that the national security
exceptions in both bills were still inadequate, arguing that the military “could not properly function” if it had to meet the procedural requirements and demanding a full exception for all activities of the
War Department and all military branches.267 The Attorney General’s
Committee, chaired by Assistant Secretary (and future Secretary) of
State Dean Acheson—who would play a key role in restructuring the
national security state after the war268—did not seem to share these
concerns, describing as “very clear” the majority bill's exceptions for
261. S. DOC. NO. 77-8 (1941); Shepherd, supra note 217, at 1632.
262. Shepherd, supra note 217, at 1632.
263. Id. at 1633-34.
264. S. 674, 77th Cong. § 111 (1941).
265. Id. § 201.
266. See KATZNELSON, supra note 40, at 41.
267. Administrative Procedure: Hearings on S. 674, S. 675, and S. 918 Before a Subcomm. of the S. Comm. on the Judiciary, 77th Cong. 36 (1941) (statement of Karl R.
Bendetson, Captain, Office of the Judge Advocate General) [hereinafter Administrative
Procedure: Hearings on S. 674, S. 675, and S. 918].
268. See STUART, supra note 217, at 5.
2014]
NATIONAL SECURITY RULEMAKING
927
“the military service, the armed forces, or the selection and discharge of
employees.”269 But the Committee, in another sense, chose to go further.
In rejecting agency-specific exceptions to rulemaking requirements, the Committee instead settled on a “functional” approach, under which any agency would be exempt from rulemaking (and adjudication) requirements to the extent that it performed certain substantive functions.270 This functional approach prevailed and was included in the APA. This approach would limit the scope of the exception
in some respects: no agency could firmly declare itself entirely free of
notice-and-comment requirements.271 On the other hand, the functional approach vastly expanded the number of agencies that could
engage in national security rulemaking. In practice, this meant that
agencies across the government could, and did, rely on the exception,
to varying degrees, to eschew notice-and-comment procedures.272 And,
as it turned out, agencies would enjoy a great deal of discretion to use
the exception when they saw fit.273
Although administrative law reform was largely put on the back
burner while the war was waged, by 1946, with the war over and
New Dealers on the defensive in Congress, the moment for comprehensive administrative law reform had finally arrived.274 The APA
would be enacted into law by year’s end.275
Nonetheless, the lessons learned from the war about the appropriate way to control the exploding administrative state were hardly
straightforward. During the war, administrative agencies had grown
even more powerful and more numerous, but significantly less popular. Americans were chafing under the price controls, comprehensive
rationing, and other rigid regulations required during wartime.276
The war revealed the ugly side of bureaucracies, which proved to be
often “inefficient, incompetent, bullying, and perhaps even captured
269. Administrative Procedure: Hearings on S. 674, S. 675, and S. 918, supra note
267, at 820 (statement of Dean Gooderham Acheson, Assistant Secretary of State for
Economic Affairs).
270. S. DOC. NO. 79-248, at 191 (1946); see also Bonfield, supra note 22, at 235-36. The
Navy was, at the time, considered a service separate from the rest of the military. See
STUART, supra note 217, at 10-15 (discussing military restructuring after the war).
271. See S. DOC. NO. 79-248, at 13 (“All departments may, and often do, exercise civil
and regulatory powers which should be subject to an administrative procedure statute.”);
see also Bonfield, supra note 22, at 235-36.
272. See Bonfield, supra note 22, at 233-34.
273. See supra Part II.C.
274. See McNollgast, supra note 217, at 192-93, 203.
275. See Kovacs, supra note 125, at 703-04 (describing the APA’s passage).
276. See KATZNELSON, supra note 40, at 216.
928
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
by the interests they were supposed to regulate.”277 One scholar concluded that the experience “weakened Americans’ faith in expertise.”278
Americans’ confidence in the administrative state had been further weakened by the struggle against fascism. Before the war, many
American political scientists had studied the governance of Nazi
Germany and Fascist Italy with great interest and even admiration.279 Many more expressed not illegitimate fears that the United
States would be left behind if it did not, in some sense, emulate these
rising foreign powers.280
During the war, such scholars either abandoned their earlier
views or fell into disrepute.281 There was a widespread belief among
the public that “administrative power could pave the road to totalitarianism” and “ ‘the claim that imposing the rule of law on agency
behavior could protect Americans from an administrative state run
amok . . . was increasingly heard across the political spectrum.’ ”282
Despite the disgust with which the fascist governments were regarded after the war, policymakers in the military and foreign affairs
establishments were well aware that the United States had, in one
sense, needed to beat the enemy at its own game in order to win the
war. Only through aggressive centralization, militarization, and, often, lack of transparency—hallmarks of fascist governance—had the
United States managed to mobilize its resources so quickly and comprehensively during the early 1940s to meet the challenge of defeating Hitler, Mussolini, and Imperial Japan.283 Pendleton Herring, a
prominent theorist of governmental organization highly influential in
the Roosevelt and Truman Administrations, had argued that “totalitarian states can be opposed only through an equally effective mobilization of resources.”284 Twenty-six new agencies with broad powers
had been created during the war.285 Moreover, the United States
277. Schiller, supra note 217, at 195; see also id. at 201 (“Too often American wartime
agencies had the appearance of incompetent bullies, captured by special interests, acting
with an autocratic disregard of due process.”).
278. Schiller, supra note 237, at 92.
279. See, e.g., Lawrence Dennis, Fascism for America, ANNALS AM. ACAD. POL. & SOC.
SCI., July 1935, at 62, 62 (arguing that the New Deal was not fascist enough and that fascism is “the inevitable alternative to chaos or communism”); see also OREN, supra note 217,
at 47-49.
280. See OREN, supra note 217, at 47-49.
281. See id. at 46-52. Lawrence Dennis was tried for treason in 1944. See KATZNELSON,
supra note 40, at 30-31.
282. Kovacs, supra note 125, at 695-96 (quoting Reuel E. Schiller, Enlarging the Administrative Polity: Administrative Law and the Changing Definition of Pluralism, 1945–
1970, 53 VAND. L. REV. 1389, 1405 (2000)).
283. See Aziz Rana, Who Decides on Security?, 44 CONN. L. REV. 1417, 1461-62 (2012).
284. PENDLETON HERRING, THE IMPACT OF WAR 14 (1941). For a definitive account of
Herring’s career and influence, see STUART, supra note 217, at 9, 27-32.
285. Schiller, supra note 217, at 190.
2014]
NATIONAL SECURITY RULEMAKING
929
could not have accomplished this task without allying itself with another totalitarian regime, the Soviet Union.286
More specifically, the disaster at Pearl Harbor had left an indelible mark on the thinking of policymakers at the top of the Roosevelt
and Truman Administrations.287 They, and ultimately the American
people as well, believed that America had been left vulnerable to
Japanese aggression by its inability to coordinate defense and military functions effectively.288 In addition, those in government responsible for national security had permitted isolationism among the public to influence critical decisions about foreign policy and national
defense in ways that weakened America’s position in the world.289
The same values that were praised by proponents of administrative
law reform that undergird the APA—transparency, legitimacy, deliberation, and accessibility, among others—were regarded as dangerous in the national security context.290
The threat of fascism and communism had therefore affected the
APA debate in complex, conflicting ways. Influential members of the
Roosevelt and Truman Administrations, such as Herring, Dean
Acheson (who had chaired Roosevelt’s Committee that proposed the
bill that became the APA), and George Marshall, adhered to a “Pearl
Harbor” view of government policymaking that treated national security matters as uniquely requiring a closed, militarized, and centralized process—just the opposite of the principles of transparency, public participation, and judicial oversight animating the APA.291
The year after Congress enacted the APA, it enacted the APA’s
mirror image—the National Security Act (NSA) of 1947.292 In contrast to the APA, the NSA, the crafting of which was very heavily influenced by Herring, was entirely animated by the Pearl Harbor
view.293 Congress debated it in the shadow of the harsh, bi-partisan
report on the Pearl Harbor attacks issued in mid-1946.294 The NSA’s
purpose was to centralize government decisionmaking in the national
security realm and, for the most part, insulate it from public scruti286. See KATZNELSON, supra note 40, at 32, 94-95 (describing the alliance and Americans’ ambivalence about the Soviet Union in the 1930s and 1940s).
287. See STUART, supra note 203, at 42.
288. Id. at 43, 70-71.
289. See id. at 39-40.
290. See id. at 70-71; Rana, supra note 283, at 1423-24 (noting “the steady emergence,
beginning during the New Deal, of the prevailing American idea of security, with its emphasis on professional expertise and insulated decision-making”).
291. See STUART, supra note 217, at 41-42, 70-71.
292. National Security Act of 1947, ch. 343, 61 Stat. 495.
293. See STUART, supra note 217, at 8-9.
294. David F. Rudgers, Creating the Secret State 138-39 (2000); see also J. Comm. on
the Investigation of the Pearl Harbor Attack, Investigation of the Pearl Harbor Attack, S.
Doc No. 79-244, at 252-53 (1946).
930
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
ny. Although the Act’s supporters were not able to accomplish their
goal of unifying the military services, the NSA created what became
known as the Department of Defense (DoD), the Joint Chiefs of Staff,
the National Security Council, and the CIA.295
The co-existence of these two radically different perspectives on
the lessons for the administrative state to be learned from World War
II—the Reform view animating the APA and the Pearl Harbor view
animating the NSA—are also reflected in the structure of the APA.
The national security exception enshrines the Pearl Harbor view
within the constitution of the modern administrative state.
The Pearl Harbor view’s influence on legislation and agency behavior was powerful, but not particularly prominent in the debates
on administrative reform.296 It was only years later that the government agencies benefitting most from the national security exception
would be required to defend it to Congress.297
In 1964, congressional committees first began to scrutinize the
national security exception and its justifications.298 By then, the Cold
War had replaced World War II as the main inspiration for national
security exceptionalism.299 And yet the arguments in favor of national
security administration remained the same. The difference is that, by
this point, they had been essentially abstracted from the post-World
War II historical context in which they had emerged and the debates
over fascism that had inspired them.
The justifications for the national security exception articulated
by agencies in 1964 are familiar ones. They appear frequently in cases and scholarly defenses of national security exceptionalism in general.300 The first rationale agencies submitted to Congress to justify
the exception is that national security often requires secrecy, which
is defeated by public participation.301 Second, the agencies argued,
the fluid nature of international relations requires agencies to react
quickly.302 Suppose, for example, that the DoD must respond to a
coup in a foreign country by immediately revising its rules regarding
295. See STUART, supra note 217, at 110-50.
296. See supra notes 287-291 and accompanying text.
297. Agencies responded to surveys about their use of the national security exception in
1957. See supra notes_130-131 and accompanying text.
298. See Letter from L. Niederlehner, Gen. Counsel of the Dep’t of Def., to James O.
Eastland, Chairman, Comm. on the Judiciary, U.S. Senate (July 23, 1964), reprinted in
Administrative Procedure Act: Hearings on S. 1663 Before the Subcomm. on Admin. Practice & Procedure of the S. Comm. on the Judiciary, 88th Cong. 492-93 (1964) [hereinafter
Administrative Procedure Act: Hearings on S. 1663].
299. See DUDZIAK, supra note 44, at 88-90.
300. See Knowles, supra note 215, at 105.
301. See Bonfield, supra note 22, at 280-82.
302. See id. at 279-80.
2014]
NATIONAL SECURITY RULEMAKING
931
military support for that country’s government.303 Notice-andcomment would make this impossible. Finally, because nations are
unitary actors on the world stage, the United States must “speak
with one voice” in foreign relations; agencies must act in a coordinated fashion and cannot be changing positions in response to public
comment without risking embarrassment and undermining U.S. foreign policy.304 For example, as a DOJ spokesman testified to Congress, “[a] requirement of public participation in . . . promulgation of
rules to govern our relationships with other nations . . . would encourage public demonstrations by extremist factions which might
embarrass foreign officials and seriously prejudice our conduct of
foreign affairs.”305
These justifications are variations on the familiar tropes that still
dominate national security and foreign relations law.306 They are
rarely challenged outside academia, and their origins and theoretical
justifications are rarely discussed. But those justifications are surprisingly brittle.307 As reasons for exempting national security rulemaking entirely from notice-and-comment, moreover, they fall short.
As Professor Arthur Bonfield observed, in some rare situations, truly
necessary secrecy would make notice-and-comment all but impossible.308 But this hardly supports removing notice-and-comment entirely from all national security rulemaking. Second, these hoary justifications rely on an illusion that the boundary between national security and ordinary rulemaking can be clearly delineated.309 Finally,
these justifications draw on overbroad principles that seek to shut
down, rather than encourage, policy debates. They are too often used
to avoid accountability and public participation in governance—even
in aspects of governance, such as rulemaking, in which such accountability and public participation are thought to be especially important.
303. Id. at 280.
304. Munaf v. Geren, 533 U.S. 674, 702 (2008) (“The Judiciary is not suited to secondguess such determinations . . . that would require federal courts to . . . undermine the Government’s ability to speak with one voice in this area.”); United States v. Curtiss-Wright
Exp. Corp., 299 U.S. 304, 319 (1936) (“[T]he President alone has the power to speak or
listen as a representative of the nation.”).
305. Administrative Procedure Act: Hearings on S. 1663, supra note 298, at 363.
306. See Knowles, supra note 215, at 111-12; Eric A. Posner & Cass R. Sunstein, Chevronizing Foreign Relations Law, 116 YALE L.J. 1170, 1202 (2007) (describing the justifications for special deference by courts in foreign relations as “often less textual than functional, based on traditional practices and understandings”).
307. See Knowles, supra note 215, at 121-23; David H. Moore, Beyond One Voice, 98
MINN. L. REV. 953, 979 (2014) (explaining why the one voice doctrine is “fatally flawed”).
308. See id. at 283.
309. See, e.g., Derek Jinks & Neal Kumar Katyal, Disregarding Foreign Relations Law,
116 YALE L.J. 1230, 1257-58 (2007) (describing the intractable boundary problems created
by attempting to apply different deference standards for foreign and domestic issues).
932
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
The next Part discusses why these tropes do not accurately reflect
the complexity of rulemaking in today’s world and what should be
done to reform national security rulemaking.
V. REFORMING NATIONAL SECURITY RULEMAKING
As the history makes clear, the modern American administrative
state was formed in the shadows of war and the existential threat
posed by the rise of totalitarian states. These regimes, by ignoring
transparency and public participation, could whip their own bureaucracies into line, leaving democracies at a disadvantage. Or so many
believed. In the face of this threat, the APA’s authors carved out
modes of national security rulemaking distinct from domestic ones.
This response was understandable, but it is not at all clear that it
was correct. One could argue that the United States and its allies
prevailed over their enemies because of, and not despite, their adherence to democratic values.310
And it is even less clear that these distinct modes of national security rulemaking are serving the United States well in the twenty-first
century. Few observers would argue that U.S. government agencies
are functioning as efficiently or effectively as they should be in the
national security realm. Obsession with secrecy, overclassification,311
unnecessary redundancy, lack of coordination, inter- and intraagency communication failures, and lack of accountability to the public—these are just a few of the criticisms regularly leveled at agencies
by observers inside and outside government.312 Legislative efforts to
reform the intelligence community after September 11—in part by
creating the Department of Homeland Security and the position of
Director of National Intelligence, who would, in theory, coordinate
intelligence-gathering—was followed by massive growth in the number and type of agencies handling secret and top-secret intelligence
and simply made these problems worse.313
Moreover, several factors have led to increasing entanglement of
the government’s national security policies with the lives of ordinary
310. See, e.g., President Franklin D. Roosevelt, Third Inaugural Address (Jan. 20,
1941) (contending that “democracy alone, of all forms of government, enlists the full force
of men’s enlightened will. . . . [I]t is the most humane, the most advanced, and in the end
the most unconquerable of all forms of human society”).
311. See supra notes 22-26 and accompanying text.
312. See, e.g., PRIEST & ARKIN, supra note 24, at 95, 125 (describing two such examples); see generally STEPHEN HOLMES, THE MATADOR’S CAPE: AMERICA’S RECKLESS
RESPONSE TO TERROR (2007) (contending that the failure by government agencies to follow
established rules and be transparent resulted in numerous policy disasters following
September 11).
313. See PRIEST & ARKIN, supra note 24, at 92, 95 (describing the creation of the Office
of the Director of National Intelligence and the belief among “leaders of the intelligence
agencies” that this restructuring “extremely reckless”).
2014]
NATIONAL SECURITY RULEMAKING
933
Americans. Globalization continues to blur, and even erase, the distinction for Americans between what is foreign and what is domestic.314 The terrorist threat emerges primarily from small groups and
individuals, rather than nation-states; in response, government more
closely monitors the lives of individuals to learn about and stop these
threats.315 Agencies’ national security rulemaking therefore falls
more often into the category of “substantive” or “legislative” than ever before: because it directly affects the public, it was meant to be
conducted through notice-and-comment procedures so that the public
can be involved.
For the founders of modern American administrative law, this
state of affairs would be seen as the worst of both worlds. National
security rulemaking today lacks both the democracy and accountability-enhancing features of ordinary rulemaking and the centralization
and efficiency that were supposed to characterize rulemaking in the
areas of military and foreign affairs.316 It is hard to justify maintaining broad substantive exceptions from notice-and-comment requirements for national security rulemaking when agencies conducting such
rulemaking are largely inefficient, uncoordinated, and ineffective.317
In light of these changes, the traditional distinction between national security and ordinary rulemaking must be re-examined. U.S.
foreign and national security policy, as well as ordinary policy, can in
many, if not most, instances best be furthered by greater transparency, public participation, and deliberation—values that the APA’s notice-and-comment rulemaking procedures advance.
The APA’s national security exception and the model of agency
behavior that it signifies are not serving the administrative state
well. The exception cannot be justified in its current form, and it
should be modified to better reflect the requirements of regulation
today. However, the best answer may not be simply making requirements for national security rulemaking identical to those for ordinary
rulemaking. The optimal reforms will encourage more openness in
rulemaking while recognizing the reality that agencies conducting
national security rulemaking have the ability, through classification
authority, to hide their activities from view. Classification reform is
314. See, e.g., Jinks & Katyal, supra note 309, at 1258-59; Knowles, supra note 215, at
110-20.
315. See Aziz Z. Huq, Structural Constitutionalism as Counterterrorism, 100 CALIF. L.
REV. 887, 908 (2012) (describing the shift from “state-based enemies” to “new threats in the
more fragmented international environment”); Neil M. Richards, The Dangers of Surveillance, 126 HARV. L. REV. 1934, 1938 (2013) (observing that democracies have “invested
heavily in surveillance technologies in the aftermath of the September 11 attacks in America, the London subway bombings of 2005, and other atrocities”).
316. See supra notes 283-299 and accompanying text.
317. See Huq, supra note 315, at 905-11.
934
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
also desperately needed, but in the past this has proven very difficult
to implement successfully.318
In fact, legislative reform in general would be a challenge. Amending the APA, given its quasi-constitutional status, is an immensely
difficult legislative task.319 And the Vermont Yankee doctrine generally prohibits courts from imposing additional rulemaking procedures
beyond those required by the APA.320
Instead, reform is more likely to take place within the executive
branch. In the past, this has been the primary means by which limitations have been placed on the national security administrative
state.321 In response to loss of public confidence and increased scrutiny after national security scandals, administrations have sought to
pre-empt legislative solutions through the use of executive orders and
regulations.322 To enact comprehensive national security rulemaking
reform, the President could issue an executive order that, like E.O.
12,866, requires additional procedural requirements for rulemaking
by non-independent federal agencies. Such an executive order would
legally bind those agencies. And it could be enforceable in court—
especially if the order itself so provided.323
One risk, of course, is that executive branch reforms are more easily undone than legislation. National security agencies, in particular,
have a tendency to revert, over time, to more insular and opaque
modes of operation as they seek to expand their power within the bureaucracy.324 Nonetheless, many past executive branch reforms have
had “stickiness”: they have persisted and become part of agency culture.325 In any event, although they may be second-best solutions,
they are better than the status quo.
318. D.A. Jeremy Telman, Intolerable Abuses: Rendition for Torture and the State
Secrets Privilege, 63 ALA. L. REV. 429, 443-45 (2012).
319. William R. Andersen, Chevron in the States: An Assessment and a Proposal, 58
ADMIN. L. REV. 1017, 1033 (2006) (“The federal APA would be difficult to amend because it
has acquired something like constitutional status . . . . ”).
320. See Vt. Yankee Nuclear Power Corp. v. Natural Res. Def. Council, Inc., 435 U.S.
519, 543-48 (1978).
321. The executive often undertakes its own reforms to forestall more sweeping reforms from Congress. See, e.g., supra notes 129-32 and accompanying text; see also Neal
Kumar Katyal, Internal Separation of Powers: Checking Today’s Most Dangerous Branch
from Within, 115 YALE L.J. 2314, 2341 (2006) (arguing that, with respect to the war on
terror, the collapse of external checks on executive power demonstrates the need for internal checks on that power).
322. See supra notes 129-138 and accompanying text.
323. See City of Carmel-by-the-Sea v. U.S. Dep’t of Transp., 123 F.3d 1142, 1166 (9th Cir.
1997) (reviewing and measuring agency action against the standards in an executive order).
324. See Dalal, supra note 132, at 83-84.
325. A prominent example is OIRA review under Executive Order 12,866, 3 C.F.R. § 638
(1993), reprinted as amended in 5 U.S.C. § 601. See supra notes 126-139 and accompanying text.
2014]
NATIONAL SECURITY RULEMAKING
935
A. Eliminating the National Security Rulemaking Exception
To begin with, it is worth exploring whether the national security
exception should be eliminated altogether. This change’s impact on
rulemaking would be vast. For the large number of agencies that,
unlike DoD, currently lack internal rules requiring notice-andcomment, their non-secret rulemaking would be opened up to public
input for the first time. In addition, many other departments or
agencies, particularly within the Treasury and Commerce Departments, that do, or might be disposed to, rely on the national security
exception would be required to undertake notice-and-comment for
many types of rules that had been previously exempt.
For those critics of notice-and-comment rulemaking, of course, this
would be moving in precisely the wrong direction, as the burden imposed on these agencies would likely be a substantial one. A complex
rule can elicit thousands of comments or more, requiring hundreds of
pages to address in the final rule’s statement of basis and purpose.326
Indeed, these critics argue that the procedure is not worth its cost in
any policy area.327 In particular, they argue, notice-and-comment may
have an anti-regulatory bias, imposing unacceptably high burdens on
government agencies when swift and decisive regulatory action is
necessary to address urgent problems.328
Moreover, even if one believes that notice-and-comment’s benefits
exceed its costs in ordinary rulemaking, national security matters
remain unique in ways that may make notice-and-comment still inappropriate—even though many of the particular concerns that drove
the creation of the national security administrative state no longer
exist. For one thing, the stakes are quite often higher where national
security is concerned, which could make an anti-regulatory bias particularly problematic.329 For example, if the National Highway Traffic
Safety Administration fails to promulgate a rule requiring more auto
safety testing, this may, over time, cost lives. But if government
agencies fail to properly regulate the disposal of nuclear material
that falls into terrorists’ hands, the consequences could present an
existential threat to the United States.330
In addition, even in a globalized world, where special interest
groups often fiercely lobby national security policymakers, secrecy—
326. See, e.g., Pierce, supra note 68, at 119-20.
327. See supra notes 88-91 and accompanying text.
328. Pierce, supra note 68, at 118-19.
329. Huq, supra note 315, at 924-25 (observing that administrative “procedures can
function as frictions on desirable agency action,” and that the APA’s “deregulatory bias . . .
may have serious unintended consequences in the national security domain”).
330. See PHILIP BOBBITT, TERROR AND CONSENT: THE WARS FOR THE TWENTY-FIRST
CENTURY 201-02 (2008) (stating that an existential threat “is precisely the sort of threat
that terror poses”).
936
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
or at least some degree of opacity—still has an important place in
national security policymaking. Suppose, as one critic has asked us to
imagine, that the notice-and-comment requirements were imposed on
decisions about military support for Taiwan and its effect on relations with China or the rules governing targeting of Al Qaeda
safehouses.331 This would expose to public (indeed, global) scrutiny
the inner workings of the government’s military and diplomatic machinery, making it much more difficult for the United States to effectively eliminate dangerous enemies or to conduct the delicate giveand-take at the heart of diplomacy.
However, in measuring notice-and-comment’s value for national
security rulemaking, it is important to take account of all the ways in
which the APA and other laws insulate these types of agency decisionmaking from public scrutiny. First, even without the national
security exception, administrative law would treat the two examples
above quite differently. Although rulemaking regarding military aid
to foreign countries would arguably require notice-and-comment, military-targeting rulemaking—along with many other rules regarding
the conduct of military operations abroad—would almost certainly
fall within the exclusion from the APA of “military authority exercised in the field in time of war or in occupied territory.”332
Second, as Professor Evan Criddle has persuasively argued, at
least some of these concerns can be addressed through the APA’s
“good cause” exception, which authorizes rulemaking without noticeand-comment when it would be “impracticable, unnecessary, or contrary to the public interest.”333 In particular, agencies could rely on
previous cases in which courts have approved agencies’ invocation of
the exception as “contrary to the public interest” when undertaking
notice-and-comment would thwart the rule’s purpose.334 Similarly,
when national security rules issued during a crisis situation must
take effect immediately, notice-and-comment rulemaking could be
considered by a court to be “impracticable.”335
At the same time, requiring agencies to rely on the “good cause”
exception instead of the national security exception would discipline
agency decisionmaking by requiring the agency to articulate a legal
basis for departing from the ordinary APA process. The APA provision establishing the “good cause” exception requires that the agency
explicitly invoke the exception when issuing the rule and explain why
331.
332.
333.
334.
335.
Pierce, supra note 68, at 119-20.
5 U.S.C. § 551(1)(G) (2012); see Kovacs, supra note 125, at 674-76.
5 U.S.C. § 553(b)(3)(B) (2012); see Criddle, supra note 34, at 192-93.
See supra notes 144-154 and accompanying text.
See 5 U.S.C. § 553(b)(3)(B); GAO REPORT, supra note 121, at 16.
2014]
NATIONAL SECURITY RULEMAKING
937
the exception applies.336 Courts have generally voided rules when the
agency failed to follow this requirement.337
However, one risk in relying on the “good cause” exception in place
of the national security exception is that it could inspire agencies to
expand the breadth of rulemaking exempt for good cause, turning the
exception into something like the all-purpose escape hatch from notice-and-comment that some critics (mistakenly) fear it already is.338
As the GAO report reveals, the most common reason agencies rely on
the “good cause” exception is that Congress has either mandated the
text of the rule or has imposed restrictions making notice-andcomment impossible.339 It is much less common for an agency to cite
an “emergency” as the reason for invoking the exception.340 But if the
national security exception were no longer available, it is easy to see
how agencies would be tempted to broaden significantly the definition of “emergency” and expand the category of rules for which notice
would be “contrary to the public interest.”
Finally, and perhaps most importantly, whether or not agencies
are relying on the “good cause” exception, eliminating the national
security exception would leave in place agencies’ power to classify
information, significantly reducing the plausibility of notice-andcomment for many rules.341 Imagine a notice of proposed rulemaking
in which the most important and specific reasons for the proposed
rule are redacted. Of course, the simple knowledge that an agency is
conducting rulemaking in a particular area may be useful to the public. This enables concerned citizens or interest groups to alert Congress, which may have the ability to further scrutinize or influence
the agency’s process.342 Outside parties could use the notice as a
means for crafting more effectively targeted FOIA requests, possibly
leading to the release of more specific information or, at the very
least, requiring the agency to articulate in court its reasons for classifying the redacted information. It is also possible, however, that in
some circumstances an agency could simply determine that the fact
that rulemaking is taking place should be classified as well.
336. See 5 U.S.C. § 553(b)(3)(B) (requiring that an agency invoking the “good cause”
exception must “incorporate[] the [good cause] finding and [include] a brief statement of
reasons therefor in the rules issued”).
337. See, e.g., Buschmann v. Schweiker, 676 F.2d 352, 356 (9th Cir. 1982); Bohner v.
Daniels, 243 F. Supp. 2d 1171, 1176 (D. Or. 2003), aff’d sub nom. Paulsen v. Daniels, 413
F.3d 999 (9th Cir. 2005).
338. See supra notes 148-154 and accompanying text; Boliek, supra note 152, at 3343.
339. See GAO REPORT, supra note 121, at 16-17.
340. See id.
341. See supra notes 22-23 and accompanying text.
342. See supra notes 85-87 and accompanying text.
938
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
The authority to classify information is a powerful one. Agencies,
like most organizations, respond to new challenges using familiar
tools. There is a risk that, if agencies are forced to open up their deliberations via notice-and-comment, they may be inclined to push
back by ramping up classification to keep those deliberations effectively closed.343 Therefore, a potential difficulty with eliminating the
national security rulemaking exception is that it could, perversely,
result in a process even less accessible to the public.
B. Encouraging Notice-and-Comment Rulemaking as a Best Practice
If eliminating the national security exception would impose unacceptable costs by burdening agencies, encouraging over-classification,
and failing to take account of the uniqueness of national security
rulemaking, an alternative approach that avoids these problems
would be to encourage agencies to use notice-and-comment for national security rulemaking whenever possible. Under this approach,
an executive order would mandate notice-and-comment as the default
mode for national security rulemaking but allow each agency to
choose to opt out for particular types of rules.344 This approach would
require agencies to assess the costs and benefits of notice-andcomment for each category of national security rule and make a determination about whether notice-and-comment will be used for that
category. It would give agencies sufficient flexibility to permit them
to take account of the uniqueness of national security policies but
would bring more regularity and deliberation to the process of departing from notice-and-comment.
The value of this approach depends in part on the accuracy of the
ex ante determinations agencies could be expected to make about
each type of rule’s suitability for notice-and-comment. For some types
of rules, the decision could be a relatively easy one. Some agencies
that regularly make rules in the national security realm already engage in notice-and-comment for certain types of rules. The State Department, for example, uses notice-and-comment in developing most
343. Cf. Eric J. Sinrod, Freedom of Information Act Response Deadlines: Bridging the
Gap Between Legislative Intent and Economic Reality, 43 AM. U. L. REV. 325, 342 (1994)
(describing agencies’ increasing use of special exceptions to avoid disclosure following a
sharp increase in FOIA requests).
344. Cf. Evan J. Criddle, Fiduciary Administration: Rethinking Popular Representation
in Agency Rulemaking, 88 TEX. L. REV. 441, 480-81 (2010) (advocating for a complete repeal of the national security exception).
2014]
NATIONAL SECURITY RULEMAKING
939
International Traffic in Arms Regulations (ITAR)345 and rules implementing the Hague Convention on Intercountry Adoption.346
However, in other contexts, agencies will often face difficulty predicting whether in any particular instance notice-and-comment
would harm national security interests. For example, one can imagine few harmful national security consequences flowing from using
notice-and-comment to promulgate rules implementing a trade
agreement with Great Britain or another liberal democratic ally. On
the other hand, given the political volatility of U.S. relationships in
the Middle East, using notice-and-comment for rules implementing a
similar agreement with a nation in that region could have more negative national security consequences.
Lacking certainty about how notice-and-comment would affect the
process of national security policymaking for a particular rule, agencies may be inclined to be cautious and avoid it—despite a mandate
that they adopt it whenever possible. On the other hand, so long as
the agency is allowed to change its mind and eschew notice-andcomment should it prove unworkable, agencies could be successfully
nudged to experiment with notice-and-comment in areas where it has
not been used.
A second factor of equal importance is agency culture. Regardless
of its value, will agencies voluntarily cede some degree of control over
the rulemaking process to regulated parties? In general, agencies
tend to function like fiefdoms—without external pressure to do otherwise, they hoard power, avoid oversight, and provide only ex post
rationales for their decisionmaking.347 Agencies making rules in the
national security area are especially likely to operate this way because their mandate to protect national security provides justification for rejecting the ordinary administrative law values of transparency and accessibility.348
345. See, e.g., Amendment to the International Traffic in Arms Regulations: Initial
Implementation of Export Control Reform, 78 Fed. Reg. 22,740, 22,740-01 (Apr. 16, 2013)
(to be codified at 22 C.F.R. pts. 120, 121, 123).
346. See Hague Convention on Intercountry Adoption; Intercountry Adoption Act of
2000; Accreditation of Agencies; Approval of Persons, 71 Fed. Reg. 8064 (Feb. 15, 2006) (to
be codified at 22 C.F.R. pt. 96); see also Anjanette Hamilton, Comment, Privatizing International Humanitarian Treaty Implementation: A Critical Analysis of State Department
Regulations Implementing the Hague Convention on Protection of Children and Cooperation in Respect of Intercountry Adoption, 58 ADMIN. L. REV. 1053, 1073 (2006).
347. See, e.g., Bruce Ackerman, The New Separation of Powers, 113 HARV. L. REV. 633,
700 (2000) (describing the politicized federal agencies as “unruly bureaucratic fiefdoms”);
Dalal, supra note 132, at 102.
348. See Dalal, supra note 132, at 99-100 (describing agencies given the “national security
mandate” as having stronger incentives to gather more power to carry out that mandate).
940
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
On the other hand, there is evidence that agencies often do voluntarily seek to learn about best practices and adopt them.349 In many
instances, a “nudge”350 may be all that is required to persuade an
agency to use notice-and-comment more often. In addition, even power-hungry agencies can benefit from increased use of notice-andcomment. The act of opening up the rulemaking process to the public—particularly in response to criticism—has a strong legitimizing
effect on the rules that are produced and the agency itself. This is
why, for example, the Department of Homeland Security has chosen
to use notice-and-comment in developing new rules for Advanced Imaging Technology at airports, despite the strong probability that it
could successfully invoke either the good cause or national security
exceptions to avoid notice-and-comment.351 Agencies may rationally
conclude that the legitimacy gained is worth the cost of increased
public awareness.
Agencies also may rationally conclude that voluntarily adopting
notice-and-comment is actually a way to avoid further congressional
scrutiny and possible legislative reforms. In fact, executive branch
reform is a strategy the President and DoD have used in the past to
head off more restrictive legislation during periods of intense public
criticism. The post-Watergate revelations of domestic spying by the
CIA and other national security-related abuses, documented in the
Church Committee Hearings and Reports, led to the passage of the
Foreign Intelligence Surveillance Act.352 But the President preempted
other types of legislative restrictions—for example, by imposing a
ban on assassinations by executive order, which arguably contained
sufficient ambiguity to allow the President to order an assassination
in some circumstances.353 And, similarly, as discussed above, during
the same reform-infused era, the Pentagon shrewdly deflected the
349. See David Zaring, Best Practices, 81 N.Y.U. L. REV. 294, 297-98 (2006) (identifying
a trend toward “best practices,” a form of regulation “in which regulated entities experiment with best practices as a way of vindicating the broad principles of various regulatory
programs, while the regulators keep track of their progress and help to celebrate and publicize particularly successful local initiatives”).
350. See RICHARD H. THALER & CASS R. SUNSTEIN, NUDGE: IMPROVING DECISIONS
ABOUT HEALTH, WEALTH, AND HAPPINESS 7-8 (2008) (proposing that regulators take account of cognitive biases to craft more effective incentives for regulated parties, including
default rules that individuals are free to opt out of).
351. The agency had not relied on these exceptions earlier. See supra notes 157-160
and accompanying text.
352. See Richard Henry Seamon & William Dylan Gardner, The Patriot Act and the
Wall Between Foreign Intelligence and Law Enforcement, 28 HARV. J.L. & PUB. POL’Y 319,
334-36 (2005).
353. William C. Banks & Peter Raven-Hansen, Targeted Killing and Assassination:
The U.S. Legal Framework, 37 U. RICH. L. REV. 667, 717-20 (2003) (contending that the
assassination ban contained in President Ford’s 1976 Executive Order 11,905 extended
only to assassinations, for their political views, of officials of foreign nations with which the
United States was not at war, and that the president may waive the ban).
2014]
NATIONAL SECURITY RULEMAKING
941
Administrative Conference’s call for elimination of the national security exemption by adopting notice-and-comment for rules “having a substantial effect upon the public” yet reserving authority to opt out.354
The skill with which agencies have used selective internal reforms
to stave off legislative reform suggests that simply encouraging agencies to use notice-and-comment cannot, by itself, fully transform the
way national security rulemaking is conducted. Indeed, it may be
that the types of rulemaking in this area most crippled by insularity
and lack of deliberation are those that agencies would be least likely
to want to expose to public scrutiny. A nudge will not always
be enough.
C. Establishing a Chenery-Type Rule for Invoking the National
Security Exception
Notice-and-comment would not be effective without the courts’ enforcing the APA requirement and, under Chenery I, requiring agencies
to sink or swim with the rationales they articulate during the rulemaking process. If agencies are required or strongly encouraged to use
notice-and-comment for national security rulemaking, the courts will
necessarily play a larger role in overseeing national security policy.
For the many critics who believe the courts are far too deferential in
national security matters, this would be a welcome development.
However, to the extent that the national security exception remains a viable option for agencies, the way in which courts treat an
agency’s invocation of the exception must also be reformed. Currently, courts allow agencies to get away with remaining silent during the
rulemaking process, issuing a rule without notice-and-comment and,
when a regulated party challenges the rule on the ground that it was
promulgated without notice-and-comment, invoking the national security exemption as a get-out-of-jail-free card.355
This permissiveness by courts encourages agencies conducting national security rulemaking to neglect notice-and-comment. And relatedly, when an agency is aware that it need not invoke the exception
during rulemaking, it will have little incentive to articulate reasons
why the exception should apply. The bottom line is that, absent unusual public or congressional attention, agencies face little pressure to
seriously deliberate about the boundaries of the national security exception and have strong incentives to rely on it whenever it could
possibly be applicable. The likely result is a largely unmentioned but
broad exception that becomes broader and broader until the very rare
case where a court determines that it does not apply.
354. See supra notes 131-138 and accompanying text.
355. See supra notes 161-165 and accompanying text.
942
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
To avoid this ever-expanding national security exception and ensure that national security rulemaking is undertaken effectively,
courts must play the role Chenery recognized for them—to test the
legality of the rationales the agency actually used to produce a rule
while it was being produced. The notice-and-comment process is fundamental to the proper functioning of the administrative state. A decision by an agency to eschew notice-and-comment should have a legally authorized basis that the agency can articulate during the
rulemaking process.
To encourage the serious deliberation that is a quality of all effective rulemaking, agencies should be required to specifically invoke
the national security exception contemporaneously with the rulemaking process and state with specificity their reasons for invoking the
exception. In addition, courts should apply a Chenery-type rule to the
exception, strictly enforcing this new requirement by refusing to consider the applicability of the exception unless the agency invoked it
during rulemaking. Moreover, courts should consider the exception’s
applicability only on the grounds articulated by the agency.356
It could be a challenge for the courts to impose such a Chenerytype rule on their own. There is not much of a textual peg for it. In
stark contrast to the “good cause” exception, the APA’s text does not
require agencies to specifically invoke the national security exception
when they proceed without notice-and-comment. As I discuss above,
Chenery does have a constitutional dimension, however, which provides it with roots deeper than the APA’s text: it is a means of enforcing the non-delegation doctrine.357 It is true that the courts have long
recognized that the non-delegation doctrine is much weaker in foreign affairs and national security matters.358 In a series of cases, most
prominently United States v. Curtiss-Wright Export Corp.,359 the
courts permitted the President to engage in lawmaking in foreign
affairs that would not have been permissible in the domestic realm.360
Nonetheless, there must be some limit to the President’s power to
make law, even where national security is concerned. To conclude
otherwise is to make the error Carl Schmitt made: if national security exceptionalism has no discernible limits, it is no longer an exception but the rule.361 Wherever that limit is must, in the end, be de356. See supra notes 109-115 and accompanying text (discussing the Chenery rule).
357. See supra notes 111-115 and accompanying text.
358. See, e.g., Gonzalez v. Reno, 212 F.3d 1338, 1349 (11th Cir. 2000) (“[T]he authority of
the executive branch to fill gaps is especially great in the context of immigration policy.”).
359. 299 U.S. 304 (1936).
360. See id. at 319-22. Curtiss-Wright is controversial and has been the subject of
“withering criticism.” HAROLD HONGJU KOH, THE NATIONAL SECURITY CONSTITUTION:
SHARING POWER AFTER THE IRAN-CONTRA AFFAIR 94 (1990).
361. See supra Part IV.
2014]
NATIONAL SECURITY RULEMAKING
943
termined by the courts. In some recent cases concerning the war
against terrorist organizations, the Supreme Court seems to have
recognized this reality.362 Asking courts to better police the boundaries of the national security exception is another way in which the
courts should ensure that the administrative state operates within
constitutional bounds.
Even if the courts will not recognize the constitutional necessity of
a Chenery-type rule, and Congress will not reform the APA to require
it, the President should impose it on agencies through an executive
order. Whether required by the Constitution or the spirit of the APA,
procedural reforms must be implemented for the national security
administrative state to be effective. And it is the President’s duty as
Commander-in-Chief to ensure that reform happens; the security of
the nation depends on it.
VI. CONCLUSION
National security rulemaking desperately needs less secrecy and
more public participation. Along with classification reform, the more
frequent use of notice-and-comment procedures, enforced by the
courts through a Chenery-type rule, can help immensely. Notice-andcomment deserves the fulsome praise it has received. A keystone of
the APA’s “bill of rights” for those affected by agency regulation, it
encourages deliberative decisionmaking, increases accountability, ensures public participation, and legitimates the final agency decision.363
In short, notice-and-comment rulemaking enhances democratic values.
Notice-and-comment rulemaking has also grown more popular as
technological advances continue to make it easier for the public to
comment on proposed regulations and the agency to respond. In fact,
the United States government now proselytizes worldwide for
the procedure.364
However, notice-and-comment rulemaking turns out to be yet another practice that the United States urges for other nations while
reserving for itself a healthy dose of American exceptionalism. Secrecy enshrouds a large portion of the federal bureaucracy performing a
broad range of national security functions.365
The APA’s national security exception plays a more important role
in establishing and maintaining the national security administrative
362. See Knowles, supra note 215, at 91-92 (discussing the Supreme Court’s refusal to
defer to the government’s claims of military exigency in the enemy combatant cases such as
Hamdan v. Rumsfeld, 548 U.S. 557 (2006)).
363. See supra Part II.A–B.
364. See Jeffrey S. Lubbers, The Transformation of the U.S. Rulemaking Process—For
Better or Worse, 34 OHIO N.U. L. REV. 469, 471 (2008); Mendelson, supra note 41, at 1344.
365. See, e.g., PRIEST & ARKIN, supra note 24, at 22.
944
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:883
state than its sporadic invocation by agencies might suggest. To be
sure, it provides a useful escape hatch for agencies whose rules are
challenged in court. But more importantly, it is the formal legal incarnation of the culture of insularity that dominates national security rulemaking. And because the exception informs the rule, understanding the scope of this widespread avoidance of notice-andcomment rulemaking on national security grounds is critical for understanding how the American administrative state operates as
a whole.
Our bifurcated administrative state evolved from Americans’ ambiguous views toward fascism and the New Deal. It may have been a
useful way to address some of that era’s challenges. But today, the
existence of two overlapping administrative law regimes often encourages agencies to select the one subjecting them to less scrutiny.
And in contexts like the NSA surveillance programs and the postSeptember 11 Alien Registration Program, the use of notice-andcomment would not only have made the agencies more effective in
achieving their goals, but it also would have protected the rights of
minorities and the American people in general.366 The present arrangement simply does not make sense in a world where the foreign
and the domestic are not only intertwined, but are often indistinguishable. As the NSA surveillance programs reveal, national security rulemaking regulates our lives as intimately as any other kind. It
deserves our sustained attention and participation.
366. See supra notes 10-12, 50-55 and accompanying text.
RAISING THE FLOOR OF COMPANY CONDUCT:
DERIVING PUBLIC POLICY FROM THE CONSTITUTION
IN AN EMPLOYMENT-AT-WILL ARENA
STEVEN J. MULROY* & AMY H. MOORMAN
I. INTRODUCTION ..................................................................................................
II. THE PUBLIC POLICY EXCEPTION TO AT-WILL EMPLOYMENT .............................
III. CONSTITUTIONAL SOURCE OF PUBLIC POLICY VERSUS OTHER LEGAL
SOURCES ...........................................................................................................
A. Generally ....................................................................................................
B. State versus Federal Constitutions ............................................................
C. Treating Private Action as State Action ....................................................
IV. THE RELEVANCE OF EXISTING STATUTORY REMEDIES......................................
V. SPECIFIC CONSTITUTIONAL PROVISIONS AS CANDIDATES FOR THE PPE ...........
A. Provisions that Should Not Form the Basis of a PPE Claim ....................
1. “Structural” Constitutional Doctrines .................................................
2. Restrictions Uniquely Focused on Government ...................................
3. Due Process/Criminal Procedure Safeguards .....................................
B. Constitutional Provisions that Should Form the Basis of a PPE Claim...
1. Free Speech: Political Expression ........................................................
(a) Forced Speech. ..............................................................................
(b) Whistleblowing .............................................................................
(c) Coverage Inspired by Statutory Speech Protection for Private
Employees ......................................................................................
(d) Toward a Balance of Employee and Employer Interests .............
2. Fourth Amendment Privacy .................................................................
3. Equal Protection ..................................................................................
4. Free Exercise ........................................................................................
C. Closer Cases ...............................................................................................
1. Reproductive Rights And Gender ........................................................
2. Substantive Due Process ......................................................................
3. Second Amendment..............................................................................
VI. CONCLUSION .....................................................................................................
945
951
956
956
962
963
965
972
973
973
973
975
979
979
982
983
986
988
992
996
999
1001
1001
1005
1006
1008
I. INTRODUCTION
Consider an employer that requires all employees to attend rallies
for a particular presidential candidate, or to sign petitions supporting
that candidate, or to lobby government for a particular position. Employees who fail to comply are subject to denial of bonuses, postponement of promotions, or even termination.
If the employer were a government agency, this would clearly violate the employees’ free speech rights under the First Amendment.1
* Steve Mulroy would like to acknowledge with gratitude the comments of Profs. Lee
Harris, Ernie Lidge, Andrew McClurg, and Eugene Shapiro of the Cecil C. Humphreys School
of Law, University of Memphis, and the research assistance of Alex Hall and Logan Klauss.
1. See, e.g., Branti v. Finkel, 445 U.S. 507, 517-19 (1980) (stating that with the exception of certain instances where political affiliation would “interfere with the discharge of
his public duties . . . continued employment of an assistant public defender cannot properly
be conditioned upon his allegiance to the political party”); Elrod v. Burns, 427 U.S. 347
946
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
In some states, a private employee might have some sort of statutory
protection.2 But, in the many cases where such explicit statutory protection is lacking, the vast majority of private employees would have
no remedy. That is so because, for the last century or two, the general
rule has been that absent explicit contract language to the contrary,
employment in the private sector is “at will.”3
However, in recent decades, courts have, on some occasions, applied a common law doctrine known as the “public policy exception”
(PPE) to the at-will employment rule to protect employees in extreme
cases when the employer’s conduct runs contrary to a definite public
policy. Originating in such relatively clear-cut cases, as where an
employer fired an employee for refusing to violate the law,4 or for
complying with a legal obligation like jury service,5 the exception has
expanded into other areas where the employer’s adverse employment
action seems contrary to a clear, widely accepted public policy in the
state. The doctrine stands as a last line of defense for an employee
who cannot point to express statutory text, explicit contractual language, or specific facts creating an implied contract. It creates a
“floor” for fair treatment by employers to which all private employees
are entitled. To the extent one considers such a floor necessary to
safeguard important public policies, then, the PPE is important.
Courts vary in their application of this doctrine in wrongful discharge claims. In the “forced speech” example above, which has actually occurred and is not strictly hypothetical, some courts have
granted the employee relief while others have not.6 While courts that
recognize the PPE may theoretically allow use of constitutional principles to inform their determination of what is clear and substantial
public policy within their states, there has been less enthusiasm for
using constitutional principles in PPE cases as opposed to statutory
(1976) (holding termination of public employees based on party affiliation unconstitutional
under the First and Fourteenth Amendments).
2. See infra Part V.B.1.c (discussing state statutory protections of free speech rights
of private employees).
3. A noteworthy exception to the general rule exists in the State of Montana, which
statutorily protects employees from being discharged absent good cause. MONT. CODE
ANN. § 39-2-904 (2013); see also infra notes 233-38 and accompanying text (describing state
statutes and case law providing protection against termination for off-duty, off-premises
conduct). Cf. ARIZ. REV. STAT. ANN. § 23-1501 (2013) (regulating employment termination
in the private sector).
4. See, e.g., Martin Marietta Corp. v. Lorenz, 823 P.2d 100, 109 (Colo. 1992) (recognizing a public-policy exception where “the employer directed the employee to perform an
illegal act as part of the employee’s work related duties”).
5. See, e.g., Nees v. Hocks, 536 P.2d 512, 515-16 (Or. 1975) (concluding that employer’s termination of employee for refusing to be excused from jury duty was driven by a “socially undesirable motive” which created a viable cause of action for wrongful discharge
based on public policy).
6. See infra Part V.B.1.a (discussing such “forced speech” cases).
2014]
COMPANY CONDUCT
947
or regulatory rules.7 In large part, this is because constitutional principles are generally considered to apply only where there is state action, and thus may be seen as less directly analogous to the setting of
a private employment dispute.8
Such a dilemma can occur in many different situations where private employers fire employees under circumstances at odds with confirmed policy behind a constitutional principle. For example, cases
have arisen in which employers allegedly fired employees for having
an abortion.9 Other cases have involved other types of “free speech”
issues, such as when employers forbade workers from speaking out
on matters of public concern,10 forced them to engage in political
speech with which they disagreed,11 or pressured them into making
political contributions.12 Controversies also have arisen over private
employer decisions to fire employees for wearing long hair,13 for
smoking off-duty and off-premises,14 and for drinking off-duty and offpremises.15 Such cases indicate that employees expect their fundamental rights to be respected in the workplace.
Similarly, society places high value on the right to vote, the freedom of worship, and the right to be free from unwarranted intrusion
into intimate matters. When employers use the coercive power at
their disposal to undermine these values, one might be inclined to
7. See infra Part III.A (discussing the use of constitutional versus other sources of
law in determining public policy under the PPE).
8. See infra Part III.A (discussing the use of constitutional versus other sources of
law in determining public policy under the PPE).
9. Barbara Presley Noble, A Firing for Cause, or for Abortion?, N.Y. TIMES, Jan.
16, 1994, at F23. Not all such cases were resolved, id., and at least one was resolved
through confidential settlement. See Alan J. Craver, Stylist Who Said Her Abortion Led to
Firing Settles Suit, THE BALT. SUN, Feb. 5, 1995, at 11C. Most of these cases went resolved
or unsettled. Id.
10. See Flesner v. Technical Commc’ns Corp., 575 N.E.2d 1107, 1111 (Mass. 1991) (applying the PPE to termination for cooperating with a federal investigation of the employer).
11. See Novosel v. Nationwide Ins. Co., 721 F.2d 894, 900 (3d Cir. 1983) (recognizing a
PPE where company fired worker for refusing to lobby state legislature for policies favored
by the company).
12. See Robert Kelner, Allegations of Employer Coercion of Political Activity Are on the
Rise, INSIDE POLITICAL LAW (Oct. 17, 2012), http://www.insidepoliticallaw.com/
2012/10/17/allegations-of-employer-coercion-of-political-activity-are-on-the-rise/ (describing
such incidents).
13. Miller v. Safeway, Inc., 102 P.3d 282, 295 (Alaska 2004).
14. Debra J. Saunders, Where There’s Smoke, You’re Fired, S.F. CHRON., Feb. 17,
2005, at B13; see also Amy H. Moorman, Employer Regulation of Off-Duty Smoking: Meeting the Needs of Employers and Employees with Smoking Cessation Programs, J.
INDIVIDUAL EMP’T RTS, 1994-95, at 243, 243 (discussing the merits of employing nonsmokers from an organizational perspective, but cautioning against the “slippery slope” of
involvement with employees’ off-duty activities).
15. Best Lock Corp. v. Review Bd. of the Ind. Dep’t of Emp’t & Training Servs., 572
N.E.2d 520, 521-22 (Ind. App. 1991) (concluding that employer did not have just cause to
fire employee for violating work policy prohibiting the off-duty consumption of alcohol).
948
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
believe that courts should provide employees with common law protections. But the law’s laissez-faire preference for “at-will” employment arrangements, and an intuitive distinction between the application of constitutional principles to public versus private employers,
tug in the opposite direction.
For example, in Novosel v. Nationwide Insurance Company, the
Third Circuit, applying Pennsylvania law in a diversity case, explained at length how constitutional principles could legitimately
form the basis for a judicially found “public policy” under the PPE.16
Several years later, after intervening Pennsylvania cases rejected
constitutionally inspired PPE claims,17 the Third Circuit sharply limited Novosel, holding that the absence of state action undercut the
utility of constitutional doctrine in deciding the contours of the state’s
public policy.18
This divergence mirrors one shown among various state courts.
While many of them either explicitly or implicitly drew upon constitutional sources,19 and no state that recognizes a broad PPE specifically
excludes constitutional sources from its public policy analysis,20 a fair
number of courts have criticized the practice over the years.21 Many,
but not all, of the later cases involved First Amendment issues.
16. 721 F.2d at 899-900.
17. See, e.g., Booth v. McDonnell Douglas Truck Servs., Inc. 585 A.2d 24, 28 (Pa. Super. Ct. 1991) (stating that certain state constitutional provisions protecting contract rights
and obligations did not support employee’s right to collect unpaid commissions as a public
policy claim because there was no state action).
18. Borse v. Piece Goods Shop, Inc., 963 F.2d 611, 620 (3d Cir. 1992).
19. See, e.g., Gantt v. Sentry Ins., 824 P.2d 680, 687-88 (Cal. 1992) (referencing only
“the constitution or statutory provisions”); Parnar v. Americana Hotels, Inc., 652 P.2d 625,
631 (Haw. 1982) (determining that a clear mandate of public policy is violated when “the
employer's conduct contravenes the letter or purpose of a constitutional, statutory, or
regulatory provision or scheme”); Fleshner v. Pepose Vision Inst., P.C., 304 S.W.3d 81, 92
(Mo. 2010) (en banc) (listing “the constitution, statutes, regulations promulgated pursuant
to statute, or rules created by a governmental body”); Dohme v. Eurand America, Inc., 956
N.E.2d 825, 829 (Ohio 2011) (noting that the Constitution is one source of public policy
in applying the PPE); Scott v. Extracorporeal, Inc., 545 A.2d 334, 342 (Pa. Super. Ct. 1988)
(stating that public policy must be articulated “in law or a constitutional provision”);
Gardner v. Loomis Armored Inc., 913 P.2d 377, 380 (Wash. 1996) (examining whether employer’s conduct “contravenes the letter or purpose of a constitutional, statutory, or regulatory provision or scheme” (quoting Parnar, 652 P.2d at 631 (emphasis added))); Birthisel v.
Tri-Cities Health Servs. Corp., 424 S.E.2d 606, 612 (W. Va. 1992) (noting a constitutional
basis for public policy in applying the PPE); Brockmeyer v. Dun & Bradstreet, 335 N.W.2d
834, 840 (Wis. 1983) (listing only “constitutional or statutory provision”). While state
courts vary, most draw upon both federal and state constitutions as sources of public policy. See infra Part III.B.
20. See generally infra Part III.A.
21. See Grzyb v. Evans, 700 S.W.2d 399, 402 (Ky. 1985) (rejecting a “freedom of association” basis for a wrongful discharge claim, because the freedom of association right only
applies to governments (citing United Bhd. of Carpenters & Joiners of Am. v. Scott, 463
U.S. 825 (1985) (dealing with a state action requirement when suing directly under the
First Amendment, not a PPE))); Cisco v. United Parcel Servs., Inc., 476 A.2d 1340, 1344
2014]
COMPANY CONDUCT
949
Some scholars have urged that federal or state constitutional provisions should be among the potential sources of relevant public policy for purposes of allegedly wrongful employment practices.22 However, they have not given the issue extensive analysis and discussion.
In recent years, scholars have examined the issue from a more topicspecific perspective, such as the right of employee privacy,23 or with a
state-specific or even case-specific focus.24
Other scholars have discussed more generally the connection between constitutional law and private law. One approach is to examine when private action may be treated as equivalent to public action,
and thus be subject to constitutional norms.25 Furthermore, international law recognizes a doctrine known as “direct horizontal effect,”
wherein the constitutional human rights guarantees of a country are
applied to govern relations between private parties.26 Just recently,
in 2012, the Supreme Court dramatically changed constitutional
criminal procedure law by importing private, common law property
(Pa. Super. Ct. 1984) (rejecting presumption of innocence-based PPE claim and stating that
constitutional rights were not meant to be “superimposed into” one’s “remaining life experiences”); Bushko v. Miller Brewing Co., 396 N.W.2d 167, 172 (Wis. 1986) (narrowly construing the PPE and rejecting a free speech-based PPE claim, lest it open a “Pandora’s box”
of due process and equal protection arguments which would “eliminate any distinction
between private and governmental employment”). Cisco echoes similar choices made by
other courts within Pennsylvania. See Booth, 585 A.2d at 28 (finding no constitutional basis for a public policy cause of action because no state action was involved), appeal denied,
597 A.2d 1150 (Pa. 1991); Veno v. Meredith, 515 A.2d 571, 581 (Pa. Super. Ct. 1986) (rejecting on similar grounds a PPE claim by newspaper editor fired for publishing an article
criticizing a judge); Martin v. Capital Cities Media, Inc., 511 A.2d 830, 844 (Pa. Super. Ct.
1986) (rejecting free speech-based PPE claim against private newspaper).
22. See, e.g., Richard L. Alfred & Ben T. Clements, The Public Policy Exception to the
At-Will Employment Rule, 78 MASS. L. REV. 88, 93-94 (1993); Henry H. Perritt, Jr., The
Future of Wrongful Dismissal Claims: Where Does Employer Self Interest Lie?, 58 U. CIN. L.
REV. 397, 402-03 (1989) (urging ready use by courts of both federal and state constitutional
sources in applying the PPE).
23. See, e.g., Adam Shinar, Public Employee Speech and the Privatization of the First
Amendment, 46 CONN. L. REV. 1, 22-24 (2013) (referencing the issue briefly as background
with respect to free speech); John B. Wefing, Employer Drug Testing: Disparate Judicial
and Legislative Responses, 63 ALB. L. REV. 799 (2000) (contrasting drug testing rules for
public and private employers); Jill Yung, Big Brother IS Watching: How Employee Monitoring in 2004 Brought Orwell’s 1984 to Life and What the Law Should Do About It, 36 SETON
HALL L. REV. 163 (2005) (discussing electronic monitoring of employees via video surveillance, GPS tracking, and Internet usage tracking software).
24. See, e.g., Ivo Becica, Note, Privacy—State Constitutional Privacy Rights Against
Private Employers: A “Hairy” Issue in Alaska. Miller v. Safeway, Inc., 102 P.3d 282 (Alaska
2004), 37 RUTGERS L.J. 1235 (2006) (discussing Alaska case involving privacy of private
employee); Maureen Binetti, Minding Your Business: Is Your Life After Work Really Private?, N.J. LAW. MAG., Apr. 2008, at 33 (discussing law in New Jersey on employer interference with off-duty, off-premises behavior such as smoking).
25. See Christian Turner, State Action Problems, 65 FLA. L. REV. 281 (2013).
26. See Willmai Rivera-Pérez, What’s the Constitution Got to Do with It? Expanding
the Scope of Constitutional Rights into the Private Sphere, 3 CREIGHTON INT'L & COMP. L.J.
189 (2012) (comparative analysis of the “direct horizontal effect”).
950
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
notions of trespass.27 In evaluating police investigative acts, the
Court held that where the common law would recognize a trespass to
private property, the acts constitute a “search” within the meaning of
the Fourth Amendment.28 Just as fruitful is an evaluation of the reverse dynamic, wherein constitutional principles inform the application of a common law cause of action.
Assuming that constitutional principles indeed have some significant role to play in evaluating such a common law claim, significant
questions remain as to which constitutional principles, and how to
apply them to arguably analogous factual situations in the private
employment sphere. If we acknowledge a strong public policy in favor
of the right to vote, should we reason similarly with respect to the
right to bear arms? If we apply free speech principles to protect an
employee from being forced to sign a petition with which she disagrees, should we apply them equally to allow her to publicly criticize
her employer and bring it into disrepute? Scholarship has not discussed the issues in this manner.
This Article discusses the use of the PPE to vindicate constitutional interests. It takes an interdisciplinary approach, mixing analysis from business law, tort law, and constitutional law. It argues that
courts can and should consider constitutional principles when deciding PPE claims. Constitutional principles have as valid a place in the
PPE analysis as do statutory principles. This Article also argues that
some constitutional rights are more amenable to a private-sector analogy, and/or more appropriate for a PPE analysis. For example, the
right to informational privacy is more translatable to the private employment context than the rights guaranteed by the Establishment
Clause. And the right to free speech enjoys more of a current public
policy consensus than the right to abortion, or to bear arms. For each
of several different constitutional provisions examined, this Article
recommends general principles to guide courts in applying the PPE.
Part II lays out background on the PPE. Part III discusses the extent to which courts have used constitutional principles to inform
their discussion, and argues that such principles are appropriate
sources of policy, at least in certain limited situations. Part IV discusses the extent to which the PPE cause of action is necessary or
advisable, given existing statutory remedies for similar types of employee complaints. Part V suggests how one might apply to the PPE
such specific constitutional provisions or doctrines as: free speech,
the Fourth Amendment, equal protection, reproductive rights, substantive due process, procedural due process, and the Second
Amendment. Part VI offers concluding thoughts.
27. See United States v. Jones, 132 S. Ct. 945 (2012).
28. Id. at 949-50.
2014]
COMPANY CONDUCT
951
II. THE PUBLIC POLICY EXCEPTION TO AT-WILL EMPLOYMENT
A fundamental rule of employment contract law is, in the absence
of a contractual term or statutory provision to the contrary, employment is presumed to be at will. That is, the employer can fire the employee for a good reason, a bad reason, or no reason at all.29
One well-recognized exception to the at-will employment rule is
the PPE.30 This judicially created exception emerged in recent decades, representing a departure from traditional common law at-will
employment principles.31
In some ways, the rise of this exception constitutes a movement
back toward the law of the eighteenth century, when masters had
implied obligations to “treat the servant humanely” and to “furnish
him with suitable lodging.”32 Back then, for example, instead of a
presumption of at-will employment, the common law presumed an
implied one-year term of employment if the term was not specified.33
This made sense in a largely agrarian society. It protected the servant from being used during the growing and harvest season only to
be let go when the ground was barren; it also protected the master
from having to support a servant during the winter only to be abandoned for a better offer once serious work had to be done.34
29. 10 LEX K. LARSON, EMP. DISCRIMINATION § 174.01 (2d ed. 2012); see, e.g., Phipps
v. IASD Health Servs. Corp., 558 N.W.2d 198, 202 (Iowa 1997) (noting that, unless a valid
employment contract exists between parties, “an employer may discharge an employee at
any time, for any reason, or no reason at all”); Fleshner v. Pepose Vision Inst., P.C., 304
S.W.3d 81, 91 (Mo. 2010) (en banc) (asserting that “[g]enerally, at-will employees may be
terminated for any reason or for no reason”);.
30. LARSON, supra note 29, § 174.01.
31. Donald C. Carroll, At-Will Employment: The Arc of Justice Bends Towards the
Doctrine’s Rejection, 46 U.S.F. L. REV. 655, 668-69 (2012) (supporting the PPE); R. Scott
Oswald & Michael L. Vogelsang Jr., The ABCs of Common Law Wrongful Termination
Claims in the Washington Metropolitan Region, 3 LAB. & EMP. L.F. 197, 199-202 (2013)
(tracing the development of the PPE from initial emergence to broad acceptance among
jurisdictions); Christopher L. Pennington, Comment, The Public Policy Exception to the
Employment-at-Will Doctrine: Its Inconsistencies in Application, 68 TUL. L. REV. 1583,
1594-98 (1994) (same, and criticizing its application); Kenneth R. Swift, The Public Policy
Exception to Employment At-Will: Time to Retire a Noble Warrior?, 61 MERCER L. REV. 551,
553-63 (2010) (opposing the PPE).
32. H.G. WOOD, A TREATISE ON THE LAW OF MASTER AND SERVANT § 83, at 165-66
(William S. Hein & Co. 1981) (1877) (emphasis omitted); see generally HENRY H. PERRITT,
JR., EMPLOYEE DISMISSAL LAW AND PRACTICE § 1.03 (5th ed. 2008) (discussing how the
Industrial Revolution influenced an economy-driven approach to employment that undermined traditional employer obligations regarding public policy).
33. 1 WILLIAM BLACKSTONE, COMMENTARIES ON THE LAWS OF ENGLAND 413 (Oxford,
Clarendon Press 1765) (“The contract between [servants] and their masters arises upon the
hiring. If the hiring be general without any particular time limited, the law construes it be
a hiring for a year.”); see also Kenneth A. Sprang, Beware the Toothless Tiger: A Critique of
the Model Employment Termination Act, 43 AM. U. L. REV. 849, 860 (1994) (discussing the
American courts’ adoption of this one-year presumption during the nineteenth century).
34. Pennington, supra note 31, at 1584-85 n.7.
952
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
With the advent of industrialization came a different approach,
one which emphasized the freedom of contract so as to enhance commerce and reduce burdens on business.35 This approach enshrined
a default at-will rule by the end of the nineteenth century.36 It
preceded a near-parallel development in constitutional law, the socalled “Lochner era,” during which the Supreme Court imposed a
laissez-faire approach to employment regulation based on a supposed
constitutional freedom of contract.37 This development is another example of the interplay between constitutional law and the common
law of employment.
Later in the twentieth century, two decades after the end of the
Lochner era, courts began to recognize policy-based exceptions to the
at-will rule.38 One prominent expert in the field has called the erosion
of the at-will doctrine “[t]he most significant employment law development in the last quarter of the 20th century.”39 Under the PPE, an
employee can bring suit for wrongful discharge where the termination was in violation of a clear public policy.40 For example, courts
may intervene when an employer fires workers because they refused
to violate the law,41 complied with a legal requirement,42 or engaged
in some form of protected activity.43 Jurisdictions vary as to which
35. Id. at 1585; see also infra notes 137-139 (discussing the perceived burden on employers as the motivating factor behind limiting statutory causes of action).
36. Pennington, supra note 31, at 1585.
37. Lochner v. New York, 198 U.S. 45, 64 (1905) (finding a statutory regulation of
labor hours within private businesses to be an unconstitutional infringement on the “freedom of contract” due process right as well as an invalid use of police power). The Lochner
era ended thirty years later when the Court rejected the notion that the freedom to contract precluded government regulation of the labor market. EDWIN CHEMERINSKY,
CONSTITUTIONAL LAW PRINCIPLES AND POLICIES § 8.2 (4th ed. 2011) (analyzing this laissezfaire philosophy as stemming from a “social Darwinism” in which the individual’s freedom
to contract limited the government’s regulatory power). See NLRB v. Jones & Laughlin
Steel Corp., 301 U.S. 1, 31-32 (1937).
38. The first case to recognize a PPE to the at-will rule was Petermann v. Int’l Bhd. of
Teamsters, 344 P.2d 25 (Cal. Dist. Ct. App. 1959). Pennington, supra note 31, at 1593.
39. Perritt, supra note 22, at 397.
40. LARSON, supra note 29, § 174.04[1]; see also Mello v. Stop & Shop Cos., 524 N.E.2d
105 (Mass. 1988) (the PPE may apply to termination of employee for reporting employer’s
criminal conduct); Collins v. Rizkana, 652 N.E.2d 653 (Ohio 1995) (establishing sexual
harassment as a valid basis for the PPE); Thompson v. St. Regis Paper Co., 685 P.2d 1081
(Wash. 1984) (holding that it violated clear public policy to discharge plaintiff for complying with federal statute on accounting procedures).
41. See, e.g., Woodson v. AMF Leisureland Ctrs., Inc., 842 F.2d 699 (3d Cir. 1988)
(upholding a PPE action when barmaid was fired for refusing to serve alcohol to visibly
intoxicated persons).
42. See, e.g., Nees v. Hocks, 536 P.2d 512 (Or. 1975) (creating a PPE where employee
was fired for refusing employer's request to ask for excuse from jury duty).
43. See, e.g., Porter v. Reardon Mach. Co., 962 S.W.2d 932, 937-38 (Mo. Ct. App. 1998)
(recognizing exception for “whistleblower” activity).
2014]
COMPANY CONDUCT
953
public policies qualify for the PPE,44 but virtually every jurisdiction
recognizes some form of this exception.45
Some courts explicitly limit the cause of action to cases where the
employee has actually been discharged.46 But there is no theoretical
reason why this doctrine could not apply to other adverse employment actions besides termination. Some courts have so held.47 However, the overwhelming majority of the cases involve full termination,
leading courts to speak of this as a “wrongful discharge” cause of action. It also sometimes is referred to as the tort of “retaliatory discharge,” at least where the complaint alleges termination in retaliation for the exercise of a protected right, or for the failure to violate a
clear law or policy upon the direction of a superior.48
The cause of action is normally characterized as a tort.49 Some authorities suggest that the cause of action also may be partially
44. Compare Sventko v. Kroger Co., 245 N.W.2d 151 (Mich. Ct. App. 1976) (creating a
PPE cause of action where en employee filed workers’ compensation claim), and McGarrity
v. Berlin Metals, Inc., 774 N.E.2d 71 (Ind. Ct. App. 2002) (finding cause of action for retaliatory discharge where employee refused to commit illegal act for which he would be personally liable), and Call v. Scott Brass, Inc., 553 N.E.2d 1225 (Ind. Ct. App. 1990) (applying
PPE where employee was fired for complying with jury duty summons), with Campbell v.
Eli Lilly & Co., 413 N.E.2d 1054 (Ind. Ct. App. 1980) (denying cause of action where employee was terminated for reporting supervisor misconduct to company officials), and Price
v. Carmack Datsun, Inc., 485 N.E.2d 359 (Ill. 1985) (finding no cognizable PPE for an employee who filed insurance claim because it did not affect the public interest), and MacDonald v. E. Fine Paper, Inc., 485 A.2d 228 (Me. 1984) (declining to find a PPE for wrongful discharge of an employee who was fired for filing a workers’ compensation claim because remedy already existed under Workers’ Compensation Act).
45. GERARD P. PANARO, EMPLOYMENT LAW MANUAL: RECRUITMENT, SELECTION,
TERMINATION ¶ 7.02 (1990); see also PERRITT, supra note 32, § 7.09[B][1] (acknowledging
that most courts recognize a PPE to at-will employment where an employee is fired for
serving jury duty).
46. See, e.g., Darchak v. City of Chi. Bd. of Educ., 580 F.3d 622, (7th Cir. 2009) (declining to extend Illinois PPE to any injury short of actual discharge); Dixon v. Denny’s
Inc., 957 F. Supp. 792, 799 (E.D. Va. 1996) (concluding “that the public policy exception
to Virginia's employment at will doctrine does not yet encompass constructive discharges
as well as actual discharges”); Roberts v. Dudley, 993 P.2d 901 (Wash. 2000) (limiting the
PPE to termination).
47. See, e.g., Hunter v. Port Auth. of Allegheny Cnty., 419 A.2d 631 (Pa. Super. Ct.
1980) (recognizing doctrine’s applicability to a refusal to hire); MISSOURI APPROVED JURY
INSTRUCTIONS (CIVIL) § 38.03 cmt. A (Robert T. Adams et al., eds., 7th ed. 2012) (contemplating such a cause of action in cases of “constructive discharge, demotion, or adverse job
consequences,” but declining to take a position on the validity of such a claim).
48. See, e.g., Turner v. Mem’l Med. Ctr., 911 N.E.2d 369, 375-76 (Ill. 2009) (using the
terms “wrongful discharge” and “retaliatory discharge” interchangeably in a case involving
a medical employee allegedly fired in retaliation for speaking out about patient safety);
Birthisel v. Tri-Cities Health Servs. Corp., 424 S.E.2d 606, 611 (W. Va. 1992) (using both
terms to describe a claim by a medical employee fired for choosing professional ethics
standards over a conflicting company policy).
49. See, e.g., Vigil v. Arzola, 699 P.2d 613, 619 (N.M. Ct. App. 1983) (acknowledging
that the majority of courts find the public policy cause of action to sound in tort, which
“provides a more appropriate rationale than one [sounding] in contract”).
954
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
grounded in the common law’s implied covenant of good faith and fair
dealing, which sounds in contract law.50
The basic rationale for the exception is that, while an employer
should generally be free from interference in choosing whom to employ, the public has an interest in ensuring that the inherently coercive power of an employer to discharge is not used to pressure employees to disregard the law, forfeit important rights, or otherwise
contravene substantial public policy interests.51 Even where such
public interests are implicated, though, the employer may still prevail where there is some “separate, plausible and legitimate reason”
for taking the challenged action.52
A typical listing of the elements of the cause of action includes:
(1) The existence of a clear public policy: i.e., the “clarity” element;
(2) The termination placed that policy in jeopardy: i.e., the “jeopardy” element;
(3) The plaintiff’s termination was actually motivated by conduct
related to the public policy: i.e., the “causation” element; and
(4) The employer lacked a legitimate business justification for
terminating the employee: i.e., the “overriding justification” element.53
While there is a theoretical difference between the second and
third elements, there is little distinguishing them in practice. Typically, the public policy at issue is placed in jeopardy precisely because
of the causal link between the termination and the policy. Thus, the
causation proxy leads to the jeopardy prong.
50. See, e.g., Norcon, Inc. v. Kotowski, 971 P.2d 158, 167 (Alaska 1999) (public policy
tort largely encompassed by implied covenant); Sterling Drug, Inc. v. Oxford, 743 S.W.2d
380, 385 (Ark. 1988) (concluding that a contract cause of action is most appropriate for a
wrongful discharge claim); Brockmeyer v. Dun & Bradstreet, 335 N.W.2d 834, 841 (Wis.
1983) (finding the remedies provided by wrongful discharge statutes to be more applicable
to contract causes of action than tort causes action).
51. MARK W. BENNETT ET AL., EMPLOYMENT RELATIONSHIPS: LAW AND
PRACTICE § 2.02[F][b] (2014) (describing the PPE as having particular significance in situations where an employee refuses to commit illegal acts, exercises a legal right, performs a
public duty, or reports an employer’s misconduct); see also Strozinsky v. Sch. Dist. of
Brown Deer, 614 N.W.2d 443, 453 (Wis. 2000) (asserting that the exception “properly balances the need to protect employees from terminations that contradict public policy with
the employer’s historical discretion to discharge employees under the freedom to contract
embodied in the at-will doctrine”).
52. Borse v. Piece Goods Shop, Inc., 963 F.2d 611, 616 (3d Cir. 1992) (citing Cisco v.
United Parcel Servs., Inc., 476 A.2d 1340, 1343 (Pa. Super. Ct. 1984)).
53. See Fitzgerald v. Salsbury Chem., Inc., 613 N.W.2d 275, 282 n.2 (Iowa 2000) (referencing this four-part elemental foundation as being a sufficient basis for a PPE cause of
action); Collins v. Rizkana, 652 N.E.2d 653, 657-58 (Ohio 1995) (same); PERRITT, supra
note 31, § 7.04 (recognizing these four elements as the basic analytical framework for a
public policy tort cause of action).
2014]
COMPANY CONDUCT
955
Courts caution that the PPE should be narrow and not expanded
lightly.54 The default position is still at-will employment. Other than
contractual terms between the parties, courts say, deviations from
this default arrangement should be specified by the legislature.55
Further, the public policy alleged to have been violated by the adverse employment action must be “substantial.”56 As one court stated,
the public policy must “strike at the heart of a citizen’s social rights,
duties, and responsibilities.”57 The existence of the policy, as well as
its contours and scope, must be relatively clear so as to place employers on proper notice of a potential violation.58 The analogy here would
be to qualified immunity for public officials, where the court must
find a violation of “clearly established law.”59
A listing of categories of typical situations in which PPE cases
arise includes those where employers fired employees:
(1) For refusing to commit an illegal act;
(2) For performing a legal duty;
(3) For exercising a legal right or privilege; or
(4) In retaliation for reporting misconduct.60
An example of category (1) would be termination for refusing to
commit perjury before an official panel overseeing or investigating
the company.61 An example of category (2) would be an instance in
which an employee must report for jury duty, or has an affirmative
54. See, e.g., Gardner v. Loomis Armored Inc., 913 P.2d 377, 380 (Wash. 1996) (stating
that courts should exercise caution when declaring public policy absent express legislative
or judicial expression on the subject (citing Thompson v. St. Regis Paper Co., 685 P.2d 1081
(Wash. 1984))); PERRITT, supra note 32, § 7.05[B] (describing courts’ general reluctance to
extend the exception beyond the express language of a statute).
55. See Horn v. N.Y. Times, 790 N.E.2d 753, 756 (N.Y. 2003) (“[S]ignificant alteration
of employment relationships . . . is best left to the Legislature . . . because stability and
predictability in contractual affairs is a highly desirable jurisprudential value.” (quoting
Sabetay v. Sterling Drug, Inc., 506 N.E.2d 919, 923 (N.Y. 1987))); see also Borden v. Johnson, 395 S.E.2d 628, 629 (Ga. Ct. App. 1990) (cautioning that the courts should not “usurp
the legislative function” by making their own decisions regarding the public policies which
would trump the general rule of at-will employment).
56. Birthisel v. Tri-Cities Health Servs. Corp., 424 S.E.2d 606, 612 (W. Va. 1992).
57. Dicomes v. State, 782 P.2d 1002, 1006 (Wash. 1989) (en banc) (quoting Palmateer
v. Int’l Harvester Co., 421 N.E.2d 876 (Ill. 1981)); see also Booth v. McDonnell Douglas
Truck Servs., Inc., 585 A.2d 24, 28 (Pa. Super. Ct. 1991) (using similar language).
58. Birthisel, 424 S.E.2d at 612.
59. Id. at 612 n.8.
60. See, e.g., Field v. Phila. Elec. Co., 565 A.2d 1170, 1176 (Pa. Super. Ct 1989) (disclosing safety problems at nuclear power plant); see also JOHN C. MCCARTHY, RECOVERY OF
DAMAGES FOR WRONGFUL DISCHARGE 2D §§ 1.5, 1.9, 1.22, 1.24 (1990) (providing an indepth analysis for each of these four distinct situations).
61. See Woodson v. AMF Leisureland Ctrs., Inc., 842 F.2d 699, 702 (3d Cir. 1988)
(upholding PPE action where barmaid was fired for refusing to serve alcohol to visibly
intoxicated persons); Petermann v. Int’l Bhd. of Teamsters, 344 P.2d 25, 27 (Cal. Dist.
Ct. App. 1959).
956
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
duty to report misconduct at the place of work.62 An example of (4)
would be any legitimate whistleblowing case.63
It is the third category that is relevant to this Article. An uncontroversial and common example for this category would be cases involving terminations for filing a workers’ compensation claim.64 But
what remedy, if any, does an employee have who is fired for exercising a constitutional right—in other words, for engaging in a constitutionally protected activity?
III. CONSTITUTIONAL SOURCE OF PUBLIC POLICY VERSUS OTHER
LEGAL SOURCES
A. Generally
Courts discussing the PPE uniformly recognize statutes and judicial decisions as valid sources of the public policy at issue in these
cases.65 Regarding the former, to get a sense of the public policy of
the state, a court can look not only to statutory text, but also to
legislative history.66
But the use of constitutional law as a source has not always been
viewed favorably. Some commentators have suggested generally that
it is more difficult to use as a source of public policy,67 or that certain
states exclude such use.68 In most of these states, rather than
being specifically hostile to constitutional sources per se, courts are
actually just hostile to the PPE in general. They either reject it out-
62. See Call v. Scott Brass, Inc., 553 N.E.2d 1225 (Ind. Ct. App. 1990) (granting cause
of action for employee’s executing statutory duty to comply with a summons to appear
for jury duty); Reuther v. Fowler & Williams, Inc., 386 A.2d 119, 119-20 (Pa. Super. Ct.
1978) (employee fired for missing work due to jury service).
63. See infra Part V.B.1.b.
64. See, e.g., Kelsay v. Motorola, Inc., 384 N.E.2d 353 (Ill. 1979); Frampton v. Cent.
Ind. Gas Co., 297 N.E.2d 425 (Ind. 1973); Hrab v. Hayes-Albion Corp, 302 N.W.2d 606
(Mich. Ct. App. 1981).
65. PANARO, supra note 45, ¶ 7.02.
66. See, e.g., Hansen v. Am. Online, Inc., 96 P.3d 950, 954 n.7 (Utah 2004) (bypassing
a statutory text analysis by looking to legislative history and debate as a legitimate means
of defining the public policy of weapons in the workplace).
67. See PANARO, supra note 45, ¶ 7.03[2].
68. See BENNETT ET AL., supra note 51, at app. 2B, 2-71 to 2-89 (2014) (listing Arizona,
Colorado, Delaware, Florida, Georgia, Louisiana, New York, Pennsylvania, and Wisconsin
as states that do not look to constitutional sources for public policy claims).
2014]
COMPANY CONDUCT
957
right,69 or else view it so narrowly as to effectively preclude use of
constitutional sources.70
Indeed, the use of constitutional principles as a source now is generally accepted, though it has received some criticism by state courts.
Some state courts explicitly refer to constitutional provisions as a potential source of the public policy. A typical formulation is to ask
whether the employer’s action “contravenes the letter or purpose of a
constitutional, statutory, or regulatory provision or scheme.”71 This
reference to at least constitutional and statutory sources of public
policy is common.72 Other courts have implicitly recognized constitutional sources while rejecting constitution-sourced PPE claims on
their facts, considering in detail claims involving employer actions
interfering with free speech,73 religious freedom,74 and the right to
69. See, e.g., Bruley v. Village Green Mgmt. Co., 592 F. Supp. 2d 1381, 1387-88 (M.D.
Fla. 2008), aff'd per curiam, 333 Fed. App’x 491 (11th Cir. 2009); Wright v. Dothan Chrysler Plymouth Dodge, Inc., 658 So. 2d 428, 431 (Ala. 1995); Pacheo v. Raytheon Co., 623
A.2d 464, 465 (R.I. 1993) (per curiam).
70. See, e.g., Poole v. In Home Health, LLC, 742 S.E.2d 492, 494 (Ga. Ct. App. 2013)
(judicially created exceptions to at-will employment disfavored, so Georgia courts defer to
legislature to create exceptions); McArn v. Allied Bruce-Terminix Co., 626 So. 2d 603, 607
(Miss. 1993) (recognizing only two circumstances in which the PPE applies: (1) where an
employee refuses to participate in an illegal act; (2) where an employee is discharged for
reporting illegal acts of his employer); Wieder v. Skala, 609 N.E.2d 105, 110 (N.Y. 1992)
(creating a narrow exception to employment at-will employment but reiterating that further alterations are “best left to the legislature”); Ed Rachal Found. v. D’Unger 207 S.W.3d
330, 333 (Tex. 2006) (PPE limited to employees terminated for whistleblowing only where
reporting the employer’s illegal conduct is a legal obligation rather than a mere civic duty);
Paul H. Tobias, State-by-State Compendium of Leading and Representative Decisions Concerning the Public Policy Tort Doctrine, 1 LIT. WRONG. DISCHARGE CLAIMS app. 5A (last
updated Dec. 2013), available at Westlaw, 1 LITWDCS App. 5A (collecting cases establishing that Indiana courts have so far limited the PPE to cases involving workers’ compensation or terminations for refusing to commit an illegal act).
71. Gardner v. Loomis Armored Inc., 913 P.2d 377, 380 (Wash. 1996) (quoting Thompson v. St. Regis Paper Co., 685 P.2d 1081, 1089 (Wash. 1984)).
72. See Gantt v. Sentry Ins., 824 P.2d 680, 687-88 (Cal. 1992) (en banc) (referencing
“constitutional or statutory provisions”); Fleshner v. Pepose Vision Inst., P.C., 304 S.W.3d
81, 92 (Mo. 2010) (en banc) (listing “the constitution, statutes, regulations promulgated
pursuant to statute, or rules created by a governmental body”); Scott v. Extracorporeal,
Inc., 545 A.2d 334, 342 (Pa. Super. Ct. 1988) (public policy must be articulated “in law
or constitutional provision”); Payne v. Rozendaal, 520 A.2d 586, 587 (Vt. 1986) (listing constitution, statute, judicial decision, and “customs and conventions of the people”); Brockmeyer v. Dun & Bradstreet, 335 N.W.2d 834, 840 (Wis. 1983) (listing only “constitutional
or statutory provision”).
73. See, e.g., Korb v. Raytheon Corp., 574 N.E.2d 370 (Mass. 1991) (holding that a
company spokesperson whose official remarks conflicted with employer interests did not
have a cognizable wrongful discharge claim on the basis of free speech).
74. See, e.g., Kolodziej v. Smith, 588 N.E.2d 634, 638 (Mass. 1992) (acknowledging the
existence of the PPE where an employer inhibits the free exercise of religion in the workplace or dismisses an employee on the basis of religion, but finding that requiring attendance at an employee training seminar referencing Christian ideology does not restrict religious exercise under this public policy).
958
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
privacy.75 Had the courts rejected constitutional sources of public policy, such extensive discussions would not have been necessary.
Given the uniform acceptance of statutory provisions as potential
sources of the public policy, recognition of constitutional provisions
has at least some intuitive appeal. After all, since statutes are “subordinate” to constitutional provisions, “it would make little sense” to
draw upon the former but not the latter for guidance.76
Of course, at first glance, the argument for not drawing upon constitutional sources in PPE cases is equally justifiable. Generally, constitutional provisions are designed to restrict government behavior.77
Arguably, constitutional principles are ill-suited to govern employment relations between private parties.78 It would make little sense,
and would, perhaps, unduly restrict commerce, to make private employers abide by all the same requirements of procedural due process
as a government employer.79
Based on this latter objection, courts in a few cases have flatly declined to draw upon constitutional sources in examining PPE claims.
In a series of cases, the Pennsylvania Superior Court rejected PPE
claims based on constitutional sources of policy, at least where applied to private employers.80 One opinion emphasized that constitu75. See, e.g., Cort v. Bristol-Myers Co., 431 N.E.2d 908, 914 (Mass. 1982) (denying
specific claim of wrongful discharge for failing to fill out employee questionnaire with arguably unwarranted and intrusive questions because employee also refused to answer
clearly relevant and warranted questions, but nonetheless recognizing that employee privacy rights limited the scope of questions for which refusal to answer constituted valid
grounds for discharge).
76. Alfred & Clements, supra note 22, at 93-94.
77. See, e.g., CHEMERINSKY, supra note 37, § 6.4 (examining the “state action doctrine”
and indicating that the Constitution applies to the government and does not regulate private conduct “no matter how discriminatory or how much [private wrongs] infringe fundamental rights”); see also Barr v. Kelso-Burnett Co., 478 N.E.2d 1354, 1356-57 (Ill. 1985)
(stating that provisions of the Illinois Constitution mandate public policy protecting free
speech restrict the actions of governmental or public officials, not of private employers);
Hatfield v. Rochelle Coal Co., 813 P.2d 1308, 1311 (Wyo. 1991) (holding that “private employers are not subject to due process claims under their state constitutions for wrongful
termination of employees”).
78. See Novosel v. Nationwide Ins. Co., 721 F.2d 894, 903 (3d Cir. 1983) (Becker, J.,
dissenting from denial of rehearing en banc); Perritt, supra note 22, at 400-03 (articulating,
but rejecting, this argument).
79. Cf. Turner, supra note 25, at 282 (“Our dinner invitations, marriage proposals,
and even business ventures certainly must be immune to at least some of the constitutional
law that requires public actors to afford equal treatment, due process, and respect for
all viewpoints.”).
80. Veno v. Meredith, 515 A.2d 571, 578-81 (Pa. Super. Ct. 1986) (rejecting on similar
grounds a PPE claim by newspaper editor fired for publishing an article criticizing a
judge); Martin v. Capital Cities Media, Inc., 511 A.2d 830, 843-45 (Pa. Super. Ct. 1986)
(rejecting argument that free speech rights could form basis of a PPE claim against a private employer); Cisco v. United Parcel Servs., Inc., 476 A.2d 1340, 1344 (Pa. Super. Ct.
1984) (rejecting presumption of innocence-based PPE claim and stating that constitutional
rights were not meant to be “superimposed onto” one’s “remaining life experiences”).
2014]
COMPANY CONDUCT
959
tional provisions applied directly only to government conduct and not
to disputes between private parties.81 Thus, the PPE claim did not
apply because there had been no allegation of state action.82
Citing this line of cases, the Third Circuit (applying Pennsylvania
law in a diversity case) declined to accept a PPE claim grounded, in
part, on Fourth Amendment-based notions of privacy against a private
employer, holding that the absence of state action was fatal to the use
of constitutional principles in a wrongful discharge suit.83 Summarizing its view of the Pennsylvania case law, the Third Circuit criticized
the use of constitution-inspired PPE causes of action, suggesting a certain level of skepticism to any further use of constitutional principles
in future applications of the PPE.84 For similar reasons, the Third Circuit reasoned, state constitutional principles were also ineligible as
evidence of public policy in evaluating a PPE claim.85
Pennsylvania is not alone. At one point, the Supreme Court of
West Virginia had stated, with respect to free speech-based PPE
claims, that “[t]he prevailing view among the majority of courts addressing the issue” was that absent state action, “state or federal
constitutional free speech cannot . . . be the basis of a PPE in wrongful discharge claims.”86 This does not appear to be the majority view
regarding free speech-based PPE claims today,87 and even West Virginia still allows theoretical consideration of constitutions in divining
public policy under the PPE.88 But, a number of other states have explicitly rejected particular constitutional sources for the PPE in many
instances, using language suggesting overall hostility to the transfer
of constitutional policy norms to the private employment setting.89
81. Booth v. McDonnell Douglas Truck Servs., Inc., 585 A.2d 24, 28 (Pa. Super. Ct. 1991).
82. Id.
83. Borse v. Piece Goods Shop, Inc., 963 F.2d 611, 619 (3d Cir. 1992).
84. Id. at 618-20.
85. Id. at 620.
86. Tiernan v. Charleston Area Med. Ctr., Inc., 506 S.E.2d 578, 589 (W. Va. 1998).
87. See infra Part V.B.1.
88. See Wounaris v. W. Va. State Coll., 588 S.E.2d 406, 414 (W. Va. 2003) (noting the
constitution as one of many sources for public policy for purposes of the PPE (citing
Birthisel v. Tri-Cities Health Servs. Corp., 424 S.E.2d 606, 612 (W. Va. 1992))).
89. See Bruley v. Village Green Mgmt. Co., 592 F. Supp. 2d 1381, 1385 (M.D. Fla.
2008) (“Florida has no exception even where termination is founded on an employee’s exercise of constitutional rights.”); Deiters v. Home Depot U.S.A., Inc., 842 F. Supp. 1023, 102729 (M.D. Tenn. 1993) (declining to make the “open courts” provision of the Tennessee constitution a basis for wrongful discharge claims because it did not apply to private employers, and further stating that even if the provision did apply to private employers, the constitutional provision “does not clearly and unambiguously create a public policy which
would prevent the discharge of at-will employees who sue their employers”); Hart v. Seven
Resorts Inc., 947 P.2d 846, 850-51 (Ariz. Ct. App. 1997) (rejecting the right of privacy as an
appropriate basis for the PPE because such a right “applies only to intrusions
by the government or where there is state action”); Edmondson v. Shearer Lumber Prods.,
75 P.3d 733, (Idaho 2003) (determining the constitutional right of free speech to be a source
960
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
In some instances, the reasoning supporting such a rejection
seems superficial. Some courts simply note authority requiring state
action for liability under the Constitution, and then reject PPE
claims where there is no state action. For example, the Kentucky Supreme Court once rejected a “freedom of association” basis for a
wrongful discharge claim, citing United Brotherhood of Carpenters
and Joiners of America v. Scott90 for the proposition that the right to
free association applies only as against government action.91
But that seems to beg the question. No one asserts that direct constitutional liability exists against private employers.92 The basis for
liability is instead a common law cause of action which indisputably
does regulate private action. The question is whether a constitutional
right suggests a strongly held public policy which courts ought to uphold as against private employers under this common law claim.
That question must be answered on some other basis besides the conclusion that a distinct constitutional claim would fail.
In like manner, courts adjudicating PPE claims often draw upon
statutory sources for public policy, even under circumstances in
which the employer has no direct liability under the statute.93 This
may be because the statutes in question are criminal and provide for
no civil liability; or because they apply only to larger employers; or
because the employee’s claim is precluded by statute of limitations
concerns; or for other reasons.94 Nonetheless, courts rely on the statutes for the underlying public policy and then hold that the common
law claim can be used to vindicate that public policy. The analysis is
no different with respect to constitutional sources of public policy.95
of public policy when applied to state regulation only, and not in the context of private
employment); McGarvey v. Key Prop. Mgmt., LLC, 211 P.3d 503 (Wyo. 2009) (rejecting free
speech-based PPE claim while noting that no Wyoming case expressly endorsed application
of Wyoming constitution to private entities).
90. 463 U.S. 825, 831-33 (1983).
91. Grzyb v. Evans, 700 S.W.2d 399, 402 (Ky. 1985).
92. But there are unusual instances in which a state constitution has been interpreted
to restrict private action. See Kraslawsky v. Upper Deck Co., 65 Cal. Rptr. 2d 297, 301-03
(Cal. Ct. App. 1997) (state constitutional guarantee of the right to pursue and obtain privacy creates a right of action against private as well as government entities). Indeed, even
the U.S. Constitution’s abolition of involuntary servitude restricts private actors. See U.S.
CONST. amend. XIII.
93. See infra Part IV.
94. Id.
95. Even where a court has rejected constitutional provisions as a legitimate source of
public policy in evaluating wrongful discharge claims, constitutional precedent may still
inform the analysis. For example, after rejecting federal and state constitutional provisions
as evidence of public policy, the Third Circuit analyzed the employee privacy claim under
state tort law—but it used Supreme Court Fourth Amendment case law to evaluate the
contours of the privacy interests at stake. Borse v. Piece Goods Shop, Inc., 963 F.2d 611,
621 n.10 (3d Cir. 1992).
2014]
COMPANY CONDUCT
961
A better argument against the use of constitutional sources is the
“slippery slope” concern. As the Wisconsin Supreme Court put it, giving courts the authority to scour constitutions for employerrestricting public policies would “open a Pandora’s box” of due process and equal protection arguments which would “eliminate any distinction between private and governmental employment.”96
However, this concern seems overblown. It is uncontroversial for
courts to look to the entirety of common law doctrine and all statutes
within the state as sources of public policy. This is a very deep well of
potential public policies to draw upon in deciding whether to add to
the list of protections afforded employees. In comparison the incremental increase in potential public policies, and thus potential extra
employee protections, from allowing constitutional analysis would
not seem to be that great. We would not be adding that much extra
employee protection simply by allowing constitutional sources along
with the already plentiful statutory and regulatory sources. This is
particularly so where courts are circumspect about which constitutional rights they will draw upon, and how broadly or narrowly they
apply them, as this Article advocates.97
On balance, while the distinction between public and private settings may justify some caution in drawing upon constitutional principles for PPE purposes, it does not justify wholesale exclusion. Constitutional principles are foundational. They often represent the most
deeply held beliefs of a society, and enshrine our most cherished
rights. Many of them are at least as important as the right to file a
workers’ compensation claim, which almost all jurisdictions recognize
as sufficiently important to form the basis of a PPE claim.98 Why
would we not look to them when deciding what public policies are
clear and substantial? The challenge is simply to decide when constitutional policies governing public entities are sufficiently apt analogies such that they can be logically related to a private setting. As
one commentator put it in a different context, the question is “whether the values protected by declared rights are threatened by concentrated private power in a way fairly analogous to the threat presently
or formerly posed by unlimited government power.”99
96. See Bushko v. Miller Brewing Co., 396 N.W.2d 167, 172 (Wis. 1986) (rejecting a
free speech-based PPE claim).
97. See infra Part V (discussing these questions and making recommendations).
98. Novosel v. Nationwide Ins. Co., 721 F.2d 894, 899 (3d Cir. 1983) (making a
similar observation).
99. See 1 JENNIFER FRIESEN, STATE CONSTITUTIONAL LAW: LITIGATING INDIVIDUAL
RIGHTS, CLAIMS, AND DEFENSES § 9-2(b)(2) (3d ed. 2000).
962
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
B. State versus Federal Constitutions
In determining the public policy underlying a PPE claim, some
state courts make no distinction between federal and state constitutional sources.100 Others expressly reference both federal and state
constitutions. For example, in outlining the elements of the PPE, the
Supreme Court of Ohio has listed the existence of a “clear public
policy . . . manifested in a state or federal constitution, statute or administrative regulation, or in the common law.”101 One of the broadest statements of sources of public policy lists, “among others:”
our federal and state constitutions, our public statutes, our judicial decisions, the applicable principles of the common law,
the acknowledged prevailing concepts of the federal and state
governments relating to and affecting the safety, health, morals
and general welfare of the people for whom government . . . is
factually established. 102
On the other hand, a few courts have looked to state constitutions
but not to the federal constitution. For example, the Supreme Court
of Illinois referred specifically (and only) to the state constitution
when listing potential sources for the public policy.103
Others take a more nuanced approach. For example, in Oklahoma,
federal statutes, standing alone, cannot provide the public policy that
triggers the exception, because the exception is geared toward state
policy.104 But Oklahoma would not preclude the use of the PPE to
protect an employee who reports the defrauding of a federal government program in violation of a federal statute, where such fraud also
would be subject to criminal or civil penalties under Oklahoma law.105
Further, a federal constitutional provision can provide such policy
guidance if it “prescribes a norm of conduct for the state.”106
This Oklahoma doctrine actually points the way toward the appropriate approach on this question. The better view is not to require
that the policy be formulated by an agent of that state’s government,
but rather to require simply that the public policy does indeed apply
within that state, regardless of its original source. The PPE is a
100. See, e.g., Borse, 963 F.2d at 618-20 (rejecting both federal and state constitutional
provisions as evidence of public policy for the PPE).
101. Dohme v. Eurand Am., Inc., 956 N.E.2d 825, 829 (Ohio 2011) (emphasis added).
102. Birthisel v. Tri-Cities Health Servs. Corp., 424 S.E.2d 606, 611 (W. Va. 1992)
(quoting Allen v. Com. Cas. Ins. Co., 37 A.2d 37, 39 (N.J. 1944) (emphasis added)).
103. Turner v. Mem’l Med. Ctr., 911 N.E.2d 369, 374 (Ill. 2009) (quoting Palmateer v.
Int’l Harvester Co., 421 N.E.2d 876, 878 (Ill. 1981)).
104. Darrow v. Integris Health, Inc., 176 P.3d 1204, 1212 (Okla. 2008).
105. Id.
106. Id.
2014]
COMPANY CONDUCT
963
common law doctrine applied by state courts; in evaluating a PPE
claim, it is only sensible to look to laws and policies which apply
within that state.
This approach suggests a distinction among different constitutional provisions as inspiration for a PPE claim. It would allow the
use of only those individual rights from the federal constitution that
have been “incorporated,” i.e., applied to the States through the Due
Process Clause of the Fourteenth Amendment.107
C. Treating Private Action as State Action
A separate ground for importing constitutional principles into
PPE claims might be cases where the private action could be viewed
legitimately as tantamount to state action. In constitutional law,
there are well-recognized exceptions to the state action doctrine that
allow courts to treat action by private entities as state action. In sum,
courts may do so when (1) the private entity is engaging in a public
function;108 (2) public and private entities are intertwined in a close
relationship;109 or (3) government ratifies and adopts the action of the
private entity.110
The best example of the “public function” exception is primary
elections. Political parties run these elections, but they nonetheless
engage in a public function while doing so. Courts have thus held
they are bound by equal protection restrictions on discrimination.111
107. See McDonald v. City of Chicago, 130 S. Ct. 3020, 3042 (2010) (discussing the incorporation doctrine). The Supreme Court has held most of the individual rights guaranteed by the Bill of Rights to be “incorporated” and applicable to the States, although some
have never been addressed by the Court. CHEMERINSKY, supra note 37, § 16.3 (discussing
the “selective incorporation” doctrine). The Court has specifically declined to incorporate
the Fifth Amendment right to a grand jury and the Seventh Amendment right to a jury
trial in civil lawsuits. Branzburg v. Hayes, 408 U.S. 665, 687 (1972) (Fifth Amendment
right to Grand Jury not incorporated); Minneapolis & St. Louis R.R. Co. v. Bombolis , 241
U.S. 211, 217 (1916) (Seventh Amendment right to civil jury trial not incorporated); see
also McDonald, 130 S. Ct. at 3035 n.13 (providing the latest statement by the Court of the
status of selective incorporation doctrine). Thus, even if the grand jury or civil jury trial
rights could sensibly and feasibly be applied to private employers under the PPE, see infra
Part V.A.3, they would be poor candidates for the PPE.
108. See, e.g., Terry v. Adams, 345 U.S. 461, 469-70 (1953) (finding state action where
private political party ran primary election).
109. See, e.g., Burton v. Wilmington Parking Auth., 365 U.S. 715, 725 (1961) (finding
that where government agency leased space to private parking authority, it created “interdependence” and “joint partnership” in the challenged discrimination sufficient to establish
state action).
110. See, e.g., Shelley v. Kraemer, 334 U.S. 1, 18-19 (1948) (finding state action where
court enforced racially restrictive covenant in private deed).
111. See, e.g., Terry, 345 U.S. at 469-70 (1953) (designating a private group’s running of
a statewide primary election as state action for purposes of constitutional limitations on
race discrimination); Smith v. Allwright, 321 U.S. 649, 663-65 (1944) (declaring a private
political party’s racially discriminatory motives in conducting a primary election to be unconstitutional because it was operating under statutory authority of the state).
964
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
Courts also have found state action present under this category in
older cases involving “company towns,” settlements owned and operated by a corporation but that functioned as municipalities for the
employee residents, with the company maintaining streets, providing
sanitation, and the like.112
The second category arose in Burton v. Wilmington Parking Authority, which involved a government agency that leased space to a
private company running a parking lot.113 The Court noted that the
government agency stood to profit from the racial discrimination being committed by the private company.114 Thus, the Court reasoned,
the two entities were “interdependen[t],” with the government essentially a “joint participant” in the discrimination.115 State action was
therefore present.116
The final category first arose in a case where a court was called
upon to enforce a racially discriminatory restrictive covenant in a
dispute between two private parties.117 The Court held that the trial
court’s action to enforce the covenant amounted to ratification of the
discrimination and thus state action.118
This line of authority may inform courts’ decisions as to when to
use constitutional principles in enforcing the PPE. Where the private
employer is engaging in a public function, as in a private prison, the
argument may be stronger that it should afford its employees constitutional rights comparable to those enjoyed by public employees.
Courts also might be more open to such arguments where the private
employer is a monopoly, wielding the kind of power analogous to that
of the government.119 Analogizing from the second category, the same
112. See, e.g., Marsh v. Alabama, 326 U.S. 501, 507-08 (1946) (enforcing First and
Fourteenth Amendment rights for a woman distributing religious material in a companyowned suburb because it was accessible to the public and functioned as any other town
within the community); cf. Citizens to End Animal Suffering & Exploitation, Inc. v. Faneuil
Hall Marketplace, Inc., 745 F. Supp. 65, (D. Mass. 1990) (holding that a private corporation
could not regulate free speech activities on land it had leased from city because the land
was a public forum for public use).
113. 365 U.S. 715 (1961).
114. Id. at 725.
115. Id.
116. Id. The continuing vitality of Burton is under serious question, and its holding
may be sharply limited to its facts or else to the specific situation of private commercial
leasing of government property. Lebron v. Nat’l R.R. Passenger Corp., 513 U.S. 374, 409
(1995) (citing L. TRIBE, AMERICAN CONSTITUTIONAL LAW § 18-3, at 1701 n.13 (2d ed. 1988)).
117. Shelley v. Kraemer, 334 U.S. 1 (1948).
118. Id. at 18-20.
119. Turner, supra note 25, at 292-93, 317 (2013) (arguing that private action should be
treated as state action where the private entity enjoys a monopoly). This argument makes
sense from a market perspective. Where an employer effectively holds a monopoly within a
particular job market, employees have markedly limited ability to seek employment with a
competitor if they consider employee policies too rights-restrictive. See Gay Law Students
Ass'n v. Pac. Tel. & Tel. Co., 595 P.2d 592, 595 (Cal. 1979) (making this observation).
2014]
COMPANY CONDUCT
965
result may be obtained if the private employer is a government contractor whose operations are closely intertwined with those of government, as are some military contractors engaged in combat operations with U.S. armed forces.120 Finally, where the government is
asked to enforce a problematic rule of a private employer—for example, an overly speech- or privacy-restrictive rule that provides not
only for termination, but also allows the employer to claim some reimbursement of recently accrued business expenses from the employee—courts may be less willing to affirmatively enforce such provisions if the employer sues the employee, and be less willing to countenance the discharge in the first place if called upon by the employer
to do so.
IV. THE RELEVANCE OF EXISTING STATUTORY REMEDIES
In evaluating the eligibility of constitutional provisions as sources
for the public policy behind the application of the PPE, one issue
bound to emerge is the applicability (if any) of statutory remedies for
the kind of wrongful discharge at issue. This is a double-edged sword.
On the one hand, the existence of relevant statutes covering the same
or similar types of alleged wrongful discharge provides powerful evidence that the related constitutional provision does indeed express a
strongly held, substantial, and clear public policy. On the other hand,
the existence of statutory remedies might suggest that the PPE
would be unnecessary.121
However, there are a number of instances in which a PPE claim
might be needed, even where a statutory claim exists covering the
same general grounds for termination. For example, the availability
of the PPE drawing upon constitutional principles is a relevant concern in those cases where the employee is precluded from pursuing a
statutory claim by missing a limitations period,122 failing to exhaust
administrative remedies,123 or working for an employer with too few
120. Of course, one rationale for using such private security forces is their ability to act
without constitutional restraints. But that very ability has become highly controversial.
See Earl. F. Martin, America’s Anti-Standing Army Tradition and the Separate Community
Doctrine, 76 MISS. L.J. 135, 135 (2006) (analyzing the “separate community doctrine” to
highlight the uncertain balance between civilian society and the armed forces).
121. Cf. Borden v. Johnson, 395 S.E.2d 628, 630 (Ga. Ct. App. 1990) (refusing to recognize the PPE for dismissal based on gender or pregnancy due to a lack of state or federal
statutes specifically addressing such rights).
122. Under Title VII of the Civil Rights Act of 1964 (“Title VII”), employees must file an
EEOC complaint within 180 days of discovering the alleged discrimination. 42 U.S.C. § 2000e5(e) (2012). Once the EEOC either fails to respond within 180 days or issues a “right to sue”
letter, the employee then must file a complaint in federal court within ninety days. § 2000e-5(f).
123. Rojo v. Kliger, 801 P.2d 373, 389 (Cal. 1990) (allowing PPE claim to proceed where
plaintiff was barred from analogous statutory claim for failing to exhaust administrative
remedies); see also Midgett v. Sackett-Chicago, Inc., 473 N.E.2d 1280, 1285 (Ill. 1984) (col-
966
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
employees to trigger coverage under the statute.124 Yet another example would be where the common law claim provides for needed
remedies that are not provided for under the corresponding statute.125
This latter problem of employer size may be a relatively common
occurrence, since most United States employers have five or fewer
employees126 and thus would be outside the scope of many federal and
state employment antidiscrimination statutes. Indeed, state courts
have recognized PPE claims in exactly such circumstances.127 But,
there is still significant variation among the states as to whether PPE
claims should go forward against employers too small to be covered by
the applicable statute, with some states rejecting such PPE claims.128
Statute of limitations issues are another salient example. Limitation periods for PPE claims vary. Generally, limitation periods for
common law tort claims range from one to three years.129 States have
lective bargaining agreement requirement that employees go through internal grievance
procedure did not bar PPE claim, since the latter sounded in tort and not contract).
124. See § 2000e(b) (defining an employer for purposes of Title VII as one with fifteen
or more employees); Thurdin v. SEI Bos., LLC, 895 N.E.2d 446, 448-49, 455, 455 n.18
(Mass. 2008) (discussing alternate liability theory of, inter alia, PPE where state pregnancy antidiscrimination statute applied only to employers with more than six employees);
Roberts v. Dudley, 993 P.2d 901, 909-11 (Wash. 2000) (allowing PPE claim based on sex
discrimination to proceed against small business where state sex antidiscrimination statute exempted small employers).
125. Even if a PPE claim was necessary to assist a particular employee plaintiff for one
of these reasons, it might suffice for the court to draw upon the existing statutes for inspiration as to the nature of the clear public policy, avoiding reference to constitutional principles as unnecessary. But because there is no good reason for treating constitutional
sources as somehow inferior to statutory ones in divining public policy, see supra Part
III.A., it seems unwise for a court to wear such blinders.
126. According to 2012 Census Bureau estimates, in nine of the ten most populous
states, the majority of employers have fewer than five employees. See Firm Size Data, U.S.
SMALL BUS. ADMIN., http://www.sba.gov/advocacy/849/12162 (last visited Feb. 16, 2014)
(spreadsheet based on census data for California, Texas, New York, Florida, Illinois, Pennsylvania, Ohio, Georgia, Michigan, and North Carolina).
127. See, e.g., Thurdin v. SEI Bos., LLC, 895 N.E.2d 446, 448-449 n.18 (Mass. 2008)
(discussing alternative liability theory of, inter alia, PPE where state pregnancy antidiscrimination statute applied only to employers with more than six employees); Badih v.
Myers, 43 Cal. Rptr. 2d 229 (Cal. Dist. Ct. App. 1995) (recognizing PPE claim involving
pregnancy discrimination, even though employer had fewer than the five individuals required to be covered by the analogous anti-discrimination statute); Molesworth v. Brandon,
672 A.2d 608 (Md. 1996) (ruling in like manner regarding sex discrimination); Collins v.
Rizkana, 652 N.E.2d 653 (Ohio 1995) (same, for sexual harassment PPE claim); Roberts,
993 P.2d at 901 (same, for pregnancy discrimination); Williamson v. Greene, 490 S.E.2d 23
(W. Va. 1997) (same, for sex discrimination and harassment).
128. See Thibodeau v. Design Grp. One Architects, LLC, 802 A.2d 731, 745-46 (Conn.
2002) (declining to find PPE extended to pregnancy-related discrimination by employers
with fewer than the three employees required by statute, and explicitly stating that common law doctrines could not trump such statutory exemptions); Weaver v. Harpster, 975
A.2d 555, 567 (Pa. 2009) (declining to find employee’s sex discrimination claims actionable
under PPE because workplace employed fewer than four employees, the amount required
to invoke remedy under Pennsylvania Human Relations Act).
129. PERRITT, supra note 32, § 9.15.
2014]
COMPANY CONDUCT
967
adopted a similar range for PPE claims, applying limitation periods
of at least one year, and often up to three years.130 Because some
state and federal antidiscrimination statutes involve requirements
that the complainant file a claim in a shorter time period,131 there
may be situations in which an employee victim of discrimination, who
otherwise might have pursued a typical discrimination claim, may
need to pursue a PPE claim instead.132
Further, states vary as to whether the limitation period begins to
run when the termination actually occurs, when the employee receives notice of termination, or at the moment the employer decides
to terminate.133 Again, where this deviates from the rule regarding an
analogous statutory claim, a PPE claim may be viable where the
statutory claim may no longer be viable. Thus, in a few jurisdictions
courts have allowed a common law action based on public policy to
proceed, even if the limitations period of the relevant employment
antidiscrimination statute has expired.134
130. See, e.g., VA. CODE ANN. § 8.01-248 (2013) (two years for Virginia); Kamen v. Int’l
Bhd. of Elec. Workers, 505 F. Supp. 2d 66, 78 (D.C. Cir. 2007) (three years for D.C., under
the “catch-all” statute of limitations); Funk v. Sperry Corp., 842 F.2d 1129, 1133 (9th Cir.
1988) (California PPE limitation period of one year); Holien v. Sears, Roebuck & Co., 689
P.2d 1292, 1301 (Or. 1984) (two years for Oregon); Barnett v. Sequim Valley Ranch, LLC,
302 P.3d 500, 505 (Wash. Ct. App. 2013) (three years for Washington).
131. See, e.g., 29 U.S.C. § 160(b) (2012) (six months for unfair labor practice claims
under the NLRA); 29 U.S.C. § 626(d) (2006) (same, for ADEA claims); 42 U.S.C. § 2000e5(e) (2012) (depending on state, federal Title VII claims must be filed within six to ten
months); OHIO REV. CODE ANN. § 4112.02(N) (West 2013) (six months to file age discrimination claim); TENN. CODE ANN. § 8-50-103 (West 2009) (six months to file employment
discrimination claim).
132. For some types of PPE claims, though, this may not matter. When a federal claim
does not have an express statute of limitation, federal courts adopt the statute of limitations of an analogous state law claim, unless doing so would interfere with the substantive
policy behind the federal claim. DelCostello v. Int’l Bhd. of Teamsters, 462 U.S. 151, 158
(1983). But for all federal statutory causes of action-based laws passed after 1990, Congress has adopted a “catch-all” limitations period of four years. 28 U.S.C. § 1658 (2012).
133. Compare Stupek v. Wyle Labs. Corp., 963 P.2d 678, 682 (Or. 1998) (determining
that limitation began to run when termination became effective), and Lorenz v. Martin
Marietta Corp., 802 P.2d 1146, 1149 (Colo. Ct. App. 1990) (same), with Huff v. Great W.
Seed Co., 885 P.2d 723, 726 (Or. Ct. App. 1994) (finding that one-year statute of limitations
began to run when plaintiff found out her dismissal was related to her workers’ compensation claim, not the date of actual termination).
134. See, e.g., Wendeln v. Beatrice Manor, Inc., 712 N.W.2d 226, 238 (Neb. 2006) (concluding that public policy-based retaliatory discharge claim based in tort was governed by
the general four-year statute of limitations period found in NEB. REV. STAT. § 25-207 (2004)
rather than employment discrimination claim under the NFEPA); Klopfenstein v. NK
Parts Indus., Inc., 870 N.E.2d 741, 747 (Ohio Ct. App. 2007) (finding at-will employee’s
independent
claim
for wrongful discharge
was
governed
by
general
fouryear statute of limitations rather than one-year statute of limitations for claims brought
under the workers’ compensation anti-retaliation statute); Pytlinski v. Brocar Prods., Inc.,
760 N.E.2d 385, 388-89 (Ohio 2002) (applying general four-year limitations period to an
independent cause of action based in common law for violation of public policy where employee’s OSHA complaint was otherwise barred by the 180-day limitations period of the
Ohio Whistleblower Act).
968
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
Similarly, a PPE claim may be relevant despite the existence of an
analogous statutory claim, because the two claims may involve different potential remedies.135 When an analogous statutory cause of
action allows for administrative remedies, courts have still recognized an independent cause of action for the public policy claim.136
Of course, even if there are procedural or remedial reasons why an
employee needs to bring a PPE claim, despite the existence of a statutory claim protecting the same type of employee behavior, the
jurisdiction may not be willing to allow the PPE claim to proceed.
After all, there are reasons why jurisdictions bar antidiscrimination
statutes’ coverage of firms with fewer than the minimum number
of employees. One is to limit the drain on enforcement resources of
the courts, the EEOC, and similar state enforcement agencies.137 Another is to limit the hardship on small “mom and pop” operations,
which may be more vulnerable to the burdens of regulation and litigation.138 Although the first concern does not really apply with respect to judicial enforcement of a common law PPE claim, the second
policy rationale does.
Similarly, there are reasons why we have statutes of limitations:
ensuring certainty and finality, saving defendants from abusive litigation, and saving courts from adjudicating stale claims with fading
evidence.139 These concerns simply do not go away when the plaintiff
pursues a common law PPE claim. Title VII, like many state antidiscrimination statutes, requires administrative procedures for similar
logistical, screening, and judicial economy reasons.140 To the extent
135. See Roberts v. Dudley, 993 P.2d 901, 909-11 (Wash. 2000) (en banc) (holding that
statutory claim, but not wrongful discharge claim, provided for administrative proceedings
and potential attorney’s fee remedies).
136. See, e.g., Holien v. Sears, Roebuck & Co., 689 P.2d 1292, 1303-04 (Or. 1984) (determining a female employee was entitled to common law tort damages based on sexual harassment because the equitable remedies available through employment antidiscrimination
statute were inadequate to compensate plaintiff for the personal nature of her injuries).
137. See LARSON, supra note 29, § 174.06[3] (discussing divergence of court decisions on
this question, and reasons for and against allowing PPE claim to proceed).
138. Id.; see also Thibodeau v. Design Grp. One Architects, LLC, 802 A.2d 731, 740-41
(Conn. 2002) (affirming that the primary purpose of exempting small employers from employment discrimination claims is to avoid subjecting such employers to the significant
financial burdens of litigation). Aside from regulatory burden, there is also the more abstract libertarian concern regarding such “mom and pop” firms, which often involve a more
intimate setting raising greater “freedom of association” issues. See infra Part V.B.1.
139. See John R. Sand & Gravel Co. v. United States, 552 U.S. 130, 133 (2008) (stating
that the primary purpose of most statutes of limitations is “to protect defendants against
stale or unduly delayed claims”).
140. See Title VII, 42 U.S.C. § 2000e-5 (2012); see also, e.g., ARIZ. REV. STAT. ANN. § 411463 (2013) (prohibiting employers from hiring, segregating, or classifying employees on
the basis of “race, color, religion, sex, age or national origin or on the basis of disability”);
NEB. REV. STAT. § 48-1104(1) (2012) (declaring discrimination on the basis of “race, color,
religion, sex, disability, marital status, or national origin” an unlawful employment practice); OR. REV. STAT. ANN. § 659A.030 (2013) (stating it unlawful to “refuse to hire or em-
2014]
COMPANY CONDUCT
969
that the coverage, time limitation, and administrative exhaustion
provisions of these statutes leave “gaps” that are conscious policy determinations rather than oversights, there may be sound policy reasons for rejecting PPE claims in such situations.141
Thus, where the same type of employee behavior may be protected
by both claims, courts may prefer to hold that the statutory remedy
renders unnecessary the common law claim, both to show deference
to the legislative branch and also to honor the common law
presumption of at-will employment. A court could thus hold that the
statute exhaustively outlines the manner in which such employee
rights are vindicated.142
However, such seeming “preemption” appears not to be the majority approach. Those courts that have addressed the issue have tended
to say that the existence of a statutory remedy protecting the same
type of employee conduct does not bar a PPE claim, at least where
that statute is not the sole source for the clear public policy the court
derives in recognizing a PPE claim under those facts.143 Instead,
courts will look to see if there is any indication of legislative intent
that the statutory remedy, or remedies, provided were designed to
replace, rather than supplement, preexisting common law remedies
like the PPE. In the absence of such indication, the common law
cause of action continues.144 Therefore, where, in a given case, the
statutory remedy is unavailable due to some procedural or coverage
ploy . . . bar or discharge” any individual from employment on the basis of “race, color, religion, sex, sexual orientation, national origin, marital status or age”).
141. For an excellent discussion of these issues, see LARSON, supra note
29, §§ 174.06[2]-.06[4].
142. See Parlato v. Abbott Labs., 850 F.2d 203, 206-07 (4th Cir. 1988) (dismissing
plaintiff’s common law claims for wrongful discharge based on race and age because statutory remedy already existed); Rupp v. Purolator Courier Corp., 790 F. Supp. 1069, 1073 (D.
Kan. 1992) (Title VII and the Kansas Act Against Discrimination provided exclusive remedies for employee's common-law claims of retaliation, constructive discharge and whistleblowing); Northrup v. Farmland Indus., 372 N.W.2d 193, 197 (Iowa 1985) (stating that “the
procedure under the civil rights act is exclusive, and a claimant asserting a discriminatory
practice must pursue the remedy provided by the act”).
143. Phillips v. J.P. Stevens & Co., Inc., 827 F. Supp. 349, 352-53 (M.D.N.C. 1993) (recognizing that the availability of statutory remedy does not affect the availability of an alternative wrongful discharge claim); Collins v. Rizkana, 652 N.E.2d 653, 660 (Ohio 1995)
(noting that a statute will not preempt common law remedies unless it is the sole source of
the public policy); Atkinson v. Halliburton Co., 905 P.2d 772, 774 (Okla. 1995) (quoting
Tate v. Browning-Ferris, Inc., 833 P.2d 1218, 1226 (Okla. 1992)) (“[W]here the common law
gives a remedy, and another is provided by statute, the latter is merely cumulative, unless
the statute declares it to be exclusive”).
144. See Collins, 652 N.E.2d at 660; Atkinson, 905 P.2d at 774. Of course, the legislature can also make express its intent to allow a parallel common law cause of action continued viability. See, e.g., N.J. STAT. ANN. § 34:19-14 (West 2013) (allowing plaintiff to
choose between suing under statute or under common law public policy claim).
970
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
problem, the common law PPE tort can fill in the gap.145 This result
can also be obtained where the analogous statutory claim is based on
a federal statute.146
A number of courts taking this approach have tended to allow PPE
claims to go forward despite parallel statutory remedies. For example,
the Kentucky Supreme Court recently acknowledged that a statute
protecting employees’ rights to bear arms explicitly recognized a civil
cause of action, yet it also held that an employer’s violation of the statute violated a wider public policy, thus creating a PPE cause of action.147 Indeed, the Supreme Court of Wisconsin has explained that in
deciding whether a clear public policy exists, courts should not focus on
the “literal language” of a statutory provision or the circumstances it
describes but should examine whether the employer “contravene[d] the
spirit as well as the letter of a constitutional, statutory, or administrative provision.”148 Thus, courts have allowed PPE claims to go forward
despite the existence of a comparable statutory claim.149
The above approaches—leaning away from a general “preemption”
rule where statutory language exists, and declining to hold that the
absence of such statutory language definitively excludes the recognition of a common law PPE claim—make sense, for reasons similar to
145. See, e.g., Thibodeau v. Design Grp. One Architects, LLC, 802 A.2d 731, 733 (Conn.
2002) (recognizing a PPE where a female employee was fired for missing work due to pregnancy even though she was one of two employees and the statutory remedy applied only to
employers with three or more employees); Mitchell v. Univ. of Ky., 366 S.W.3d 895, 903
(Ky. 2012) (allowing claim for wrongful discharge based on gun ownership despite analogous statutory claim).
146. Wheeler v. Caterpillar Tractor Co., 485 N.E.2d 372, 375-76 (Ill. 1985) (holding that
the federal nuclear safety statute with whistleblower protection provision did not preempt
state law PPE claim).
147. Mitchell, 366 S.W.3d at 903.
148. Strozinsky v. Sch. Dist. of Brown Deer, 614 N.W.2d 443, 455 (Wis. 2000) (quotation marks omitted).
149. Phillips v. J.P. Stevens & Co., Inc., 827 F. Supp. 349, 352-53 (M.D.N.C. 1993) (recognizing that the availability of statutory remedy does not affect the availability of an alternative wrongful discharge claim); Collins v. Rizkana, 652 N.E.2d 653, 660 (Ohio 1995) (noting
that a statute will not preempt common law remedies unless it is the sole source of the public
policy); Atkinson, 905 P.2d at 774 (quoting Tate v. Browning-Ferris, Inc., 833 P.2d 1218, 1226
(Okla. 1992)) (“[W]here the common law gives a remedy, and another is provided by statute,
the latter is merely cumulative, unless the statute declares it to be exclusive.”).
The converse is also true: the absence of statutory language barring the kind of
conduct at issue does not necessarily preclude recognition of the PPE. Though a court
could take an expressio unius est exclusio alterius approach and infer a decision to exclude
other common law remedies where a statutory remedy exists, courts do not seem to do so.
Payne v. Rozendaal, 520 A.2d 586, 588 (Vt. 1986) (recognizing PPE cause of action for age
discrimination even though termination occurred prior to the passage of state legislation
barring such discrimination).
This approach is sensible. At bottom, the question is a straightforward one of
statutory construction. Sometimes the legislature intended to “preempt” common law causes of action, and sometimes it did not. Courts can, of course, use tools of statutory construction to divine legislative intent
2014]
COMPANY CONDUCT
971
the continued recognition of common law causes of action in the first
place. The legislature cannot anticipate all potential situations when
drafting statutory language. Case-by-case, incremental development
of legal doctrine is useful for that reason and also because it provides
the flexibility needed for adapting older principles to new situations.
These new situations emerge constantly, and the legislature cannot
always keep up.
Such judicial latitude is nonetheless consistent with the general
preference for having the legislature make the kinds of policy
decisions involved in formulating a PPE.150 Where the legislature intends a statutory remedy to be exhaustive and exclusive of common
law remedies, it need only say so explicitly to prevent judicial overreach. Such language effectively preempts otherwise applicable
common law causes of action.151 The legislature also can assert such
preeminence after the fact: when a court applies an overly broad
PPE, the legislature may legislatively overrule the decision with
proper preemption language.152
Also, recognition of PPE claims based in whole or in significant
part on constitutional principles may be especially appropriate. After
all, constitutional principles are foundational. They are designed to
supersede statutory and regulatory principles. And by its very na150. See, e.g., Borden v. Johnson, 395 S.E.2d 628, 629 (Ga. Ct. App. 1990) (discussing
the court’s precedent resting on the principle that the legislature is the body most qualified
to set public policy for purposes of interpreting the PPE); Horn v. N.Y. Times, 790 N.E.2d
753, 756-57 (N.Y. 2003) (same).
151. See, e.g., ARIZ. REV. STAT. ANN § 23-1501B (West 2012) (remedies provided for in
statutes creating liability for wrongful discharge at the sole and exclusive remedies for
vindicating any “public policy . . . arising out of the statute”).
152. Analogizing once again from constitutional principles, courts can also find “implied preemption” in some circumstances. Where the state has so pervasively regulated in
an area to create an inference that it intended to “cover the field,” courts may decide that
the existence of statutory remedies for a given situation may indeed crowd out a PPE
claim. Similarly, where the recognition of a PPE claim would stand as a substantial “obstacle” to the fulfillment of the legislature’s underlying policy in an area, courts may infer a
different category of implied preemption. See Arizona v. United States, 132 S. Ct. 2492,
2502-03 (2012) (describing “field” and “obstacle” preemption).
For the reasons stated immediately above, these kinds of “implied preemptions” of
PPE claims should not be the norm. Nor do they seem to be the existing norm, according to
the case law. See supra notes 143-149 and accompanying text. The mere existence of a
statutory remedy based on a comparable ground for termination need not create a negative
inference of an intent to have such ground be the sole remedy. The law is filled with examples of parallel statutory and common law remedies, and having alternate theories of liability can serve a salutary purpose. Nonetheless, there may indeed be especially compelling
occasions where legislative history, statutory purpose, or the overall structure of the statutory scheme in a given state evince a legislative purpose to crowd out all common law
claims, or at least PPE claims, even in the absence of express preemption language in the
statutes. Allowing courts the ability to recognize such “implied preemption” provides needed flexibility. It also provides adequate deference to legislative prerogatives, effectively
rebutting criticisms that recognition of a PPE cause of action parallel to an analogous statutory remedy represents judicial usurpation.
972
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
ture, PPE liability exists only when the public policy at issue is both
clear and substantial. If any public policy deserves vindication even
when the facts place it in a “gap” within the statutory framework, it
would be a policy that is clear, substantial, and foundational.
In sum, the existence of statutory remedies for comparable employer actions involving similar public policies may or may not warrant precluding a supplemental PPE claim. Reasonable arguments
exist on both sides. Resolution of the question should most appropriately turn on the specifics of the intent of the legislature which
passed the statute in question: whether it intended to supplant or
supplement common law remedies.153 But at least where such remedies do not “crowd out” the PPE, courts should still be able to look to
constitutional as well as statutory sources of authority.
V. SPECIFIC CONSTITUTIONAL PROVISIONS AS CANDIDATES FOR THE
PPE
If constitutional provisions can indeed inform a court’s analysis of
the prevailing public policy under the PPE, which types of constitutional provisions can most properly fill that function? Are there any
constitutional doctrines that are particularly appropriate, or inappropriate, for this role?154 To begin to answer this question, this Article uses provisions of the U.S. Constitution as examples, keeping in
mind that state constitutions often have analogous provisions.155
153. See supra note 144 and accompanying text.
154. One overarching consideration might be the constitutional standard of review
applied when evaluating alleged deprivations of a specific constitutional right. Obviously,
the higher the standard of review, the more important the right, and therefore, the stronger the employee’s argument for the PPE. Most constitutional claims of alleged deprivations
of individual rights are evaluated using (from most lenient to most exacting) “rational basis” review, “intermediate scrutiny,” or “strict scrutiny.” 3 RONALD D. ROTUNDA & JOHN E.
NOWAK, TREATISE ON CONSTITUTIONAL LAW: SUBSTANCE AND PROCEDURE § 18.3(a)(v) (4th
ed. 2008). In most of the examples considered infra Parts V.B-C, as potential candidates for
the PPE, the applicable standard is either “strict scrutiny,” e.g., free speech, or “intermediate scrutiny,” e.g., gender discrimination. See United States v. Virginia, 518 U.S. 515, 570
(1996) (Scalia, J., dissenting) (noting that the Court applies intermediate scrutiny in gender discrimination cases); Perry Educ. Ass’n v. Perry Local Educators’ Ass’n, 460 U.S. 37,
45 (1983) (“For the state to enforce a content-based exclusion it must show that its regulation
is necessary to serve a compelling state interest and that it is narrowly drawn to achieve that
end.” (citing Carey v. Brown, 447 U.S. 455, 461 (1980)). In a few cases, the standard of review
is outside the traditional three tiers of constitutional review (abortion) or still undefined (the
Second Amendment). See McDonald v. City of Chicago, 130 S. Ct. 3020, 3046-48 (2010) (applying individual Second Amendment right to bear arms to the states but failing to specify a
particular standard of review); Planned Parenthood of Se. Pa. v. Casey, 505 U.S. 833, 869
(1992) (applying unique “undue burden” standard for abortion rights).
155. See, e.g., W. VA. CONST. art. III, § 7 (“No law abridging the freedom of speech, or
of the press, shall be passed.”).
2014]
COMPANY CONDUCT
973
A. Provisions that Should Not Form the Basis of a PPE Claim
1. “Structural” Constitutional Doctrines
There are certain types of constitutionally based policies which
are obviously inappropriate sources for the PPE. Some are structural
policies that do not implicate individual rights and would thus afford
no basis in drawing an analogy with respect to the interests of an individual employee fired from her job. For example, the doctrine of
separation of powers is a well-respected, longstanding policy that is
arguably both substantial and clear.156 But no terminated employee
could plausibly claim her termination violated this policy. The same
is true with respect to the doctrine of federalism.
2. Restrictions Uniquely Focused on Government
Similarly, inherent in Establishment Clause doctrine is the idea
that it is only the government, not private entities that may not endorse religion.157 While as a legal matter, all federal constitutional
provisions restrict only government action, the Establishment
Clause, by its very nature, focuses uniquely on government action.
Thus, the doctrine stands in contrast to constitutional provisions
guaranteeing individual rights of privacy, free speech, or nondiscrimination, all of which can easily be analogized to a private setting.
True, one could imagine situations in which a terminated employee might coherently invoke the values affirmed by this Clause: say, a
Jewish employee fired for objecting to a “Keep Christ in Christmas”
sign outside a workplace, to printing Christian Bible verses on all
156. Actually, while the existence of a separation of powers doctrine is clear, its contours are not. See ROTUNDA & NOWAK, supra note 154, § 3.12(a) (describing the concept of
separation of powers as one that evades “precise legal definition” and often provides nebulous solutions to intra-governmental disputes); see also Martin H. Redish & Elizabeth J.
Cisar, “If Angels Were to Govern”: The Need for Pragmatic Formalism in Separation of
Powers Theory, 41 DUKE L.J. 449, 450 (1991) (characterizing the modern court’s treatment
of the doctrine as “something of a split personality,” seemingly wavering from strict and
“formalistic” to the use of a lenient, “functional” approach). The uncertain state of the law
in this area is beyond the scope of this piece.
157. See Bd. of Educ. of Westside Cmty. Schs. v. Mergens, 496 U.S. 226, 250 (1990)
(explaining that: (1) the Establishment Clause forbids only government speech endorsing
religion, while private speech endorsing religion is actually constitutionally protected; and
(2) that “there is a crucial difference between government speech endorsing religion, which
the Establishment Clause forbids, and private speech endorsing religion, which the Free
Speech and Free Exercise Clauses protect”); Cooper v. U.S. Postal Serv., 577 F.3d 479, 484
(2d Cir. 2009) (affirming an Establishment Clause violation only where religious material
was displayed on postal units serving a public function, but declaring the private business
surrounding in which the units operated free from such restrictions); Ams. United for Separation of Church & St. v. City of Grand Rapids, 980 F.2d 1538, 1553 (6th Cir. 1992) (holding that a private organization’s menorah display in a public forum did not constitute an
endorsement of religion for purposes of Establishment Clause restrictions).
974
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
paychecks, or to holding weekly Christian evangelical rallies on employer premises. Indeed, at a certain point, such situations might
trigger employer liability for religious harassment under Title VII or
analogous state antidiscrimination statutes.158 But applying a “no
endorsement of religion” restriction on a private employer seems
overly intrusive. Indeed, a court order preventing an employer from
endorsing religion would likely violate the employer’s rights under
the Free Speech and Free Exercise Clauses.159
Certainly, one might have legitimate concerns about an abusive
employer coercing employees into engaging in unwelcome religious
observances or preventing employees from engaging in their own
non-disruptive religious practices. Can the employer require its Jewish employee to attend Christian services, make him sport a “Keep
Christ in Christmas” button, or prevent him from wearing a yarmulke? But such concerns can be addressed adequately by reliance on the
Free Exercise Clause, discussed below, which is indeed designed to
protect individual rights.160 Thus, the Establishment Clause makes a
poor candidate for a source of public policy underlying a PPE claim.
For the same reason, the Fifth Amendment Takings Clause guarantee of “just compensation” also seems an unlikely source for policy
under the PPE. It clearly focuses uniquely on the responsibility of
government to reimburse private parties for “takings” by that government. Disputes under the Clause typically arise in cases involving
eminent domain, a doctrine uniquely applicable to the right of the
government to take private property for public uses.161 Moreover,
where it is a private employer who has engaged in some form of analogous “taking,” a common law PPE claim is entirely superfluous. A
more mundane common law claim in contract, property, or tort (e.g.,
conversion, tortious damage to property) will be a far more straightforward path to just compensation.
158. See EEOC v. Townley Eng’g & Mfg. Co., 859 F.2d 610, 620-21 (9th Cir. 1988) (recognizing the ability of employers to express themselves in the workplace but distinguishing
this ability from coercing employees to conform with conflicting religious convictions);
Thomas C. Berg, Religious Speech in the Workplace: Harassment or Protected Speech?, 22
HARV. J.L. & PUB. POL'Y 959, 1002 (1999) (“[W]hen an employer encourages a worker to
attend church, the worker is likely to feel more pressure to do so than if a co-worker urged
it. These activities, then, are quite likely to be prima facie covered by Title VII or analogous
state anti-harassment statutes.”).
159. Mergens, 496 U.S. at 250 (“[T]here is a crucial difference between government speech
endorsing religion, which the Establishment Clause forbids, and private speech endorsing
religion, which the Free Speech and Free Exercise Clauses protect.”).
160. See CHEMERINSKY, supra note 37, § 12.1.1 (stating that the Free Exercise Clause
“clearly safeguards individual liberty” while the Establishment Clause “seems directed at
the government”).
161. Armstrong v. United States, 364 U.S. 40, 49 (1960) (describing the aim of the Takings Clause as preventing the government "from forcing some people alone to bear public
burdens which, in all fairness and justice, should be borne by the public as a whole”).
2014]
COMPANY CONDUCT
975
If, for example, an employer required employees to contribute
their own money, property, or uncompensated services toward the
employer’s enterprise, or toward the employer’s own pockets, without
compensation, one could plausibly consider this a deduction from the
employee’s pay, one for which there is either no liability or liability in
contract, property, or tort. An employee fired for protesting such
harsh contractual terms would have either no PPE remedy or else a
remedy analyzable under Free Speech principles.162 As an example of
the latter, consider an employee terminated for reporting or protesting “wage theft;”163 if she has a PPE claim at all, it would seem to lie
better within a whistleblower theory than under some sort of analogy
to a “taking.”
At least one court has explicitly distinguished constitutional provisions which by their very nature apply only to government action.
For example, in Booth v. McDonnell Douglas Truck Services, Inc., the
Pennsylvania Superior Court rejected a wrongful discharge claim
based on a provision of the state constitution prohibiting any “laws
which impair the obligations of contracts.”164 The court specifically
emphasized that by its plain terms, the constitutional provision bars
only the passage of a “law” impairing contract rights; thus, since the
case involved no allegation of state action, the constitutional provision did not apply.165
3. Due Process/Criminal Procedure Safeguards
Procedural safeguards guaranteed by the Bill of Rights might conceivably form the basis of a PPE claim, most likely because the employer failed to provide an analogous version of them in any investigation of employee misconduct. But it is unreasonable to expect that
every private employer provide every employee charged with jobrelated misconduct the full panoply of procedural protections guaranteed by the Bill of Rights.
In some cases, this is obvious. For example, as applied to a private
employer, the Sixth Amendment right to a trial by jury would seem
to be either inherently inapplicable, overly burdensome, or both.
Courts hardly could be expected to require firms to assemble “juries”
of fellow employees to settle disputed claims of employee misconduct.
162. See infra Part V.B.1.
163. “Wage theft” occurs when an employer unlawfully “deprives a worker of legally
mandated wages,” either by not paying proper overtime compensation, paying less than
what the employee deserves, or not paying the employee at all. KIM BOBO, WAGE THEFT IN
AMERICA 6 (2009); see also Barbosa v. Impco Techs., Inc., 101 Cal. Rptr. 3d 923, 927 (Cal.
Ct. App. 2009) (qualifying an employer’s duty to pay overtime wages as a “well-established
fundamental public policy affecting the broad public interest”).
164. 585 A.2d 24, 28 (Pa. Super. Ct. 1991) (citing PA. CONST. art. I, § 17).
165. Id.
976
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
Similarly inapplicable and/or unduly burdensome are the rights to a
speedy trial, a public trial, and a trial with a proper venue and jurisdiction, as well as the rights to counsel and compulsory process in
one’s defense.166 The right to a grand jury and a jury trial in civil lawsuits are inapt sources of policy for similar reasons and for the additional reason that they do not apply to state governments.167 In other
cases, the inapplicability of traditional criminal procedure safeguards
is less obvious and perhaps less convincing; for example, a Pennsylvania court has rejected a PPE claim based on the right of an accused’s presumption of innocence.168
However, there is no shortage of PPE cases protecting employees
discharged for completing jury service.169 Most of these cases rely on
statutory provisions requiring persons to perform jury service and
protecting those who do from adverse employment action.170 However,
at least one court has based this PPE claim on the constitutional
right to a jury trial.171 To the extent courts can draw upon the constitutional basis of the jury trial right to underscore the importance of
protecting employees’ ability to perform jury service free from retaliation, such use of that constitutional provision is certainly appropriate and uncontroversial. But given that jury service is both required
by statute and protected from retaliation by statute in virtually all
jurisdictions, the use of this constitutional provision is unnecessary
as a practical matter.
The Fifth Amendment’s privilege against self-incrimination
should likely be on this same “ineligible” side of the line, for a somewhat distinct reason. Arguably, it could apply by analogy to private
employment situations. For example, it could be said to apply when
an employer investigating workplace misconduct calls an employee in
166. See U.S. CONST. amend. VI. Protection under double jeopardy doctrine also falls
into this category. See U.S. CONST. amend. V.
167. See supra note 107.
168. Cisco v. United Parcel Servs., Inc., 476 A.2d 1340, 1344 (Pa. Super. Ct. 1984).
169. See, e.g., Wiskotoni v. Mich. Nat’l Bank-W., 716 F.2d 378, 383 (6th Cir. 1983)
(stating that discharge based on subpoena to testify before a grand jury violates public
policy); Call v. Scott Brass, Inc., 553 N.E.2d 1225, 1230 (Ind. Ct. App. 1990) (recognizing
valid PPE claim where employee was terminated for fulfilling duty to appear for jury duty);
Shaffer v. Frontrunner, Inc., 566 N.E.2d 193, 196 (Ohio Ct. App. 1990) (establishing valid
claim for employee fired for jury service); Hodges v. S.C. Toof & Co., 833 S.W.2d 896, 899
(Tenn. 1992) (reinstating lost wages to employee fired for serving on jury duty).
170. See, e.g., Call, 553 N.E.2d at 1230 (reasoning that the public policy was “expressed
by the statutes enacted by the legislature”); Wright v. Faggan, 773 S.W.2d 352 (Tex.
App. 1989) (considering the underlying statutory purpose as a basis for making jury service
a PPE).
171. See Nees v. Hocks, 536 P.2d 512, 516 (Or. 1975); see also Reuther v. Fowler &
Williams, Inc., 386 A.2d 119, 120-21 (Pa. Super. Ct. 1978) (relying on statute making jury service mandatory, as well as state constitutional provision granting the right of
trial by jury, in defining the state’s public policy).
2014]
COMPANY CONDUCT
977
for questioning. This is especially the case if the employer’s grievance
procedures or a collective bargaining agreement provide for some sort
of hearing prior to termination. Furthermore, it would not necessarily be fatally burdensome on an employer to forbid firing (or threatening to fire) all those who wish to remain silent until they can adequately prepare for such a grievance hearing, or until they can consult with a union official, or lawyer, for example.
The Supreme Court has made clear that the reason an accused
has a right to remain silent is that custodial interrogation is inherently coercive.172 It is the ability of the state to use force to detain a
suspect, as well as its ability to threaten a suspect with jail time, that
makes the interrogation process so fraught, and the need for a brightline right to keep silent so imperative.173 Certainly, being summoned
to the boss’s office with one’s job on the line can be an intimidating
experience. But the boss can neither physically force the employee to
stay and cooperate with questioning, nor credibly threaten a significant curtailment of physical liberty in the future (although the threat
of termination can certainly create substantial psychological pressure). For this reason, the right to remain silent seems a less than
compelling candidate for a source of public policy under the PPE.
Finally, the Eighth Amendment’s protections against excessive
fines and “cruel and unusual punishment”174 seem an unlikely PPE
candidate for a different reason. First, in jurisdictions recognizing the
PPE only in cases of actual discharge,175 it would not apply to cases
where an employer imposed fines or other punishments for employee
misconduct short of termination.176 Where adverse employment actions short of termination could give rise to a PPE claim, or where
termination itself is considered the “cruel and unusual punishment,”
applicability still would seem to be rare. Even as to clear state action,
172. Miranda v. Arizona, 384 U.S. 436, 455, 502 (1966) (explaining that such interrogation “exacts a heavy toll on individual liberty and trades on the weakness of individuals”
through “psychologically coercive pressures”).
173. Id. at 512 (emphasizing the important premise that “pressure on the suspect must
be eliminated”).
174 U.S. CONST. amend. VIII.
175. See supra note 47.
176. Of course, were the punishment excessive enough, it might constitute constructive
discharge. Pa. St. Police v. Suders, 542 U.S. 129, 147 (2004) (determining the requirements
for “constructive discharge” to be satisfied where “working conditions [become] so intolerable that a reasonable person would have felt compelled to resign”); see also LARSON, supra
note 29, § 15.08 (stating that most courts apply some form of this “reasonable person” test).
Given that Eighth Amendment claims succeed only in cases of the most extreme, “grossly
excessive” fines and punishments, it may rarely matter whether a jurisdiction recognizes a
PPE claim for action short of termination.
978
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
these clauses apply only in the most extreme cases of grossly excessive fines and punishments.177
This area closely tests the requirement that only policies that “go
to the heart of a citizen’s rights, duties, and responsibilities” can support a PPE claim.178 The public may have a strong interest in ensuring that private employees are not forced to forfeit their privacy,
their right to choose to participate in (or refrain from participating
in) worship, or their freedom of speech about matters of public concern. Does it have a similarly strong interest in policing the presumably rare instance in which an abusive employer, for example, docks
a month’s pay because an employee forgot to properly log her start
time on the sign-in sheet?179 Given the strong presumption toward atwill employment, this seems a less viable candidate for the PPE.
While some of the procedural protections from the Bill of Rights
seem implausible sources of the public policy applicable to the PPE
cases, it still may be that society has an interest in promoting basic
notions of fairness in the workplace, such that the procedural due
process protections created by the Due Process Clause itself may be a
legitimate source of policy. Where a company is large enough to afford internal grievance procedures, for example, it may be appropriate to require that employees using such procedures should be informed of the basis for the allegation of misconduct, and be given notice and an opportunity to rebut the accusation before being discharged.180 This is certainly a requirement for state action under procedural due process, applicable not only in criminal matters, but also
177. See Cooper Indus., Inc. v. Leatherman Tool Grp., Inc., 532 U.S. 424, 425 (2001)
(finding this constitutional constraint to be applicable only where punishments are “grossly
disproportional to the gravity of the offense”); Rummel v. Estelle, 445 U.S. 263, 272 (1980)
(acknowledging that “[o]utside the context of capital punishment, successful challenges to
the proportionality of particular sentences have been exceedingly rare”); see also BMW of N.
Am., Inc. v. Gore, 517 U.S. 559, 559-61 (1996) (declaring a punitive damages award “excessive” only after concluding that none of the elements associated with reprehensible conduct
were present, the ratio of punitive damages to actual harm was five hundred to one, and the
economic sanction was substantially greater than any civil penalty available in the state
where the harm occurred); TXO Prod. Corp. v. Alliance Res. Corp., 509 U.S. 443, 444 (1993)
(admitting that a $10 million punitive damage award alongside a $19 thousand actual damages award was “certainly large” but refusing to classify it as “grossly excessive”).
178. Booth v. McDonnell Douglas Truck Servs., Inc., 585 A.2d 24, 28 (Pa. Super. Ct. 1991).
179. Many state statutes forbid employers from deducting their employees’ pay unless
authorized to do so by federal law, state law, or contract. See, e.g., ARIZ. REV. STAT.
ANN. §§ 23-352, -355 (2013); COLO. REV. STAT. § 8-4-105 (2013); OR. REV. STAT. § 652.610(3)
(2013). An arbitrary, spontaneous employer decision of punishment “overkill” in such a
situation might trigger a statutory cause of action for the employee, arguably mooting any
potential PPE issue. But if such overkill is part of existing managerial policy or practice,
these statutes would be of no help to the employee, for they do not create any minimum
“floor” of protection against unreasonable deductions from pay.
180. Cf. NLRB v. J. Weingarten, Inc., 420 U.S. 251, 266-68 (1975) (holding that National Labor Relations Act supported administrative rule entitling union employee to have
union representative present at any meeting which might lead to disciplinary action).
2014]
COMPANY CONDUCT
979
when a government employer purports to act against a government
employee.181 The requirement of notice, at least, is one that private
employers may need to respect, even where the at-will doctrine otherwise applies.182 Or, if the company perversely punishes an employee for availing herself of those grievance procedures, it may be seen
as violating fundamental fairness.183 In like manner, at least one
court has stated that where the employee bypasses the usual grievance channels, he is less likely to have a claim.184
But this seems to be an area calling for great caution. Due process
cases involving government employer-employee relationships seem
more closely analogous to the PPE situation than due process cases
involving individual defendants in criminal or civil cases. And, in
Due Process doctrine, a government employee has no legally cognizable due process interest, and thus may not even assert a due process
claim, if that employee is an at-will employee as opposed to one who
has a right to continued employment (absent bad behavior) under
either contract or statute.185 It would be incongruous, at the least, and
possibly unduly burdensome and economically inefficient, to apply
stricter standards to private employers via the PPE than are applied
to public employers under the analogous constitutional doctrine.
B. Constitutional Provisions that Should Form the Basis of a PPE
Claim
1. Free Speech: Political Expression
Let us begin with the First Amendment right of free speech. Can
it support a PPE claim? An illustrative case is Novosel v. Nationwide
Insurance Company.186
Novosel, a district claims manager at an insurance company, was
fired after he refused to participate in his employer’s political agenda
181. See Cleveland Bd. of Educ. v. Loudermill, 470 U.S. 532, 542-43 (1985) (balancing
employer interests with employee interests in concluding that due process requires that a
government employee receive notice and an opportunity to respond before being denied a
constitutionally protected property or liberty interest); CHESTER JAMES ANTIEAU &
WILLIAM J. RICH, MODERN CONSTITUTIONAL LAW § 35.18 (2d ed. 1997) (discussing procedural due process rights in public employment dismissal proceedings as requiring proper
notice and an opportunity to respond).
182. See Person v. Bell Atl.—Va., Inc., 993 F. Supp. 958, 961 (E.D. Va. 1998) (characterizing Virginia common law at-will employment doctrine as incorporating requirement
that employer give reasonable notice before termination. in the absence of any agreement
to the contrary).
183. Cf. Wounaris v. W. Va. St. Coll., 588 S.E.2d 406 (W. Va. 2003) (upholding PPE
claim against state actor when college rehired employee after she won an ALJ proceeding,
only to immediately re-fire her while an appeal was still pending).
184. See Geary v. U.S. Steel Corp., 319 A.2d 174,179-80 (Pa. 1974).
185. Bishop v. Wood, 426 U.S. 341, 344-47 (1976).
186. 721 F.2d 894 (3d. Cir. 1983).
980
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
and efforts to lobby the Pennsylvania House of Representatives.187
Defending against a PPE claim, the insurance company argued that
a wrongful discharge action depended upon the violation only of a
statutorily recognized policy, and that a constitutional provision
could not provide the source of a PPE claim.188 The Third Circuit rejected this argument and held for the employee. Even though the
dismissal did not violate a specific statute, it was enough that it contravened a “clearly mandated public policy” that arose “directly from
the Constitution,” i.e., political and associational freedoms.189 Accordingly, the court reasoned that a corporation’s interest in supporting
its state legislature did not outweigh societal interests in free political expression.190 As the Third Circuit explained, “the protection of an
employee’s freedom of political expression would appear to involve no
less compelling a societal interest than the fulfillment of jury service
or the filing of a workers’ compensation claim.”191
Novosel represents perhaps the high-water mark for aggressive
use of constitutional principles for the PPE. The Third Circuit declined to extend Novosel a decade later, holding that intervening
Pennsylvania cases cast doubt on the availability of constitutional
sources, and that, at any rate, the narrowness of the PPE counseled
against extending Novosel to reach the First and Fourth Amendments.192 Instead, the Third Circuit relied on normal tort law (in that
case, invasion of privacy) rather than the “search and seizure” provision of the Fourth Amendment.193
Of course, a too robust PPE may actually interfere with the free
speech rights of employers. Arguably, an employer who wishes to further ideological aims with his business ought to be able to hire (and
fire) based on whether an employee shares the employer’s sense of
mission. An eco-friendly energy company should be able to decide to
hire only “green” employees and to terminate the employee who denies the existence of human-induced climate change. For that matter,
ought not a diehard, conservative, Republican small-business owner
have the right to decide to spend his workdays only with like-minded
employees who share his values—including his party affiliation? At
least, in a small business setting, such an employer may have a free
association interest in choosing the persons with whom he spends
forty hours a week. It is partly for this reason that antidiscrimination
187.
188.
189.
190.
191.
192.
193.
Id. at 895-96.
Id. at 898.
Id. at 899-901.
Id. at 899.
Id.
Borse v. Piece Goods Shop, Inc., 963 F.2d 611, 619-20 (3d Cir. 1992).
Id. at 621-22.
2014]
COMPANY CONDUCT
981
statutes exempt small businesses with only a very small number of
employees.194 The Supreme Court has recognized a right of freedom of
association in membership organizations that is greater for smaller,
more intimate groups than it is for larger, more widespread groups.195
Such a freedom of association is not as great when it comes to choosing with whom one works,196 but it should enjoy some respect.197
It is precisely the right to hire and fire based on party affiliation
which the law denies to a public employer, except for high-ranking
political appointments.198 A government employer cannot fire an
employee for being of the wrong political party.199 But a private employer presumably can: there is no direct First Amendment
protection given the lack of state action, and while some state antidiscrimination statutes list party affiliation as a protected status for
private employees,200 Title VII201 and many state anti-discrimination
statutes do not.202
194. Leroy D. Clark, Employment Discrimination Testing: Theories of Standing and a
Reply to Professor Yelnosky, 28 U. MICH. J.L. REFORM 1, 15 (1994). It is certainly for this reason that part of the Civil Rights Act’s public accommodations provision, 42
U.S.C. § 2000a(b)(1) (2012), and part of the Fair Housing Act, 42 U.S.C. § 3603(b)(2) (2012),
have the so-called “Mrs. Murphy’s boardinghouse” exemption for dwellings with at most four
or five units or families. James D. Walsh, Reaching Mrs. Murphy: A Call for Repeal of the
Mrs. Murphy Exemption to the Fair Housing Act, 34 HARV. C.R.-C.L. L. REV. 605, 606 (1999).
195. Roberts v. U.S. Jaycees, 468 U.S. 609, 621-22 (1984) (stating that a large national
organization with large, otherwise nonselective memberships did not have a First Amendment Free Association right to engage in sex discrimination illegal under state law).
196. Id. (“[T]he Constitution undoubtedly imposes [free association] constraints on the
State’s power to control the selection of one’s spouse that would not apply to regulations affecting the choice of one’s fellow employees.”); see also Guesby v. Kennedy, 580 F. Supp. 1280,
1284 (D. Kan. 1984) (“[I]n the context of an employment relationship, the scope of the right of
association is much more limited than in the context of membership in an organization.”).
197. See Julie Manning Magid & Jamie Darin Prenkert, The Religious and Associational Freedoms of Business Owners, 7 U. PA. J. LAB. & EMP. L. 191 (2005) (arguing that
employers with an avowed religious or theological mission and orientation, even if not
themselves a religious organization, should enjoy a free association-based latitude to prefer
co-religionists as employees).
198. See, e.g., Branti v. Finkel, 445 U.S. 507, 517-19 (1980) (stating that with the exception of certain instances where political affiliation would “interfere with the discharge of
his public duties . . . the continued employment of an assistant public defender cannot
properly be conditioned upon his allegiance to the political party”).
199. See Elrod v. Burns, 427 U.S. 347, 373 (1976) (plurality opinion) (stating that termination of public employees based on party affiliation unconstitutional under the First
and Fourteenth Amendments).
200. CAL. LAB. CODE § 1101 (West 2012) (prohibiting private employers from controlling
or directing the political affiliations of employees); D.C. CODE § 2-1401.01 (2013) (prohibiting
private employers from making employment decisions based on political affiliation).
201. 42 U.S.C. § 2000e-2 (2012).
202. See, e.g., 775 ILL. COMP. STAT. ANN. 5/1-102 (2012) (not including political affiliation
among categories protected from employment discrimination); IOWA CODE § 216.6 (2013)
(same); TENN. CODE ANN. § 4-21-401 (2012) (same); MISS. CODE ANN. § 33-1-15 (2013) (same).
982
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
So it seems sensible to say that not every free speech-centered limit constitutionally imposed on a government employer action ought to
apply to private employers. But that should not mean that there are
no proper free speech-type rights enjoyed by employees. Some should
exist, subject to limitations, based on the considerations below.
(a) Forced Speech
The Novosel case illustrates a prime candidate for such protection,
namely when employers force employees to engage in unwelcome
speech. The problem is real, and in recent years, reports of employers
who have pressured employees into making political campaign contributions or even attending political rallies have risen.203 These employers make attendance at political rallies mandatory, and strongly
encourage monetary contributions.204 They keep track of who cooperates and who does not, with the latter group subject to sanctions like
demotions and the withholding of bonuses.205
Given the rising influence of large corporations on the political
scene,206 this potential to arm-twist armies of surrogate speakers
bodes ill for the political process. This danger to the political process
justifies the application of the PPE to protect free speech values, at
least where the employer is using the employer-employee relationship to unduly influence employees. So, while a private employer may
be able to burden employee free speech values more than the government, this ability is still somewhat constrained: at a minimum,
they should not be able to force their employees to engage in expressive political activity.
Existing FEC regulations prohibit such coercion if designed to result in a financial contribution in a federal election.207 However, such
protections at the state level for state and local elections are far more
rare. Even at the federal level, the regulations apply only to coerced
203. See, e.g., Robert Kelner, Allegations of Employer Coercion of Political Activity Are on
the Rise, INSIDE POLITICAL LAW (Oct. 17, 2012), http://www.insidepoliticallaw.com/2012/
10/17/allegations-of-employer-coercion-of-political-activity-are-on-the-rise/; Alec MacGillis, Coal
Miner’s Donor, THE NEW REPUBLIC (Oct. 4, 2012), http://www.newsrepublic.com/
article/politics/108140/coal-miners-donor-mitt-romney-benefactor (describing pressure tactics
used by Murray Energy to get employees to attend political rallies and to make contributions).
204. MacGillis, supra note 203.
205. Id.
206. See Citizens United v. Fed. Election Comm’n, 558 U.S. 310, 343 (2010) (recognizing the free speech rights of corporations through political spending and holding invalid
campaign finance reform rules which treat corporate entities different from natural persons); see also Developments in the Law: Corporations and Society, 117 HARV. L. REV. 2272
(2004) (discussing the prominence of corporations in the political sphere that has led to an
increase in their social accountability).
207. 11 C.F.R. § 114.2(f)(2)(E)(iv) (2013).
2014]
COMPANY CONDUCT
983
contributions, and not forced attendance at rallies, pressure to sign
petitions, or the like.
Most state provisions ban discrimination based only on a refusal
to contribute rather than upon the act of making a financial contribution.208 States more commonly protect the right to vote,209 sometimes
classified as “the right of suffrage.”210 However, some courts construe
this protection to extend to other forms of political activity besides
actually casting a vote.211 This construction supports the tendency
among jurisdictions to provide protection to employees seeking to express political beliefs in a way that does not substantially interfere
with their employer’s interests.
(b) Whistleblowing
Another prime candidate for protection is “whistleblowing.” Certainly, there are no shortage of whistleblower cases, where employees
reporting misconduct obtain relief either through statutes212 or the
PPE itself.213 The case for a clear and substantial public policy here is
compelling. First, there is the impressive number of statutes at the
208. ALA. CODE §§ 10A-21-1.01(b), 17-5-17 (2012); IDAHO CODE ANN. § 67-6605 (2012);
LA. REV. STAT. ANN. § 18:1461.1(A)(2), :1483, :1505.2 (2011); MASS. GEN. LAWS ch. 56, § 33
(2012); MO. REV. STAT. §§ 130.028.1(2)-(3) (1997); N.H. REV. STAT. ANN. § 664:4-a(II)
(2012); S.C. CODE ANN. § 8-13-1332(2) (2011); TEX. ELEC. CODE ANN. § 253.102 (West
2011); WASH. REV. CODE § 42.17A.495(2)(a) (2011); WIS. STAT. § 12.07(4) (2011); WYO.
STAT. ANN. § 22-26-111(a)(ii) (2011).
209. HAW. REV. STAT. § 19-3 (2011); IDAHO CODE ANN. § 18-2305 (2012); KY. REV. STAT.
ANN. § 121.310(1) (West 2011); 25 PA. CONS. STAT. § 3547 (2012); TENN. CODE ANN. § 2-19134(b) (2012); W. VA. CODE § 3-8-11 (2012); WYO. STAT. ANN. § 22-26-111 (2011).
210. See BLACK’S LAW DICTIONARY 712 (9th ed. 2009) (defining “suffrage” as “[t]he
right or privilege of casting a vote at a public election”).
211. See, e.g., Am. Indep. Party in Idaho, Inc. v. Cenarrusa, 442 P.2d 766, 768 (Idaho
1968) (recognizing the right of citizens to gather and express political beliefs as inherent to
the term “suffrage”); Md. Green Party v. Md. Bd. of Elections, 832 A.2d 214, 228 (Md. 2003)
(indicating that the term “suffrage” includes the right to sign a referendum petition).
212. See, e.g., ALASKA STAT. § 39.90.100 (2013); ARIZ. REV. STAT. ANN. § 23-1501(3)(c)(ii)
(2013); N.D. CENT. CODE § 34-01-20(1)(a) (2012); R.I. GEN. LAWS § 28-50-3(4) (2012).
213. See, e.g., N.J. STAT. ANN. § 34:19-4 (West 2013) (allowing plaintiff to choose between suing under statute or under common law public policy claim); Porter v. Reardon
Mach. Co., 962 S.W.2d 932, 938 (Mo. Ct. App. 1998) (recognizing whistleblower exception
to doctrine of employment at will).
984
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
federal214 and state215 levels providing whistleblower protection of one
manner or another. Recognizing this as a clear and substantial public
policy is almost tautological, inasmuch as the activity is often recognized as whistleblowing only when the misconduct alleged is a matter of significant public concern or implicates significant matters of
public policy.216
While whistleblowing activities are often mentioned as the kind of
socially desirable activity that might trigger the PPE, in practice,
whistleblowing has not always sufficed. Courts often distinguish between whistleblowing to vindicate a public interest versus exposing
mere private wrongdoing. For example, the Supreme Court of Oklahoma expressly distinguished between the employee reporting of
“crimes,” “violation of health or safety laws,” or other actions breaching a public trust, that would be protected under the PPE, and the
employee reporting of conduct harming “private or proprietary interests,” that would not.217 Thus, exposing an employer’s Medicare fraud
was protected, but exposing a co-employee’s embezzlement from a
private employer was not.218 Examples of cases applying the PPE for
“whistleblowing” activities aimed at actual violations of law
214. See, e.g., Asbestos Emergency Response Act, 15 U.S.C. § 2651 (2012) (protects
employees from discrimination or retaliation for reporting alleged violations of environmental laws relating to asbestos in elementary and secondary school systems); SarbanesOxley Act, 18 U.S.C § 1514(A) (2012) (protects employees of publicly traded companies, and
their contractors, subcontractors, and agents from discrimination or discharge for reporting
violations of the federal mail, wire, bank, or securities fraud statutes, or any other provision of federal law relating to fraud against shareholders); Occupational Safety and Health
Act, 29 U.S.C § 660(c) (2012) (protects employees from discrimination or retaliation for
reporting workplace or safety hazards); 33 U.S.C. § 1367 (2006); Energy Reorganization
Act, 42 U.S.C. § 5851 (2006) (protects employees in the nuclear power and nuclear medicine industries against discrimination or discharge for reporting violations of nuclear safety laws); 42 U.S.C § 6971 (2006) (each protecting employees from discrimination or discharge for reporting violations of certain environmental laws or regulations); 42
U.S.C § 7622 (2006); International Safe Container Act, 46 U.S.C. App. § 1506 (2006) (protects employees who are discharged or discriminated against for filing an OSHA complaint
alleging unsafe containers for internal cargo); Surface Transportation Assistance Act, 49
U.S.C. § 31105 (2006) (protects employees of the trucking industry who are discharged,
disciplined, or discriminated against for reporting commercial motor vehicle safety hazards, health, or safety concerns).
215. See, e.g., FLA. STAT. § 448.102(1) (2013) (prohibiting “retaliatory personnel action”
where an employee reports employer’s unlawful conduct or refuses to engage in illegal activity); NEV. REV. STAT. § 281.631 (2013) (prohibiting the use of “authority or influence to
prevent disclosure of improper governmental action by another state or local governmental
officer or employee”); TENN. CODE ANN. § 50-1-304 (2013) (providing that “[n]o employee
shall be discharged or terminated solely for refusing to participate in, or for refusing to
remain silent about, illegal activities”); W. VA. CODE § 6C-1-1 (2013) (providing protection
to a discharged employee “where the employer's motivation for the discharge is to contravene some substantial public policy principle”).
216. See supra notes 142-145.
217. Darrow v. Integris Health, Inc., 176 P.2d 1204, 1213-15 (Okla. 2008).
218. Id. (citing Hayes v. Eateries, Inc., 905 P.2d 778, 785 (Okla. 1995)).
2014]
COMPANY CONDUCT
985
abound.219 But, the employee is less likely to be protected when the
speech concerns employment grievances or generic concerns about
quality control not involving illegality.220
The extent to which courts recognize a whistleblowing-based PPE
suit may involve whether the employee had a legal duty to report the
alleged impropriety; whether the employee followed proper internal
reporting procedures or violated the chain of command; and whether
the employee had the requisite expertise in the area in question to
recognize a legitimate need for whistleblowing.221 At least one court
has held that an employee enjoys no right to champion “product safety”
where the employee is not qualified to render an expert opinion on
product safety, or where safety assessment was not part of his duties.222
The somewhat tepid protection given to whistleblowing activities
is actually consistent with constitutional rules. In free speech cases
involving public employees who criticize their organizations, the Supreme Court has provided First Amendment protection only when (1)
the speech relates to a matter of “public concern;” (2) it is not spoken
by the employee as part of the employee’s job duties; and (3) its value
outweighs any potential damage to the government office’s reputation, efficiency, or the like.223 This is because public employers have a
legitimate interest in controlling speech that can directly affect the
performance of their agency or office.224 If the speech satisfies conditions (1) and (2) above, and the court determines that the legitimate
interests of the government employer in efficient operations do not
219. See, e.g., Sheets v. Teddy’s Frosted Foods, Inc., 427 A.2d 385 (Conn. 1980) (state
labeling and licensing laws); Palmer v. Brown, 752 P.2d 685, 689-90 (Kan. 1988) (Medicaid
fraud); Harless v. First Nat’l Bank in Fairmont, 246 S.E.2d 270 (W. Va. 1978) (state and
federal consumer credit protection laws).
220. Compare Flesner v. Technical Comm’n Corp., 575 N.E.2d 1107, 1111 (Mass. 1991)
(applying exception to termination for cooperating with a federal investigation of the employer), with Wright v. Shriners Hosp. for Crippled Children, 589 N.E.2d 1241, 1245 (Mass.
1992) (declining to apply PPE to termination for criticizing employer to national medical
accrediting body), and Hrehorovich v. Harbor Hosp. Ctr., Inc., 614 A.2d 1021, 1033 (Md.
App. 1992) (declining to apply PPE where hospital employee, claiming right to “promote
quality health care,” engaged in extensive criticisms of the hospital). See also Suchodolski
v. Mich. Consol. Gas Co., 316 N.W.2d 710 (Mich. 1982) (corporate management dispute
regarding internal accounting procedures did not amount to violation of a clearly mandated
public policy).
221. Field v. Phila. Elec. Co., 565 A.2d 1170, 1180-81 (Pa. Super. Ct. 1989)
222. Geary v. U.S. Steel Corp., 391 A.2d 174, 180 (Pa. 1974).
223. Garcetti v. Ceballos, 547 U.S. 410, 423 (2006); Connick v. Myers, 461 U.S. 138, 147
(1983) (“When employee expression cannot be fairly considered as relating to any matter of
political, social, or other concern to the community, government officials should enjoy wide
latitude in managing their offices, without intrusive oversight by the judiciary in the name
of the First Amendment.”).
224. Garcetti, 547 U.S. at 421-23.
986
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
outweigh the employee’s free speech interests, then the employer
may not engage in any adverse employment action.225
Something analogous ought to apply in the private employment
sphere under the PPE. Speech on matters not of a public concern
ought not to trigger PPE protection. And, if a job’s duties require an
employee to make presentations, write memos, etc., the employer
must fire the employee based on the content of what she says. The
tougher question arises when the speech concerns a matter of public
interest and is not made pursuant to job duties.
(c) Coverage Inspired by Statutory Speech Protection for Private
Employees
While the above principles apply equally to state government
employees,226 neither federal nor state constitutional principles
directly provide free speech protection to private employees.227 Further, most states lack any explicit statutory free speech protection for
private employees.
However, a minority of states have promoted similar free speech
principles in varying degrees within the realm of private employment.228 Over a half-dozen states generally prohibit employers from
punishing employees for engaging in political activities.229 A few prohibit discrimination based narrowly on election-related speech,230 and
a few others extend employment protection to the expression of “political” ideas, opinions, or beliefs.231
Another area of variation is whether anti-discriminatory provisions apply only to expressions occurring outside of the employer’s
225. Id.
226. See Gitlow v. New York, 268 U.S. 652, 630 (1925) (Fourteenth Amendment Due
Process Clause incorporates First Amendment right to free speech); Smith v. Novato Unified Sch. Dist., 59 Cal. Rptr. 3d 508 (Cal. Ct. App. 2007).
227. See Hudgens v. NLRB, 424 U.S. 507, 513 (1976) (stating that the Constitution
does not directly provide free speech protection to private corporations or persons); Jacobs
v. Major, 407 N.W.2d 832, 836 (Wis. 1987) (stating that the free speech clause of the state
constitution provides protection only against state action).
228. See Eugene Volokh, Private Employees’ Speech and Political Activity: Statutory
Protection Against Employer Retaliation, 16 TEX. REV. L. & POL. 295 (2012) (discussing the
ways in which state statutes provide protection for employee political speech and activity).
229. See CAL. LAB. CODE § 1101 (West 2012); COLO. REV. STAT. § 8-2-108 (2013); LA.
REV. STAT. ANN. § 23.961 (2011); MO. ANN. STAT. § 115.637(6) (West 2012); NEB. REV.
STAT. § 32-1537 (2012); NEV. REV. STAT. § 613.040 (2012); S.C. CODE ANN. § 16-17-560
(2011); W. VA. CODE § 3-8-11(b) (2012). See also SEATTLE, WASH. MUN. CODE § 14.04.040
(2011) (creating similar general free speech rights for private employees at the municipal
level); MADISON, WIS. MUN. CODE §§ 39.03(1), .03(2)(cc), .03(8)(c), .03(8)(d)(1) (2010)
(same).
230. See 5 ILL. COMP. STAT. ANN. § 5/29-4 (2012); N.Y. LAB. LAW § 201-d (McKinney
2012); WASH. REV. CODE § 42.17A.495(2) (2012).
231. See MINN. STAT. § 10A.36 (2012); N.M. STAT. ANN. § 1-20-13 (West 2012).
2014]
COMPANY CONDUCT
987
premises, or additionally, to expressions taking place at the work
site.232 Even where the statute expressly provides protection to offpremises speech, there is variation as to degree. For example, in
mandating employers to terminate employees only for “[g]ood cause,”
Montana clarifies that the scope of protection applies, but is not limited to, “[t]he legal use of a lawful product by an individual off the
employer’s premises during nonworking hours.”233 While such statutes are often applied to smoking and drinking after hours and offsite,234 they also can be applied to more inherently expressive products like political signs and social media programming.235 Colorado
and North Dakota broaden the reach of this protection by forbidding
employers to terminate employees for engaging in “any lawful activity” that occurs off the employer’s premises.236 Finally, state statutes
also vary the breadth of protection afforded by the use of the term
“political,”237 and whether protection extends to hiring practices, or to
only disciplinary actions against current employees.238
The variations above alert one to several salient issues regarding
the proper breadth of free speech protections in the private sphere.
Courts using free speech principles to inform their PPE analysis
would do well to consider such issues as, for example, the extent to
which the expression is political, off-site, or after-hours.
232. Compare Cotto v. United Techs. Corp., 711 A.2d 1180 (Conn. App. Ct. 1998)
(finding statutory language that does not explicitly limit protection to off-the-job speech
extends to on-the-premises speech through implication), with Dixon v. Coburg Dairy, Inc.,
330 F.3d 250, 262 (4th Cir. 2004) (interpreting a similar state statute to apply only to offpremises speech, for fear of making the workplace a “constitutionally protected forum for
political discourse”).
233. MONT. CODE ANN. § 39-2-903 (2011).
234. See, e.g., MINN. STAT. § 181.938 (2012) (employers may not discipline or discharge
employee based on use of lawful products off-premises during nonworking hours).
235. See, e.g., Angel v. Rayl, No. 04CV3420, 2005 WL 6208024, at *2 (D. Colo. Dec. 1,
2005) (interpreting lawful activity off the job premises to include “attend[ing] certain political or social-activism events”); Joseph Lipps, State Lifestyle Statutes and the Blogosphere:
Autonomy for Private Employees in the Internet Age, 72 OHIO ST. L.J. 645, 652-54 (advocating for a distancing between blogs and social networking activities and adverse employment action). But see Kolb v. Camilleri, No. 02-CV-0117A(Sr), 2008 WL 3049855, at *13
(W.D.N.Y. Aug. 1, 2008) (concluding that picketing did not constitute a “recreational activit[y]” for purposes of a statute protecting such activity when conducted off the job premises).
236. COLO. REV. STAT. § 24-34-402.5(1) (2013); N.D. CENT. CODE ANN. § 14-02.4-03, 408 (West 2011).
237. Compare Thompson v. Borg-Warner Protective Servs. Corp., No. C-94-4015 MHP,
1996 WL 162990, at *9 (N.D. Cal. Mar. 11, 1996) (finding the term “political activities” to
include opinions on alcohol and drug policies within the meaning of sections 1101 and 1102 of
the California Labor Code), with Vanderhoff v. John Deere Consumer Prods., Inc., No. C.A.
3:02–0685–22, 2003 WL 23691107, at *2 (D.S.C. March 13, 2003) (holding that a confederate
flag decal was not a “political opinion” as intended by S.C. CODE ANN. § 16-17-560).
238. Compare Bools v. Gen. Elec. Co., 70 F. Supp. 2d 829, 832 (S.D. Ohio 1999) (limiting statute’s protection to persons already employed), with Gay Law Students Ass’n v. Pac.
Tel. & Tel. Co., 595 P.2d 592, 610 n.16 (Cal. 1979) (interpreting a statute that protects
“employees” to include job applicants).
988
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
As noted above,239 in jurisdictions protecting private employees’
speech, the existence of such statutory claims need not prevent a
court from considering a PPE claim. And, the existence of such statutory remedies can help a court decide that a clear public policy to that
effect exists, such that a PPE claim should be recognized. But, even
where such statutory remedies do not exist, the constitutional basis
alone for free speech protections ought to be sufficient to convince
courts that a clear public policy favoring free speech rights supports a
PPE claim.
(d) Toward a Balance of Employee and Employer Interests
It is axiomatic that the right of free speech is “fundamental” in our
country, and the Supreme Court has repeatedly so held.240 It is the
very first right listed in the Bill of Rights.241 It was the first substantive (i.e., non-procedural) right to be recognized as “fundamental,”
and thus, “incorporated” through the Fourteenth Amendment’s Due
Process Clause and applied to the states.242 There is no individual
right more celebrated in the United States. A leading scholar has
termed it “the Constitution’s most majestic guarantee,” one which
helps to preserve all other rights.243 If constitutional principles can
ever inform a courts’ application of the PPE, then such constitutional
inspiration must include free speech rights.244
Certainly, the kind of free speech protection offered to conventional government whistleblowers also should apply to private employees
who blow the whistle on illegal acts. But should the PPE extend beyond such a straightforward type of employer harassment, to adverse
employment action taken against an employee for engaging in other
types of speech? If one is committed to a vibrant, free speech-oriented
society, the answer should be yes.
239. See supra Part III.A.
240. See, e.g., McDonald v. City of Chicago, 130 S. Ct. 3020, 3061 (2010); Near v. Minnesota, 283 U.S. 697, 707 (1931); Gitlow v. New York, 268 U.S. 652, 666 (1925).
241. U.S. CONST. amend. I.
242. See Gitlow, 268 U.S. at 666.
243. Nelson v. McClatchy Newspapers, Inc., 936 P.2d 1123, 1129 (Wash. 1997) (quoting
LAWRENCE H. TRIBE, AMERICAN CONSTITUTIONAL LAW § 12-1 (2d ed. 1988)).
244. Cf. San Diego Bldg. Trades Council v. Garmon, 359 U.S. 236, 244-45 (1959) (National Labor Relations Act’s protection of union employees’ picketing rights preempted
contrary state law principles). The existence of strong NLRA preemption of state law
might suggest that much of the Free Speech-based PPE discussion in this Section was unnecessary: perhaps federal labor law statutes preempted the analysis entirely. But, analogizing from Supreme Court First Amendment case law, this Article suggests that the PPE
should not apply to job-related speech. This approach would thus apply the PPE only in
situations in which federal labor law did not apply, and could thus raise no legitimate issue
of federal preemption.
2014]
COMPANY CONDUCT
989
First, the case for protecting employees from “forced speech” is
quite compelling. Even in relatively trouble-free economic times, employees are simply not that portable. Being fired from one’s job and
being forced to find new employment is a traumatic experience for
many, especially those who live paycheck to paycheck. Employers,
thus, have a potent coercive influence over their employees.245 Allowing them to abuse that power to influence the political process would
skew our political debate and undermine our democratic process.
Even beyond forced speech situations, there is a strong public policy interest in granting private employees a certain amount of free
speech latitude. We protect free speech not merely as a means of
promoting discussion and participation in democratic government,
and not merely to further the discovery of truth through “the marketplace of ideas,” but also because individual self-expression is good
for its own sake.246 It leads to happier, more fulfilled lives: a better
quality of life for the individuals doing the expressing, and because
these individuals are more fulfilled, a more pleasant environment for
the friends and coworkers around them.
The vast majority of Americans are employees who spend about
one-third of their lives, and almost half their waking lives, at their
place of work. It is a prime place for them to communicate with
their peers, and for them to receive information from their peers.
For workers to be denied a certain range of motion in their expression could seriously dampen this individual self-fulfillment, not to
mention sabotage the search for truth and diminish the vitality of
our democracy.
Again, it is no answer to suggest that anyone who does not like the
free speech restrictions of their employer can find other employment.
For most of these people, the prospect of losing their jobs is a significant hardship, and has a formidable chilling effect on their speech.
Current law recognizes certain limitations on the PPE doctrine to
recognize the legitimate interests of employers. Again, Novosel
245. A similar analysis informs federal labor law. In NLRB v. Gissel Packing Co., 395
U.S. 575 (1969), the Supreme Court explained that the “economic dependence” of employees required that employers be very careful in expressing anti-union sentiments prior to
union elections, lest employees be unduly influenced by coercive “intended implications” of
such efforts. Id. at 617. And in NLRB v. Exch. Parts Co., 375 U.S. 405 (1964), the Court
held that the NLRA kept employers from granting benefits on the eve of a union election,
even if they were unconditional and permanent. The potential coercive influence of such
employer action was so great, the Court reasoned, that even the implicit threat of such a
move—i.e., the suggestion that future benefits could dry up if employees did not vote the
way the employer wanted—could constitute a “fist inside the velvet glove” and thus violate
the NLRA. Id. at 409-10.
246. See Curtis Publ’g Co. v. Butts, 388 U.S. 130, 147 (1967) (free speech protection advances “truth, science, morality, and arts in general as well as responsible government” (quotation marks omitted)); 25 JOHN STUART MILL, ESSAY ON LIBERTY 246-49 (Charles W. Eliot
ed., 1909) (1859) (famously expounding the “utilitarian” rationale for free speech protections).
990
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
provides a good example. Even as it drew upon constitutional principles to extend free speech protection to private employers via the
PPE, the court stated guidelines to protect the employers’ prerogatives.247 It called for other courts considering such claims to weigh,
inter alia, the employer’s contractual interests; whether the time,
place, and manner of the speech interferes with business operations,
or with “essential and close working relationships;” and whether it
prevents the employer or employee from efficiently carrying out the
requisite responsibilities.248
Although such employer interests do not typically qualify for the
bona fide occupational qualification exception, state legislatures have
included provisions within their statutes to constrain broad employee
protection in order to protect employer interests.249 As an illustration,
courts have held that a newspaper could prevent its reporters from
engaging in political activity when such activity frustrates the employer’s purpose by creating an apparent conflict of interest,250 or
from engaging in speech when that speech was placing an advertisement in a rival newspaper.251 Similarly, a New York court found a
discharge lawful where an employee’s affiliation with Holocaust deniers threatened to harm the reputation of the employer.252
Even where the speech involved does not present a conflict of interest or bring the company into public disrepute generally, it may be
unprotected if it is a publicly expressed opinion at odds with that of
the employer’s and potentially injurious to the employer’s financial
interests. For example, a Massachusetts court found it legitimate for
a defense contractor to fire its spokesperson for publicly advocating a
decline in defense spending, a topic of “acute concern” to the company
in which it had a direct “financial stake.”253 In general, then, courts
247. Novosel v. Nationwide Ins. Co., 721 F.2d 894, 901 (3d Cir. 1983).
248. Id.
249. See MONT. CODE ANN. § 39-2-903(5) (2013) (establishing “good cause” for dismissal
where employees cause “disruption of the employer's operation, or other legitimate business
reason”); N.D. CENT. CODE § 14-02.4-03 (2011) (limiting employee protection to activities that
are “not in direct conflict with the essential business-related interests of the employer”).
250. See Nelson v. McClatchy Newspapers, Inc., 936 P.2d 1123, 1127 (Wash. 1997)
(determining that a state law prohibiting employment discrimination based on an employee's political conduct was not constitutionally applicable to newspaper publishers).
251. See Martin v. Capital Cities Media, Inc., 511 A.2d 830, 844 (Pa. Super. Ct. 1986).
252. See Berg v. German Nat’l Tourist Office, 248 A.D.2d 297, 298 (N.Y. App. Div.
1998) (affirming the lower court’s dismissal of the wrongful discharge suit, without referencing the initial grounds for the discharge); Salvatore Arena, Tourist Office KOs Damage
Suit, N.Y. DAILY NEWS (Feb. 12, 1997, 12:00 AM), http://www.nydailynews.com
/archives/news/tourist-office-kos-damage-suit-article-1.760157 (explaining that the employee was fired for Holocaust denial associations, and that the lower court upheld the termination because Holocaust denial conflicts with the business interests of the German National Tourist Office).
253. Korb v. Raytheon Corp., 410 N.E.2d 581, 584-85 (Mass. 1991).
2014]
COMPANY CONDUCT
991
may reject free speech-based PPE claims where the employee speaks
“against the interests” of the employer.254 There is no protection
for speech which actively disrupts the employer or provides assistance to competitors.
A middle-ground approach is needed, one which balances the public interest in protecting employees’ rights of expression with those of
the employer. Existing PPE case law already points the way. For example, protected whistleblowing activity is more likely to be protected when it involves actual illegality, as opposed to mere criticism of a
firm’s products or procedures. Employees should not have the right to
trash their employers publicly or to engage in conduct that diminishes their reputation, but they certainly should have the right to point
out illegal employer conduct. At the same time, even where the Supreme Court recognizes a legitimate business prerogative on the part
of an employer, it has shown its willingness to withdraw such deference when the facts show that the employer is using its business rationale as a pretext for more ordinary unfair labor practices.255
Drawing upon existing statutory and case law protections for
guidance, we recommend some basic guidelines for PPE claims based
on free speech principles. First, an employer should not be able to
discharge an employee for expression protected by the First Amendment taking place while the employee is off-duty and off-premises,
unless such activity is against the employer’s interest. Expression
can run counter to the employer’s interest if it (1) aids a competitor;
(2) exposes confidential information; (3) runs counter to the avowed
mission of the firm or organization; (4) harms the reputation of the
employer; or (5) materially and substantially disrupts productivity
(e.g., by harming morale). The latter two exceptions would not apply
if the employee speech was not merely criticism but actual whistleblowing, i.e., accurate factual assertions exposing illegal or unethical
practices, or practices threatening the public health or welfare. Further, the above exceptions would not apply where the facts showed
that the employer’s proffered business justification was pre-textual.
254. Id. at 284. See also Rossi v. Pa. St. Univ., 489 A.2d 828 (Pa. 1985) (employee who
lost PPE claim had continuously complained to supervisors about financial waste in his
department, thus becoming “troublesome”). See, e.g., COLO. REV. STAT. § 24-34-402.5(1)(b)
(2012) (prohibiting employers from firing employees based on their lawful conduct off the
employment premises unless such conduct conflicts with their responsibilities to the employer); N.Y. LAB. LAW § 201-d(3)(a) (2011) (making it unlawful for an employer to discriminate based on employees’ off-duty political activities unless such activities interfere with
the employer’s interest).
255. See Assoc’d Press v. NLRB, 301 U.S. 103, 131 (1937) (NLRA restrictions on firing
for attempting to organize a workforce did not violate media employer’s free press rights
where the employer acted not out of editorial prerogative but out of a desire to suppress
union membership).
992
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
Tying the PPE to expression protected by the First Amendment
means an employer can fire an employee for off-premises, off-duty
speech that, e.g., is obscene or defamatory, or constitutes “fighting
words.” Speech that is highly offensive but constitutionally protected—
e.g., racist or anti-Semitic slurs, or sexually “indecent” but non-obscene
speech—could be grounds for termination to the extent it was open,
public, and notorious, and could plausibly be connected back to the
employer by the public, thus harming the employer’s reputation.
For speech made on-duty, on-premises, or using the employer’s
facilities, resources, or materials, the employer should have wide latitude to discipline or terminate an employee. Courts should intervene
only where the employer is using its power to pressure employees to
make political contributions, vote a certain way, or express a certain
political opinion.
2. Fourth Amendment Privacy
The personal privacy rights256 guaranteed by the Fourth Amendment’s protection against unreasonable searches and seizures are
also excellent candidates for constitutional rights driving a PPE
analysis. These rights are well known to citizens at large, as the recent controversy over domestic surveillance by the National Security
Agency demonstrates.257 The Fourth Amendment’s privacy protections apply to the states through incorporation, of course.258 Moreover, all state constitutions have some comparable provision.259 And,
256. The privacy concerns protected by the Fourth Amendment are analytically distinct
from the “right to privacy” held to apply under the Due Process Clause as part of substantive
due process. The latter protects individuals from unwarranted state interference in deeply
personal matters such as childrearing, family, reproduction, and intimate relations. See infra
Part V.C.1-2. The Fourth Amendment search and seizure brand of privacy is sometimes referred to as “informational privacy.” See Planned Parenthood of S. Ariz. v. Lawall, 307 F.3d
783, 789 (9th Cir. 2002) (defining “informational privacy” as encompassing “privacy interest
in avoiding disclosure of sensitive personal information”); Francis S. Chlapowski, Note, The
Constitutional Protection of Informational Privacy, 71 B.U. L. REV. 133, 135 (1991) (asserting
that an individual’s “interest in informational privacy is a right that the Constitution protects
under the due process clauses' protection of both liberty and property”).
257. See David Aronofksy & Matthew Cooper, The War on Terror and International
Human Rights: Does Europe Get it Right?, 37 DENV. J. INT'L L. & POL'Y 567, 576 (2009)
(describing the NSA’s surveillance program as having deleterious effect on privacy rights
by “reaching into the homes and offices of Americans all throughout the United States
and elsewhere”); Joshua Kopstein, Senators Strike at Surveillance Laws Authorizing
NSA’s Warrantless Data Collection, THE VERGE (June 24, 2013, 6:25 PM),
http://www.theverge.com/2013/6/24/4460334/senators-strike-at-nsa-surveillance-laws.
258. See Mapp v. Ohio, 367 U.S. 643, 655 (1961).
259. ALA. CONST. art. I § 5; ALASKA CONST. art. I § 14; ARIZ. CONST. art. II § 8; ARK.
CONST. art. II § 15; CAL. CONST. art. I § 13; COLO. CONST. art. II § 7; CONN. CONST. art.
I § 7; DEL. CONST. art. I § 6; FLA. CONST. art. I § 12; GA. CONST. art. I, § 1, § XIII; HAW.
CONST. art. I § 7; IDAHO CONST. art. I § 17; Ill. CONST. art. I § 6; IND. CONST. art. I § 11;
IOWA CONST. art. I § 8; KAN. CONST. BILL OF RIGHTS § 15; KY. CONST. § 10; LA. CONST. art.
2014]
COMPANY CONDUCT
993
all states are subject to certain federal statutes protecting privacy
interests of even private employees. For example, Congress has restricted the use of polygraph tests for pre-employment screening,260
inquiries about employee disabilities,261 and the use of genetic information in employment decisionmaking.262
At least one state constitution includes a right of privacy that applies to private action.263 Further, as with free speech rights, many
states have passed statutes guaranteeing private employee privacy,
at least in specific contexts, such as drug testing.264 This privacy interest is perhaps even stronger than free speech as a PPE candidate,
given that the common law recognizes a variety of privacy-based
causes of action.265
The strength of this interest is intuitive. We all have certain sensitive information concerning our lives away from the office—off-duty
matters which have no effect on one’s work productivity. It should
not be controversial that society has an interest in protecting the
ability of workers to keep that information private. As the Third Circuit put it:
It may be granted that there are areas of an employee’s life in
which [the] employer has no legitimate interest. An intrusion into
I § 5; ME. CONST. art. I § 5; MD. CONST., DECLARATION OF RIGHTS, art. XXVII; MASS.
CONST. PT. 1, art. XIV; MICH. CONST. art. I § 11; MINN. CONST. art. I § 10; MISS. CONST. art.
III § 23; MO. CONST. art. I § 15; MONT. CONST. art. II § 11; NEB. CONST. art. 1, § 7; NEV.
CONST. art. I § 18; N.H. CONST. PT. 1, art. XIX; N.J. CONST. art. I § 7; N.M. CONST. art.
II § 10; N.Y. CONST. art. I § 12; N.C. CONST. art. I § 20; N.D. CONST. art. I § 8; OHIO CONST.
art. I § 14; OKLA. CONST. art. II § 30; OR. CONST. art. I § 9; PA. CONST. art. I § 8; R.I. CONST.
art. I § 6; S.C. CONST. art. I § 10; S.D. CONST. art. VI § 11; TENN. CONST. art. I § 7; TEX.
CONST. art. I § 9; UTAH CONST. art. I § 14; VT. CONST. CH. 1, art. I § 11; VA. CONST. art.
I § 10; WASH. CONST. art. I § 7; W. VA. CONST. art. III § 6; WIS. CONST. art. I § 11; WYO.
CONST. art. I § 4.
260. Employee Polygraph Protection Act of 1988, 29 U.S.C. §§ 2001–2009 (2012).
261. Americans with Disabilities Act, 42 U.S.C. § 12112 (2012).
262. Genetic Information Nondiscrimination Act of 2008, 42 U.S.C. § 2000ff-1
(Supp. II 2008).
263. See, e.g., CAL. CONST. art. I § 1; Luck v. S. Pac. Transp. Co., 267 Cal. Rptr. 618,
627-28 (Cal. Ct. App. 1990) (forcing employee to offer urine sample violated her right of
privacy as articulated in state constitution).
264. See, e.g., CAL. CIV. CODE § 56.20 (West 2013) (providing standards of confidentiality regarding employee medical records); CONN. GEN. STAT. §§ 31-51t to -51aa (2012) (regulating an employer’s right to enforce drug testing of current employees in the private sector); HAW. REV. STAT. § 378-26.5(5) (2013) (making it unlawful for an employer “to [u]tilize
any device that intrudes into any part or cavity of the body for the purpose of truth verification”); MASS. GEN. LAWS ch. 214, § 1B (2013) (protecting an individual from “unreasonable, substantial or serious interference with his privacy”); see generally Wefing, supra note
23 (comparing drug testing policies of public employers with that of private employers).
265. See RESTATEMENT (SECOND) OF TORTS § 652A-E (1977) (listing the types of torts
grounded in the right of privacy). See e.g., GA. CODE. ANN. § 16-11-62(1) (2013) (prohibiting
by law any form of eavesdropping or surveillance performed in a “clandestine manner”);
West v. Media Gen. Convergence, Inc., 53 S.W.3d 640 (Tenn. 2001) (recognizing the tort of
false light invasion of privacy).
994
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
one of these areas by virtue of the employer’s power of discharge
might plausibly give rise to a cause of action, particularly where
some recognized facet of public policy is threatened.266
Some courts have recognized a privacy-based PPE cause of action.
In some cases, courts have merely drawn on statutory sources,267 but
sometimes they have drawn, at least in part, on constitutional interests.268 These courts have often engaged in a balancing-of-interests
approach, weighing the privacy interests of the employee with the
legitimate business interests of the employer.269 In the context of
drug testing, for example, courts have been willing to recognize a
wrongful discharge claim where a mandatory drug-testing policy
tested for drug use outside the workplace and work hours, was not
based on individualized suspicion, and did not involve sensitive safety-related positions.270 These factors are also central to the Supreme
Court’s Fourth Amendment analysis of government-imposed drug
testing programs on public employees.271 Similarly, some courts have
recognized a privacy-related PPE cause of action where plaintiffs
266. Borse v. Piece Goods Shop, Inc., 963 F.2d 611, 622 (3d Cir. 1992) (quoting Geary v.
U.S. Steel Corp., 456 A.2d 174, 180 (Pa. 1974)). The Third Circuit actually had rejected use
of the Fourth Amendment as a source for public policy, but acknowledged that the privacy
concerns under state tort law and the Constitution were similar. See id. at 621 n.10.
267. See, e.g., Doyon v. Home Depot USA, Inc., 850 F. Supp. 125 (D. Conn. 1994)
(private employer's post-accident drug-testing policy violated statutory provision prohibiting employers from requiring an employee to submit to urinalysis testing without
reasonable suspicion).
268. See, e.g., Luedtke v. Nabors Alaska Drilling, Inc., 768 P.2d 1123, 1130 (Alaska
1989); Hennessey v. Coastal Eagle Point Oil Co., 609 A.2d 11, 16-17 (N.J. 1992); Twigg v.
Hercules Corp., 406 S.E.2d 52, 55 (W. Va. 1990).
269. Twigg, 406 S.E.2d at 55-56; but see Jennings v. Minco Tech. Labs, Inc., 765 S.W.2d
497, 502 (Tex. App. 1989) (rejecting a balancing approach, and holding drug testing within
an employer’s prerogative). The analysis is different, and more employee-favorable, in California, where the state constitution’s privacy right actually applies directly to private employers. See Kraslawsky v. Upper Deck Co., 65 Cal. Rptr. 2d 297 (1997) (state constitutional guarantee of the right to pursue and obtain privacy creates a right of action against private as well as government entities). Some cases have also recognized a PPE cause of action against employee drug testing based on non-constitutional statutory sources. See, e.g.,
Doyon, 850 F. Supp. at 125 (private employer's post-accident drug testing policy violated
statutory provision prohibiting employers from requiring an employee to submit to urinalysis testing without reasonable suspicion).
270. Compare Luedtke, 768 P.2d at 1123 (rejecting PPE claim where plaintiffs performed
safety-sensitive jobs), with Twigg, 406 S.E.2d at 55 (accepting PPE claim because the positions did not involve safety and the employer lacked “reasonable good faith objective suspicion”). Cf. Borse, 963 F.2d at 625-26 (these factors were relevant and probative but their absence was not necessarily fatal to a private employer’s mandatory drug testing policy).
271. See Skinner v. Ry. Labor Execs.’ Ass’n, 489 U.S. 602, 603, 624 (1989) (justifying
employee drug tests by asserting the government’s interest in regulating “safety-sensitive
tasks” which outweighs the “requirement of individualized suspicion”).
2014]
COMPANY CONDUCT
995
challenged employer policies mandating polygraph testing272 and personal property searches.273
Not all PPE claims involving employee drug testing have been
successful. Sometimes, the court explicitly rejects a constitutional
privacy right as a basis for a PPE claim.274 In others, the court examines only statutory sources of public policy before rejecting the
employee’s claim.275
As a general matter, the privacy right would normally not extend
to information reasonably related to activity at the workplace for
which the employer would have a legitimate business interest, especially where the investigative methods were not overly intrusive. For
example, a federal district court in Pennsylvania rejected a privacybased PPE claim brought by an employee fired after an allegedly intrusive investigation of a personal matter, an alleged affair with a coworker.276 Noting that the employer confined its investigation to interviewing other employees and examining company records, the
court found that the employer’s methods were reasonable and rejected the wrongful discharge claim.277
Furthermore, although it declined to use constitutional principles
as evidence of public policy, the Third Circuit has analyzed privacy
issues (based on common law privacy torts) as affected by employer
drug-testing policies.278 The court found an employer’s policy requiring employees to provide urine samples for drug screenings to be a
significant intrusion of privacy.279 This was both because of the extremely intrusive manner in which a urine sample is collected, and
because urine samples can yield “a host of private medical facts about
an employee,” together creating a very strong expectation of privacy.280
272. See, e.g., Leibowitz v. H.A. Winston Co., 493 A.2d 111, 115 (Pa. Super. Ct. 1985).
273. See, e.g., Borse, 963 F.2d at 625.
274. See, e.g., Wilkinson v. Times Mirror Corp., 264 Cal. Rptr. 194 (Cal. Dist. Ct. App.
1989) (private employer's offer of employment conditional on submission to the drug test
not a violation of the right to privacy under the state constitution); Hennessey v. Coastal
Eagle Point Oil Co., 589 A.2d 170, 176 (N.J. Super. Ct. App. Div. 1991), aff’d, 609 A.2d 11
(N.J. 1992) (declining to apply federal and state constitutional privacy protections to private
employers); Stein v. Davidson Hotel Co., 945 S.W.2d 714, 718 (Tenn. 1997) (no constitutionbased public policy protecting employees from random drug testing-based terminations).
275. See, e.g., Webster v. Motorola, Inc., 637 N.E.2d 203 (Mass. 1994) (holding private
employer's universal drug testing program did not violate employee driver's statutory privacy rights); Folmsbee v. Tech Tool Grinding & Supply, Inc., 630 N.E.2d 586 (Mass. 1994)
(holding private employer's universal drug testing program did not violate employee driver's statutory privacy rights).
276. Rogers v. Int’l Bus. Mach. Corp., 500 F. Supp. 867, 868 (W.D. Pa. 1980).
277. Id. at 870.
278. Borse, 963 F.2d 611.
279. Id. at 620-21.
280. Id. at 619-21.
996
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
Drawing again on statutory and case law protection for guidance,
and analogizing to the free speech rules set out above, we propose
Fourth Amendment-inspired contours to the PPE claim. Where an
employee enjoys a reasonable expectation of privacy regarding particular information,281 an employer should not be able discharge an
employee for failure to disclose such information unless it could reasonably affect productivity, or materially and substantially affect the
work environment. For example, for drug testing, intrusive employee
questionnaire inquiries, or searches of personal areas owned by the
employee like the employee’s car or purse, the employer must have
some form of individualized suspicion that the “search” will uncover
information relevant to actual employer interest. In the case of drug
testing, individualized suspicion would not be required if the employment positions involved were safety-related. In like manner, the
level of individualized suspicion required in any manner might decline for particularly sensitive positions, e.g., inquiries into an employee’s financial affairs for positions entrusting employees with
large sums of money.
3. Equal Protection
The Fourteenth Amendment’s guarantee of “equal protection” is
another substantive right that quite logically translates to the private work sphere. The salience of employment discrimination issues
is obvious, as demonstrated by the plethora of federal and state statutes providing protection against various types of employment discrimination. There is no serious question that all states have a strong
public policy against invidious discrimination in employment, at least
with respect to such commonly protected categories as race, color,
creed, national origin, and sex.282 There is a widespread societal consensus that such types of discrimination are invidious.283 While disapproval of these other types of discrimination may not be as
281. This “reasonable expectation of privacy” is a well-recognized concept in Fourth
Amendment jurisprudence. Katz v. United States, 389 U.S. 347, 360 (1967) (Harlan, J.,
concurring). Courts applying the PPE as suggested above would be able to draw upon a
rich body of analogous case law to help make the initial determination of whether an employee had such an expectation of privacy.
282. See, e.g., 42 U.S.C. § 2000e-2 (2006) (declaring the discharge or refusal to hire on
the basis of “race, color, religion, sex, or national origin” an “unlawful employment practice”); LA. CONST. art. I, § 12 (all prohibiting discrimination on the basis of race, religion, or
nationality); MICH. COMP. LAWS § 37.2202 (2013); OHIO REV. CODE ANN. § 4112.02 (2013).
283. E.g., Frank Newport, Little “Obama Effect” on Views About Race Relations,
GALLUP (Oct. 29, 2009), http.//www.gallup.com/poll/123944/Little-Obama-Effect-on-RaceRelations.aspx (addressing views on equal job opportunities, or the lack thereof, among
non-whites); Regina Corso, Three in Five Americans Say U.S. Has Long Way to Go to Reach
Gender Equality, THE HARRIS POLL (Aug. 16, 2010), http://www.harrisinteractive.com/News
Room/HarrisPolls/tabid/447/ct/ReadCustomDefault/mid/1508/ArticleId/452/Default.aspx (addressing views on gender equality in the workplace).
2014]
COMPANY CONDUCT
997
longstanding, both age discrimination and disability discrimination
should probably also be counted as categories for which there is a
clear policy of nondiscrimination.284
Moreover, the rights of women are implicated whenever employers
engage in sexual harassment. Although such cases are normally resolved by reference to federal or state antidiscrimination statutes,
procedural issues or the absence of statutory protection due to the
small size of an employer may leave a common law claim like the
PPE as the only available course.285 And, courts have recognized PPE
claims in the context of sexual harassment.286 There also would seem
to be no reason to exclude racial or religious harassment from coverage under the PPE. For commonly recognized workplace harassment
claims, the PPE seems appropriate where necessary to vindicate constitutional principles regarding sex discrimination.
Indeed, in this general area of employment discrimination, statutory remedies may be so abundant that they may very well “crowd
out” much of the use of the PPE. In Georgia, an appellate court held
that the absence of state statutory protection for pregnancy discrimination ruled out recognition of a wrongful discharge cause of action
on that ground.287 However, in Vermont, the state’s highest court
ruled that the absence of state statutory protection against age discrimination did not prevent recognition of a common law PPE action.288 In addition, Washington’s Supreme Court recognized a wrongful discharge claim based on sex discrimination where the employer
had too few employees to fall within the state’s otherwise applicable
antidiscrimination statute.289 Even though the state had recognized a
public policy of protecting small businesses by limiting statutory
remedies against them, the state Supreme Court allowed a sexdiscrimination-based PPE claim to proceed against a small business.290 The court noted salient differences between the statutory and
common law causes of action: only the former provided for adminis284. See, e.g., Age Discrimination in Employment Act, 29 U.S.C. § 631(a) (2012) (forbidding discrimination against anyone at least forty years of age); Americans with Disabilities Act, 42 U.S.C. § 12112 (2012) (stating that a “covered entity shall not discriminate
against a qualified individual on the basis of disability”); GA. CODE ANN. § 34-6A-4 (2013);
OR. REV. STAT. § 659A.112 (2013) (all generally forbidding employer discrimination based
on certain disabilities); VA. CODE ANN. § 51.5-41 (2013); Payne v. Rozendaal, 520 A.2d 586,
589 (Vt. 1986) (“[D]ischarge . . . solely on the basis of age is a practice so contrary to our
society’s concern for providing equity and justice that there is a clear and compelling public
policy against it.”).
285. See, e.g., Collins v. Rizkana, 652 N.E.2d 653, 660-61 (Ohio 1995).
286. Id. at 660-62.
287. Borden v. Johnson, 395 S.E.2d 628, 629-30 (Ga. Ct. App. 1990).
288. Payne, 520 A.2d at 593-94.
289. Roberts v. Dudley, 993 P.2d 901, 909-11 (Wash. 2000).
290. Id. at 908; Griffin v. Eller, 922 P.2d 788, 792 (Wash. 1996).
998
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
trative proceedings, attorney fee remedies, and potential liability for
adverse employment action other than full termination.291
Certainly, the availability of a PPE cause of action under common
law could help a wronged employee where a state or federal statute
covers the same type of discrimination, i.e., the same protected class,
but the common law claim provides better remedies, has a more forgiving limitations period, or allows suit against small employers with
too few employees to be statutorily covered.292 Moreover, the availability of the PPE also might be desirable if the protected class in
question is not covered by statute. For example, suppose an employer
fires a worker for living with a romantic partner without being married, or for having children out of wedlock. Or, suppose the employer
discriminates against the employee because the employee had been
born out of wedlock, or because the employee proceeded to marry
someone the employer disapproved of against the employer’s advice.
In some jurisdictions, this might be covered by statutes barring discrimination on the basis of marital status.293 However, in most states,
no statutory relief would be available.294
Such are prime examples of the importance of using constitutional
principles as evidence of public policy. The Supreme Court has long
held that government discrimination against children of unmarried
parents violates the Equal Protection Clause,295 and has subjected
such types of discrimination to heightened, so-called “intermediate,”
scrutiny.296 It also has held that a combination of sex plus marital
discrimination violates the Equal Protection Clause.297 Further, the
Court has held that marriage is a “fundamental right.”298
It might very well be the case that, despite the absence of any
substantive statutory protection for that type of discrimination, an
employee fired under any of the circumstances described above could
291. Id. at 911.
292. See supra Part IV.
293. See, e.g., MASS. GEN. LAWS ch. 151B, § 4 (2013); MINN. STAT. § 363A.08 (2013);
OR. REV. STAT. § 659A.030 (2013) (all prohibiting employment discrimination based on
marital status).
294. Federal antidiscrimination law does not provide employment discrimination protection on the basis of marital status. See 42 U.S.C. § 2000e-2 (2012) (Title VII bars discrimination on the basis of race, color, religion, sex, or national origin).
295. See, e.g., N.J. Welfare Rights Org. v. Cahill, 411 U.S. 619, 620-21 (1973); Weber v.
Aetna Cas. & Sur. Co., 406 U.S. 164, 175-76 (1972).
296. United States v. Virginia, 518 U.S. 515, 570-71 (1996) (analyzing “intermediate
scrutiny” as an inquiry into whether a statutory classification is “substantially related to
an important governmental objective”).
297. Weinberger v. Wiesenfeld, 420 U.S. 636, 652-53 (1975) (holding that a statute that
treats widower less favorably than widows is an invalid “sex plus marital status” type
of discrimination).
298. Zablocki v. Redhail, 434 U.S. 374, 381 (1978); Loving v. Virginia, 388 U.S. 1,
11 (1967).
2014]
COMPANY CONDUCT
999
make out a PPE claim, using constitutional principles as a basis.299
Under such circumstances, the constitutionally-inflected common law
claim would actually expand the scope of protected categories from
those listed in the applicable antidiscrimination statutes. Rather
than deviating from analogous statutes in such procedural ways as
statute of limitations or types of remedies available, this use of the
PPE would depart from an analogous statute in substance.
If successful, such a use of the PPE might have a dramatic effect
on the gay rights debate. Many states, like the federal government,
lack any statutory protection for employees discriminated against
because of their sexual orientation. However, governmental discrimination against homosexuals has come under fire from the Supreme
Court in recent decades, and this past year’s decisions have only increased the pressure.300 Such decisions, plus the large number of lower court decisions striking down anti-gay discrimination,301 may establish a strong public policy against discrimination on the basis of
sexual orientation, such that discharging a private employee would
create wrongful discharge liability.
4. Free Exercise
The Free Exercise Clause prevents the government from interfering in the worship practices of individuals.302 At a minimum, it cre299. However, at least one court rejected a PPE claim alleging termination because the
employee had married a coworker, rejecting the idea that an employee had a right to marry
the person of their choice free of employer action. McCluskey v. Clark Oil & Ref. Corp., 498
N.E.2d 559, 561 (Ill. App. Ct. 1986).
300. See United States v. Windsor, 133 S. Ct. 2675, 2693 (2013) (invalidating on equal
protection and federalism grounds a provision of a federal statute denying federal benefits
to same-sex couples legally married in their home state); Lawrence v. Texas, 539 U.S.
558, 578 (2003) (invalidating on substantive due process grounds a statute criminalizing
same-sex sodomy); Romer v. Evans, 517 U.S. 620, 635 (1996) (invalidating on equal protection grounds a state constitutional amendment which barred local jurisdictions from passing gay rights ordinances).
301. See, e.g., MacDonald v. Moose, 710 F.3d 154, 166 (4th Cir. 2013) (concluding that
statute prohibiting sodomy between two persons is facially unconstitutional); Finstuen v.
Crutcher, 496 F.3d 1139, 1156 (10th Cir. 2007) (declaring Oklahoma's adoption statute unconstitutional under the Full Faith and Credit Clause because it denied recognition to valid
out-of-state adoptions by same-sex couples); Perry v. Schwarzenegger, 704 F. Supp. 2d 921,
1003 (N.D. Cal. 2010) (state’s same-sex marriage ban violated both federal equal protection
and due process); Jegley v. Picado, 80 S.W.3d 332, 353 (Ark. 2002) (determining that a statute
prohibiting certain sexual conduct between members of same sex violated the fundamental
right to privacy and equal protection clause of the state constitution); Goodridge v. Dep’t of
Pub. Health, 798 N.E.2d 941, 969 (Mass. 2003) (granting same-sex couples the protections,
benefits, and obligations of civil marriage on equal protection grounds).
302. Thomas Berg, Free Exercise of Religion, in THE HERITAGE GUIDE TO THE
CONSTITUTION 309-11 (Edwin Meese III et al. ed., 2005) (summarizing the Free Exercise
Clause as protecting a person’s religiously motivated conduct and belief from government
action); see also Emp’t Div., Dep’t of Human Res. of Or. v. Smith, 494 U.S. 872, 877 (1990)
(defining the free exercise of religion as “the right to believe and profess whatever religious
doctrine one desires”).
1000
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
ates a clear and substantial public policy against religious discrimination, one which is mirrored in federal and state statutes barring
religious discrimination in employment.303 If use of the Free Exercise
Clause in the PPE did no more than this, it would not contribute
much, since antidiscrimination statutes of this type are so widespread. However, as with the Equal Protection Clause discussed
above, it could provide the basis for a PPE claim in those instances
where the comparable statute(s) do not apply, e.g., where the employer is too small for statutory coverage, or where the plaintiff seeks
a different remedy from the one provided by the statute.
Beyond that, its greatest impact might be to protect employees
who are fired for refusing to engage in worship practices. If an employer holds weekly prayer services during working hours, can it require employee attendance on pain of termination? A Free Exercise
Clause-inspired PPE would suggest not. 304
Such a PPE cause of action would have to give way if the employer
were itself a religious organization. Antidiscrimination statutes often
create exemptions for religious organizations, which are allowed to
favor coreligionists, or even exclude persons outside the faith, when
hiring for certain positions. These exemptions are motivated, in significant part, by a desire to avoid interfering in the Free Exercise interests of the religious organization employers.305
Another scenario would cover an employee dismissed for engaging
in non-disruptive worship practices, either on or off premises. For
example, consider an employer who employs many Muslims, and
treats them the same as non-Muslims, but does not permit them to
pray on prayer mats during the day, even if they are on break; or, an
employer who has many Jewish employees, but does not permit them
to wear yarmulkes because he bars all employees from wearing
headgear of any kind. They may not be guilty of religious discrimination, but they may be interfering with the employees’ rights to worship as they see fit.
Of course, this is one scenario where antidiscrimination statutes
may provide more protection than the PPE. Title VII, and many comparable state statutes, require that employers provide a “reasonabl[e]
303. For examples of state statutes prohibiting employment discrimination based on
religion, see NEB. REV. STAT. § 23-2513 (2012); OR. REV. STAT. § 659A.003 (2013); TEX. LAB.
CODE ANN. § 21.125 (West 2013); UTAH CODE ANN. 1953 § 34A-5-106 (West 2013).
304. Of course, such a situation might already be covered under a “forced speech” theory. See supra Part V.A.1.
305. Hosanna-Tabor Evangelical Lutheran Church & Sch. v. EEOC, 132 S. Ct. 694, 710
(2012) (holding that the Free Exercise Clause required the “ministerial exception” to bar
teacher from suing her school under workplace discrimination laws).
2014]
COMPANY CONDUCT
1001
accommodat[ion]” of employee religious practices.306 Thus, the employer might be required to take affirmative steps to make exemptions for a particular employee so as to avoid a substantial burden on
that employee’s religious practices. Thus, if there were an otherwise
neutral dress code which did not permit men to wear head garments
indoors, an employer might be required to exempt men wearing turbans or yarmulkes for religious reasons.307
This is not the case under the Free Exercise Clause, where the
rule is that the government has no obligation to modify neutral laws
of general application, even if their effect is to place a substantial
burden on religious practice.308 If the PPE followed the contours of
the constitutional right in this manner, it would not require reasonable accommodations for religious employees. If this were in fact the
rule, it arguably addresses the concern that this application of the
PPE would be too burdensome on employers.
C. Closer Cases
The constitutional provisions discussed immediately above are
strong candidates for inclusion in a PPE analysis. The provisions discussed below also represent very important liberty interests, but are
controversial enough, or sufficiently ill-defined, to warrant caution
regarding their usage by courts adjudicating PPE claims.
1. Reproductive Rights And Gender
The Due Process Clause gives rise not only to procedural due process protections, but substantive due process protections as well.
More specifically, there is a long line of cases stemming from the early to mid-twentieth century where the Supreme Court recognized a
right of “privacy” that protects individuals from state intrusions into
certain deeply private decisions. These decisions include both basic
ones about whether to bear children,309 and also, those involving chil-
306. See 42 U.S.C. § 2000e(j) (2012); see also D.C. CODE § 1-607.01 (2013) (mandating
reasonable accommodation for the free exercise of religion); TEX. LAB. CODE ANN. § 21.108
(West 2013) (prohibiting religious discrimination unless reasonable accommodation unduly
burdens the employer).
307. See, e.g., EEOC v. Reads, Inc., 759 F. Supp. 1150, 1158 (E.D. Pa. 1991) (describing
“religious garb” as falling within the ambit of statutory protection).
308. See Emp’t Div., Dep’t of Human Res. of Or. v. Smith, 494 U.S. 872, 879
(1990) (holding that neutral laws of general applicability do not implicate the Free
Exercise clause).
309. Griswold v. Connecticut, 381 U.S. 479, 484-85 (1965) (invalidating statute prohibiting the use of birth-control devices by holding that the penumbras of the guarantees in
the Bill of Rights create “zones of privacy”); Skinner v. Oklahoma, 316 U.S. 535, 541-42
(1942) (striking down a forced sterilization act while addressing marriage and procreation
as being “fundamental to the very existence and survival of the race”).
1002
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
drearing,310 family living arrangements,311 and intimate relations.312
The strength of these rights is bolstered by statutes that protect
against discrimination on the basis of marital status.313
Within the general area of substantive due process, one area ripe
for PPE analysis is the area of reproductive rights. In a variety of
cases around the country, wrongful discharge plaintiffs have claimed
they were fired for having abortions.314 Most of these cases either
went unresolved or were settled.315
The array of statutory remedies for persons in this situation is unclear. Employees discriminated against for having abortions likely
have a claim under the Pregnancy Discrimination Act (PDA).316 The
Third and Sixth Circuits have so held,317 and the EEOC regulations
so indicate.318 This right, apparently, also extends to adverse employment actions taken because an employee merely considers having
an abortion.319
A related potential issue involves an employer taking an adverse
action against an employee for not having an abortion. Such an employer may prefer that the employee not bring the pregnancy to term
for practical reasons: for example, out of fear the employee will ask for
maternity leave and flexible hours, and/or will not be able to devote the
same energy to the job. Or, the employer may disapprove of the employee becoming pregnant out of wedlock, or through an illicit affair.
Several state courts have recognized the PPE for pregnancy discrimination. Both Washington and California have recognized such
310. Pierce v. Soc’y of Sisters, 268 U.S. 510, 534-35 (1925) (rejecting statute requiring
all children to attend public elementary school); Meyer v. Nebraska, 262 U.S. 390, 399
(1923) (finding statute requiring all classes to be taught in English unconstitutional under
a substantive due process analysis).
311. Village of Belle Terre v. Boraas, 416 U.S. 1, 8 (1974) (upholding limitations on
single-family dwellings to traditional families or to two or fewer unrelated persons).
312. Lawrence v. Texas, 539 U.S. 558, 578 (2003) (holding that intimate consensual
sexual conduct was part of the liberty protected by substantive due process).
313. See, e.g., Federal Equal Credit Opportunity Act, 15 U.S.C. § 1691 (2012); ALASKA
STAT. § 18.80.220 (2013) (both making it unlawful for employers to discriminate on the
basis of, inter alia, marital status).
314. Noble, supra note 9. Not all such cases were resolved, id., and at least one was
resolved through confidential settlement. See Craver, supra note 9.
315. Noble, supra note 9.
316. See 42 U.S.C. § 2000e-2(a)(1) (2012) (barring sex discrimination); see also § 2000e(k) (defining sex discrimination to include discrimination on account of “pregnancy, childbirth, or related medical conditions”). In 1978, Congress passed the Pregnancy
Discrimination Act as an amendment to Title VII. See Turic v. Holland Hospitality, Inc., 85
F.3d 1211, 1213 (6th Cir. 1996).
317. Doe v. C.A.R.S. Prot. Plus, Inc., 527 F.3d 358, 364 (3d Cir. 2008); Turic, 85
F.3d at 1214.
318. 29 C.F.R. § 1604 app. (2013).
319. Turic, 85 F.3d at 1214.
2014]
COMPANY CONDUCT
1003
claims, drawing from their state constitutions to derive the requisite
public policy where the employer’s small size made the relevant state
statute inapplicable.320 By contrast, an intermediate appellate court
in Georgia rejected such a pregnancy discrimination PPE claim, reasoning that courts generally should defer to the legislature in deciding what public policies could trump the general rule of at-will employment.321 Perhaps cognizant of the PDA, the Georgia opinion refers to the fact that in addition to state statutes, there are unspecified “federal statutes” that also limit the at-will doctrine.322 But the
extent to which such federal statutes should inform the “public policy” analysis for PPE purposes, the court reasoned, is a decision best
left to the state legislature.323
This seems to be a proper area for the PPE. The constitutional
right of persons to make decisions about reproduction has a long lineage. Early in the twentieth century, the Supreme Court recognized
an unenumerated federal constitutional right to decide to reproduce
free from government interference.324 The Court extended the protection for the right to reproduce to the decision not to reproduce, guaranteeing a right of access to contraception.325 In order to recognize such
rights despite their not being expressly laid out in the Constitution’s
text, the Court had to find that these rights were “fundamental,” i.e.,
that they were “implicit in the concept of ordered liberty.”326 Forty
years ago, the Court recognized that the constitutional right to decide
not to reproduce extended to the decision to have an abortion, a right
which was reexamined and reaffirmed twenty years later and which is
still recognized today.327 Given the long history of reproductive rights,
320. Badih v. Myers, 43 Cal. Rptr. 2d 229, 293 (Cal. Dist. Ct. App. 1995); Roberts v.
Dudley, 993 P.2d 901, 911 (Wash. 2000).
321. Borden v. Johnson, 395 S.E.2d 628, 629 (Ga. Ct. App. 1990).
322. Id.
323. Id.
324. See Skinner v. Oklahoma, 316 U.S. 535, 541 (1942) (invalidating law providing for
forced sterilization of recidivist offenders).
325. See Carey v. Population Servs. Int’l, 431 U.S. 678, 694 (1977) (extending this right
to persons over age sixteen); Eisenstadt v. Baird, 405 U.S. 438, 453 (1972) (extending this
right to unmarried couples); Griswold v. Connecticut, 381 U.S. 471, 485-86 (1965) (recognizing this right for married couples).
326. Washington v. Glucksberg, 521 U.S. 702, 703 (1997) (citing multiple Supreme
Court cases using this phrase). The Court has used several different formulations of this
test for being a “fundamental” right. See id. at 721 (observing fundamental rights to be
“deeply rooted in this Nation's history and tradition”); see also Planned Parenthood of Se.
Pa. v. Casey, 505 U.S. 833, 851 (1992) (noting that these fundamental rights “involv[e] the
most intimate and personal choices a person may make in a lifetime, choices central to
personal dignity and autonomy”).
327. See Casey, 505 U.S. at 869-74 (reaffirming the core holding of Roe regarding the
right to an abortion, and adopting an “undue burden” test for all restrictions on abortion
prior to fetal viability); Roe v. Wade, 410 U.S. 113, 153 (1973) (holding a constitutional
right to abortion).
1004
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
it seems a good candidate for the kind of public policy which courts can
rely on to protect at-will employees under the common law.
As a matter of constitutional rights, it should make no difference
whether the right in question is the right to bear a child or not to
bear a child. The Court has made clear that among the most fundamental rights is the “decision whether to bear or beget a child.”328 Arguably, though, abortion may be properly separated from the rights
to be pregnant or to use contraception. Because a difficult moral
question exists as to exactly when a fetus becomes a “person,” it is
distinct. This philosophical question is still largely unresolved in our
society, making abortion rights extremely controversial. For this reason, perhaps courts should be cautious about wading into this area
with the PPE.
Further, because so many people have strongly held religious beliefs about abortion, court-imposed limits on employers via the PPE
might trigger constitutional issues about the employers’ own religious freedom rights.329 If an employer considers abortion to be murder, perhaps it should not be required to (from its viewpoint) support
murderers by employing them. Indeed, the awkwardness of continued employment after the abortion would seem most acute in small
firms with only a few employees, where employer and employee likely
personally interact. These are precisely the types of firms where the
PPE might be most relevant, given its utility as a backup where
statutory remedies are unavailable.
The solution may be to recognize abortion rights only narrowly
through the PPE. Under current precedent, a constitutional right to
abortion exists only before fetal viability.330 Outside that period, the
right exists only where necessary to protect the woman from serious
threats to life or health.331 Public opinion polls show a consistent majority of Americans favor abortion rights, at least during the very early stages of pregnancy.332 But such support drops during later stages
of the pregnancy.333 If PPE protection were to exist for abortion at all,
328. Eisendstadt, 405 U.S. at 453.
329. See Michael Dichiara, A Case of First Impression: The Third Circuit Recognizes
that Having an Abortion Is Protected by Title VII, 33 RUTGERS L. REC. 12, 14-15 (2009)
(raising this concern with respect to statutory protection for employment discrimination
targeting abortion).
330. Casey, 505 U.S. at 846.
331. Id.
332. Abortion, GALLUP HISTORICAL TRENDS (May 13, 2014), http://www.gallup.com/poll/
1576/abortion.aspx (over 60% of Americans believed abortions should be legal in the first
trimester; percentage dropped to less than 30% with respect to the second trimester).
333. Lydia Saad, Majority of Americans Still Support Roe v. Wade Decision, GALLUP
(Jan. 22, 2013), http://www.gallup.com/poll/160058/majority-americans-support-roe-wadedecision.aspx (acknowledging that “support drops off sharply, to 27%, for second-trimester
abortions, and further still, to 14%, for third-trimester abortions”).
2014]
COMPANY CONDUCT
1005
it should be limited to abortions taking place to prevent serious and
legitimate health risks to women, or which take place relatively early
on during pregnancy.
2. Substantive Due Process
A Native-American sales clerk wears shoulder-length hair as an
expression of his identity and spirituality. His private employer, a
grocery store, orders him to cut his hair; when he refuses, it fires
him. Does the employee have any legal remedy? The Alaska Supreme
Court said no, rejecting a claim that the employee had any right of
privacy or free expression as against a private employer, given the
constitutional requirement of state action.334
The court resolved the issue by rejecting the threshold notion that
constitutional principles had any analogous relevance to the PPE’s
regulation of purely private conduct. But the case presents an interesting test for the use of substantive due process as a source of policy
under the PPE.335
Substantive due process doctrine suggests that there exist certain
individual rights that are so fundamental that the state may not burden them, even if they are not expressly enumerated in the text of the
Constitution.336 Substantive due process is controversial precisely because it is open-ended. The recognition of rights untethered to constitutional text creates the potential for subjective “judicial activism.”337
Thus, with the exception of those reproductive rights like contraception for which there has emerged a sound and broad judicial and
public opinion consensus, courts should not use substantive due process as a “catch-all” measure to broaden the PPE, lest the exception
swallow the rule and all non-merit-related grounds for termination
become invalid violations of an employee’s right to be free from arbitrary action. If at-will employment is to retain any robust meaning,
334. Miller v. Safeway, Inc., 102 P.3d 282, 288-90 (Alaska 2004).
335. One can safely put aside consideration of this as a Free Speech issue. Where conduct as opposed to pure speech is concerned, the First Amendment’s Free Speech Clause
affords no protection unless the conduct is properly considered “expressive conduct.” Here,
while the Native American may have intended his long hair to express an idea, it may not
be the case that most people would reasonably interpret the long hair as principally expressive. See Texas v. Johnson, 491 U.S. 397, 409 (1989) (defining “expressive conduct”
under the First Amendment so as to require that persons would reasonably interpret the
conduct as being expressive).
336. See James W. Ely Jr., Due Process Clause, in THE HERITAGE GUIDE TO THE
CONSTITUTION 395 (Edwin Meese III et al. ed., 2005) (describing these fundamental rights
to be “beyond the reach of government”).
337. See Luís Roberto Barroso, The Americanization of Constitutional Law and its Paradoxes: Constitutional Theory and Constitutional Jurisdiction in the Contemporary World,
16 ILSA J. INT'L & COMP. L. 579, 580, 599-604 (2010) (using this term to describe an era in
which “contemporary democracies” sought to advance rights not listed in the Constitution).
1006
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
employers must be able to fire people for irrational, arbitrary, idiosyncratic, even quirky reasons, as long as those reasons do not
threaten broadly, strongly recognized rights.338
3. Second Amendment
Another candidate for the PPE is the Second Amendment. Within
the last few years, the Supreme Court has recognized that this
Amendment does indeed guarantee a right to bear arms even for an
individual with no connection to militia service.339 This right extends
to the use of arms reasonably related to individual self-defense, and
includes those arms, including handguns, which are not “dangerous
and unusual weapons.”340 It applies to state governments as well as
the federal government.341 While the Supreme Court cases to date
have dealt with the use of guns in the home, credible arguments exist
for protecting the right to carry arms outside the home into public
areas to further the self-defense rationale.342 Should the PPE protect
the right of employees to bring weapons to work?
Courts are just beginning to address the general issue of the right
to bear arms at work, and are split. In Mitchell v. University of Kentucky, the Supreme Court of Kentucky extended the PPE to the right
to bear arms, finding wrongful discharge where a public university
fired a properly licensed at-will employee for keeping a weapon in his
car on university property in violation of university rules.343 In that
case, conflicting statutes seemed to side with both employee and employer in this situation, and the court concluded, through a general
338. See Hunter v. Port Auth. of Allegheny Cnty. 419 A.2d 631, 634-35 (Pa. Super. Ct.
1980) (state constitutional provision protecting freedom to seek a livelihood prevented refusal to hire based on prior criminal convictions unrelated to job merit; though the employer was government agency, language suggests possible broader application); see also
Fellhauer v. City of Geneva, 568 N.E.2d 870, 877 (Ill. 1991).
339. District of Columbia v. Heller, 554 U.S. 570, 635 (2008).
340. Id. at 627.
341. McDonald v. City of Chicago, 130 S. Ct. 3020, 3036-44 (2010).
342. United States v. Masciandaro, 638 F.3d 458, 467 (4th Cir. 2011) (“[T]here now
exists a clearly-defined fundamental right to possess firearms for self-defense within the
home. But a considerable degree of uncertainty remains as to the scope of that right beyond the home and the standards for determining whether and how the right can be burdened by governmental regulation.”); Michael P. O’Shea, Modeling the Second Amendment
Right to Carry Arms (I): Judicial Tradition and the Scope of “Bearing Arms” for SelfDefense, 61 AM. U. L. REV. 585, 615 (2012) (interpreting Heller to suggest that “the defensive role for which handguns are uniquely suited—routine carry outside the home—is also
constitutionally protected”). Cf. Scott v. Extracorporeal, Inc., 545 A.2d 334, 342-43 (Pa.
Super. Ct. 1988) (declining to apply exception to asserted right of self-defense in workplace
fistfight, deferring to management to assign responsibility to some or all participants in
disruptive workplace altercations).
343. 366 S.W.3d 895, 903 (Ky. 2012).
2014]
COMPANY CONDUCT
1007
review of still other statutes, that the termination violated Kentucky’s public policy.344
Although the employer in this case was a public entity, the court
found a public policy that extended to private employers as well.345 The
court did not rely expressly on the Second Amendment or a Kentucky
Constitution provision; indeed, it said it would “refrain from reaching
constitutional issues when other, non-constitutional grounds can be
relied upon.”346 Since the employer was a state actor, this statement
could merely indicate an unwillingness to interpret directly either the
state or federal constitution; it may not explicitly address whether the
court felt comfortable in drawing upon Second Amendment or related
constitutional doctrines in reaching the PPE merits.
Contrast Hansen v. America Online, Inc., where the Supreme Court
of Utah rejected a similar claim involving the keeping of guns in an
employer’s parking lot.347 The court looked to the Utah Constitution in
evaluating the state’s public policy. That constitution did protect a
right to bear arms, but did not specifically address rights of employees
during working hours.348 Given the court’s general preference to apply
the PPE “parsimoniously” in deference to the default at-will doctrine,
it declined to extend the right to bear arms to the workplace.349
The two cases might be reconciled by noting that the Kentucky
statutes provided protection for employees of private employers,
whereas the legislative history of the Utah statutes suggested no application to private employers.350 However, the at-will-friendly Utah
case arose before the U.S. Supreme Court decisions recognizing for
the first time an individual right to bear arms,351 while the PPEfriendly Kentucky case arose afterwards. The difference may be
simply due to evolving constitutional doctrine.
A related issue arose in West Virginia, where the state supreme
court recognized a PPE claim based on a policy guaranteeing a right
of self-defense.352 In that case, the plaintiff was discharged for disarming a woman who was attempting to rob the employer’s premises,
an act that violated company policy against interfering with a store
robbery.353 The court looked to the state constitution, state statutes,
344. Id. at 900-03.
345. Id. at 901.
346. Id. at 898 (quoting Baker v. Fletcher, 204 S.W.3d 589, 597-98 (Ky. 2006)).
347. 96 P.3d 950, 956 (Utah 2004).
348. Id. at 953.
349. Id. at 954.
350. Compare Mitchell, 366 S.W.3d at 901, with Hansen, 96 P.3d at 954-55.
351. McDonald v. City of Chicago, 130 S. Ct. 3020, 3036-44 (2010); District of Columbia
v. Heller, 554 U.S. 570, 635 (2008).
352. Feliciano v. 7-Eleven, Inc., 559 S.E.2d 713, 723-24 (W. Va. 2001).
353. Id. at 716.
1008
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
and West Virginia case law, finding the existence of a public policy
favoring a right of self-defense.354 However, the court held that this
right must be balanced against the employer’s legitimate interests in
keeping order on its premises.355 Thus, firing an employee for defending himself creates a presumption of PPE wrongful discharge liability, but the employer may rebut that presumption by showing it had
a legitimate business reason for the termination.356
This latter approach seems more reasonable, especially in the context of bringing guns to work. Although some states have recognized
by statute a right of employees to bring weapons to work,357 in the
absence of such statutes, there is a paucity of case law supporting
such a right as a constitutional matter. That is not surprising, since
reasonable people differ as to whether an employer has a legitimate
interest in eliminating the risk of gunplay on its premises or preferring other means to ensure employee safety. Thus, standing on its
own and without state statutory support, the Second Amendment,
and similar constitutional provisions regarding the right to bear
arms, seem unlikely candidates for the PPE.358 This may change, given that the law in this area is just beginning to take shape. For example, the Supreme Court has not yet established the applicable
standard of review for Second Amendment rights. Nor has the Court
determined what interest an employer has in regulating arms in the
workplace. How the federal courts ultimately treat this employer interest will inform the development of the PPE with respect to the
Second Amendment.
VI. CONCLUSION
Courts considering constitutional claims borrow from common law
doctrine. There is no reason to treat with suspicion the reverse dynamic. Constitutional principles are compelling sources of our public
policy. They stem from foundational principles embodying our most
deeply held beliefs. They should certainly inform the analysis when
courts apply the PPE to the at-will employment rule.
354. Id. at 718-22.
355. Id. at 723.
356. Id. at 723-24.
357. See, e.g., GA. CODE ANN. § 16-11-135 (2013) (certifying that private and public
employers do not have the right to prohibit their employees from keeping lawful fir earms locked in their vehicles); VA. CODE ANN. § 15.2-915 (2012) (“[N]o locality shall
adopt any workplace rule . . . that prevents an employee of that locality from storing at
that locality’s workplace a lawfully possessed firearm and ammunition in a locked pr ivate motor vehicle.”).
358. See Gardner v. Loomis Armored Inc., 913 P.2d 377, 386 (Wash. 1996) (PPE claim
succeeded where company fired security guard for violating company rule in order to save
woman from life-threatening hostage situation).
2014]
COMPANY CONDUCT
1009
But courts should be circumspect in applying to the private employment context constitutional principles designed to cabin abuses
by government. In doing so, they should limit their gaze to constitutional rules creating rights held by individuals—rights which are
clear, uncontroversial, and readily analogous to the situation of a
private employee. Matters of free speech, Fourth Amendment privacy, and equal protection fall easily into this category. Criminal procedure rights, substantive due process concerns, and Second Amendment issues do not.
Traditionally, business and employment law questions are viewed
as inquiries distinct from, and perhaps more pragmatic than, the
seemingly more abstract doctrines of constitutional law. The PPE
provides a means for an open exchange of ideas between these two
ostensibly distinct areas of the law. At the same time, the use of constitutional values in the PPE creates a fail-safe, a means of filling the
gap left by statutory and contract theories to curb the most troubling
employer abuses.
Understandably, employers will be concerned about adding yet
another set of grounds for employees to sue them. The employeremployee relationship is, and should remain, a business decision and
the at-will doctrine, therefore, should not be undermined. However, if
we are to continue recognizing a PPE—and we should, to curb the
most outrageous subversions of public policy by employers while also
acknowledging the economic vulnerability of employees—then it
makes no sense to consider “lesser” sources of public policy but suddenly become shy about using the “greater” source of the Constitution.
The law sets a floor for minimum fair treatment of employees. A
properly robust, and properly selective, use of constitutional law to
inform the PPE raises that floor where it could use some raising.
1010
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:945
MAKING ROOM FOR COOPERATIVE INNOVATION
LIZA S. VERTINSKY
ABSTRACT
Patent law, created in response to a constitutional mandate to encourage innovation,
may be discouraging important forms of cooperative innovation. Advances in technology
have enabled new ways of pooling knowledge and computational capabilities, facilitating
cooperation among many participants with complementary skills and motivations to collectively solve complex problems. But emerging models of cooperative innovation increasingly
run into patent roadblocks.
Why might patent law sometimes thwart instead of support socially beneficial cooperative innovation? The problem lies in the tensions between the market-based incentives that
patent law creates and the mechanisms that support emerging models of cooperative innovation. The complexity and cost of solving contemporary public challenges are nudging diverse
participants together to collectively build their knowledge, but patents often serve to keep
them apart. While digital technologies enable new forms of massively distributed, open and
collaborative intellectual production, patents threaten the vitality and even the viability of
these promising types of innovation.
In this Article I use two examples—the risk of crowding out crowd science and the battle
between proprietary software companies and free open source software platforms—to illustrate how patent law in its current form may sometimes impede beneficial cooperation in
innovation. I then suggest how we might limit the negative effects of patents in contexts of
cooperative innovation without undermining the patent system.
I. INTRODUCTION ..................................................................................................
II. AN ALTERNATIVE WAY OF INNOVATING ............................................................
A. Distinct Features of Cooperative Innovation .............................................
B. Two Examples ............................................................................................
1. Crowd Science: The Example of Foldit ................................................
2. Open Source Software: The Example of Linux ....................................
III. PATENT ROADBLOCKS .......................................................................................
IV. PATENT REMEDIES WITH COOPERATION IN MIND .............................................
A. Remedies Without Context .........................................................................
B. Principles for the Design of Remedies ........................................................
1. First Principle ......................................................................................
2. Second Principle ..................................................................................
3. Third Principle.....................................................................................
V. CONCLUSION .....................................................................................................
1067
1072
1073
1076
1076
1086
1095
1103
1103
1110
1112
1115
1121
1123
I. INTRODUCTION
“In the long history of humankind (and animal kind, too), those
who learned to collaborate and improvise most effectively have prevailed.”—Charles Darwin
Systems of decentralized, massively distributed open innovation
are emerging with increasing frequency and with the ability to harness new resources in powerful new ways.1 People with no prior in1. See, e.g., Yochai Benkler, Law, Policy, and Cooperation, in GOVERNMENT AND
MARKETS: TOWARD A NEW THEORY OF REGULATION 299 (Edward J. Balleisen & David A.
Moss eds., 2010); see also ERIC VON HIPPEL, DEMOCRATIZING INNOVATION (2005). For a list
1068
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
volvement in biotechnology contribute hours of their time to solve
protein-folding puzzles posed by Foldit, an online video game that
uses crowd science to solve complex scientific problems.2 Data gathered from volunteer bird watchers through the global orthinological
network eBird are used to detect important environmental changes
that might otherwise go unnoticed.3 The U.S. Air Force is harnessing
collaborative online platforms to solicit input from students on complex
technological problems,4 and the National Institute of Health is enlisting the help of citizen scientist volunteers to collect and analyze data
from bacteria samples as part of the American Gut project.5 All of
these processes of crowd science rely on cooperation among many participants with diverse motivations and skills to collectively and openly
develop solutions to complex scientific or technological problems.6
Advances in technology have enabled greater computational capabilities and new ways of pooling knowledge. They have facilitated the
growth of massively distributed open access innovation processes
that rely upon voluntary participation by large numbers of people
who bring with them a diversity of skills and perspectives.7 These
participants share their ideas and discoveries openly with other
members of the innovation community in order to collectively and
cumulatively advance the innovation process. I refer to these kinds of
innovation processes as “cooperative innovation.” Cooperative innovation can harness underutilized and unused human resources and
combine existing disciplines and perspectives in new ways to solve
previously intractable scientific problems. In some cases it may comof some crowd science projects, see Citizen Science, SCI. AM., http://www.scientific
american.com/citizen-science/ (last visited Mar. 22, 2014).
2. See The Science Behind Foldit, FOLDIT, http://fold.it/portal/info/about (last visited
Mar. 22, 2014).
3. Jim Robbins, Crowdsourcing, for the Birds, N.Y. TIMES (Aug. 19, 2013),
http://www.nytimes.com/2013/08/20/science/earth/crowdsourcing-for-the-birds.html?page
wanted=all&_r=0 (eBird aggregates data about bird sightings that could not be collected other than through individual observations and uses it to uncover changes in
the environment).
4. Jane L. Levere, Air Force Asks Students to Solve Real World Problems, N.Y. TIMES
(July 28, 2013), http://www.nytimes.com/2013/07/29/business/media/air-force-asks-stud
ents-to-solve-real-world-problems.html.
5. See American Gut, HUMAN FOOD PROJECT, http://humanfoodproject.com/americangut/
(last visited Mar. 22, 2014).
6. While these and other examples of what I am calling “cooperative innovation”
have important similarities, an equally important distinction can be drawn between
projects that rely primarily on cooperative data gathering, such as in the eBird and
American Gut project examples, and projects that involve cooperative problem solving and
free sharing of ideas, such as the Foldit and U.S. Air Force examples. The concerns that I
address in this paper apply most strongly to the latter cases, where at least some
participants are engaging in more than collecting data, although the lines between
activities that generate data and activities that lead to invention are often blurred.
7. See, e.g., VON HIPPEL, supra note 1; Benkler, supra note 1.
2014]
COOPERATIVE INNOVATION
1069
plement existing modes of market-driven innovation, and in other
cases it may challenge them, increasing competition in important areas of intellectual production.8 However, emerging models of cooperative innovation sometimes run into patent roadblocks.
The problems that patents pose for cooperation in processes of
innovation are becoming difficult to ignore.9 The patent litigation
wars between major players in the smart phone industry such as
Apple, Samsung, Google, and Microsoft illustrate the divisive role
that patents can play in an industry that relies upon the shared
development and use of technology standards to achieve interoperability, particularly when network effects are important. By some
accounts Apple and Google now spend more on patent litigation than
they do on R&D.10 These problems are even greater for systems
of cooperative innovation such as open source software and crowd
science. Open source software systems like Linux challenge proprietary products like Microsoft Windows in the marketplace, only to
find their viability threatened by patent lawsuits.11 Ironically,
the open source software community finds it necessary to spend
substantial resources acquiring patents as a way of protecting free
software use.12 While harder to detect and measure, the problems
8. Scholars such as Yochai Benkler have argued that cooperative innovation is
valuable in itself, offering a way of democratizing innovation. In this article I am not
arguing that cooperative innovation is better than existing approaches, or even that
cooperation is always good. I am simply arguing that some kinds of cooperative innovation
have the potential to increase social welfare and deserve a chance to compete with
alternative approaches.
9. See, e.g., Rochelle Cooper Dreyfuss, Does IP Need IP? Accommodating Intellectual
Production Outside the Intellectual Property Paradigm, 31 CARDOZO L. REV. 1437 (2010)
(describing the need to change IP law to accommodate new forms of collaborative intellectual production); Aija Leiponen & Justin Byma, If You Cannot Block, You Better Run:
Small Firms, Cooperative Innovation, and Appropriation Strategies, 38 RES. POL’Y 1478
(2009) (“Earlier research has emphasized patents and trade secrets as key strategies of
appropriation, yet these strategies do not appear to be very beneficial for small firms engaged in cooperative innovation. These results raise policy questions regarding the functionality of the existing system of intellectual property rights.”); Mark A. Lemley, The Myth
of the Sole Inventor, 110 MICH. L. REV. 709, 709 (2012); Katherine J. Strandburg, Evolving
Innovation Paradigms and the Global Intellectual Property Regime, 41 CONN. L. REV. 861,
863-65 (2009) (arguing that current IP regimes are not designed to cope with changes in
the innovative process).
10. See Charles Duhigg & Steve Lohr, The Patent, Used as a Sword, N.Y. TIMES (Oct. 7,
2012), http://www.nytimes.com/2012/10/08/technology/patent-wars-among-tech-giants-can-stif
le-competition.html?pagewanted=all.
11. See Roger Parloff, Microsoft Takes on the Free World, FORTUNE (May 14, 2007, 9:35
AM), available at http://money.cnn.com/magazines/fortune/fortune_archive/2007/05/28/
100033867/ (“Microsoft claims that free software like Linux, which runs a big chunk of
corporate America, violates 235 of its patents. It wants royalties from distributors and users.
Users like you, maybe.”).
12. See Clark D. Asay, Enabling Patentless Innovation, MD. L. REV. (forthcoming 2014)
(manuscript at 47-48), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_
id=2289326); Keith Bergelt & William Wong, Interview: Keith Bergelt Discusses Open Invention
1070
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
that patents create for systems like Foldit that rely most heavily on
non-market mechanisms might prove to be even more costly. Patents
may interfere with the growth and vitality of these cooperative systems, limiting the opportunities that they provide for socially beneficial intellectual production.
Scholars of innovation such as Yochai Benkler and Eric von Hippel have already challenged us to consider what changes to the design of the legal and institutional system are necessary to sustain
cooperative innovation.13 They have identified intellectual property
law, particularly patent law, as threatening the open, inclusive and
collaborative nature of these systems. But they and other scholars
following in their wake have left the precise contours and magnitude
of the patent threats and specific proposals for patent law change for
further study.14 This Article responds to the challenge by identifying
and addressing specific ways in which patent law may interfere with
non-market mechanisms that support cooperative innovation. It focuses on three mechanisms that play a critical role in systems of cooperative innovation: trust, benefit-sharing, and reciprocity.15 Patents interfere with these mechanisms by increasing incentives for
group members to defect from the group and by increasing threats
from third parties against the group. This can make it harder to sustain the intrinsic motivations and the related norms of trust and sharing that cooperative systems rely upon, as well as increasing the cost
and risk of participation to potentially unsustainable levels. I argue
that courts should take these negative effects into account when fashioning patent remedies in contexts of cooperative innovation.16
Network, ELEC. DESIGN (Mar. 6, 2014), http://electronicdesign.com/embedded/interview-keithbergelt-discusses-open-invention-network.
13. “Policymakers . . . can design institutions and social systems to foster cooperation
by shaping social and psychological dynamics, rather than by focusing on individual incentives. The question then becomes, what aspects of the design of an institution or system—
be it technical platform, legal rule, business process, or policy intervention—are likely to
lead to a stable cooperative social dynamic?” Benkler, supra note 1, at 302; see also VON
HIPPEL, supra note 1.
14. See Katherine J. Strandburg, Intellectual Property at the Boundary (N.Y.U. Sch. of
Law Pub. Law & Legal Theory Res. Paper Series, Working Paper No. 13-60, 2013) (setting
out an agenda for investigating how IP rules impact cooperative innovation systems).
15. While these basic mechanisms seem to capture important ways in which
many cooperative systems operate, there are many other aspects of cooperative systems
that distinguish them from market-based systems, including a rich variety of non-economic
motivations and incentives for participating. For a discussion of the variety of incentives,
as well as some of the mechanisms, driving open source software systems, see, for
example, Stephen M. Maurer & Suzanne Scotchmer, Open Source Software: The New
Intellectual Property Paradigm, in 1 HANDBOOK OF ECON. & INFO. SYS. 285 (Terrence
Hendershott ed., 2006).
16. Preventing cooperation may, however, sometimes be socially desirable. See, e.g., F.
Scott Kieff, On Coordinating Transactions in Intellectual Property: A Response to Smith’s
Delineating Entitlements in Information, 117 YALE L.J. POCKET PART 101 (2007); Klaus
2014]
COOPERATIVE INNOVATION
1071
The Article proceeds as follows. Part I explores the potential of
cooperative innovation as an alternative to the traditional marketdriven approach. It provides two examples of cooperative innovation,
crowd science and open source software, to illustrate both the power
of cooperative systems and their vulnerabilities to patent threats.
Part II examines how and why patents may create problems for cooperative innovation.17 Part III shows how patent law could be adapted
to accommodate cooperative innovation through limited changes to
patent remedies. It provides three principles for contextualizing patent remedies in ways that respond to and limit the costs of patents
for socially beneficial cooperative innovation.18 These principles
would be triggered and applied by the courts only in contexts of cooperative innovation,19 leaving other areas of patent law unchanged.
Making room for cooperative innovation in this way will facilitate
Kultti et al., Patents Hinder Collusion (Helsinki Ctr. of Econ. Res., Discussion Paper No.
144, 2007); see also Jonathan Barnett, Dynamic Analysis of Intellectual Property: Theory,
Evidence and Policy (USC Gould Sch. of Law Ctr. in Law, Econ. & Org. Res. Papers Series,
Research Paper No. C13-4, 2013) (noting that periods of weak patent protection may actually disadvantage new entrants).
17. See, e.g., Dreyfuss, supra note 9; see also Gregory N. Mandel, Left-Brain Versus
Right-Brain: Competing Conceptions of Creativity in Intellectual Property Law, 44 U.C.
DAVIS L. REV. 283 (2010) (noting IP law has failed to recognize insights from psychology,
neurobiology and cultural research about how to promote creativity, resulting in laws
based on distorting stereotypes of creativity); Gregory N. Mandel, To Promote the Creative
Process: Intellectual Property Law and the Psychology of Creativity, 86 NOTRE DAME L.
REV. 1999, 2000 (2011) (noting patents may have psychological effects that alter incentives
to innovate, particularly in collaborations); Michael Mattioli, Communities of Innovation,
106 NW. U. L. REV. 103, 103 (2012) (examining collective patent licensing as a form of collective behavior in the patent system).
18. For discussions of the need for principles to guide policy determinations about
patent remedies, see, for example, Thomas F. Cotter, Four Principles for Calculating
Reasonable Royalties in Patent Infringement Litigation, 27 SANTA CLARA COMPUTER &
HIGH TECH. L.J. 725 (2011); John M. Golden, Principles for Patent Remedies, 88 TEX. L.
REV. 505 (2010) (providing principles to guide policymakers in assessing the relative merits
of alternative policy proposals); Paul J. Heald, Optimal Remedies for Patent Infringement:
A Transactional Model (Univ. of Chi. Law Sch. John M. Olin Law & Econ., Working Paper
No. 431, 2008), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1278062.
19. The literature includes an increasing number of proposals for tailoring patent
remedies in ways that take the public interest in supporting innovation into account. See,
e.g., Sarah R. Wasserman Rajec, Tailoring Remedies to Spur Innovation, 61 AM. U. L. REV.
733 (2012); Ted Sichelman, Purging Patent Law of ‘Private Law’ Remedies, 92 TEX. L. REV.
517 (2014); Samson Vermont, Basing Patent Remedies on Harm to the World Instead of
Harm to the Patentee (May 12, 2012) (unpublished manuscript) (on file with author). It also
includes proposals for tailoring patent remedies to reflect the relational or transactional
aspects of patents. See, e.g., Thomas F. Cotter, Patent Holdup, Patent Remedies, and
Antitrust Responses, 34 J. CORP. L. 1151 (2009) (focusing on the relational value of patents
and the importance of employing practical reason as an approach to patent remedies);
Heald, supra note 18, at 1176 (stating that the goal of patent remedies is to provide
incentives for efficient transactions to occur while minimizing cost of transacting).
1072
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
competition among alternative modes of innovation without materially undermining the patent system.20
II. AN ALTERNATIVE WAY OF INNOVATING
“Gettin’ good players is easy. Gettin’ ‘em to play together is the hard
part.”—Casey Stengel
In a controversial presidential campaign speech, now known as
the “you didn’t build that” speech, President Obama told his audience
that “[i]f you were successful, somebody along the line gave you some
help,” referring to the essential role of shared infrastructure and collective knowledge in supporting individual achievement.21 Since then,
Obama has continued to challenge the image of the rugged individual
inventor and entrepreneur, emphasizing the social context in which
innovation takes place and the importance of cumulative contributions to business success.22 An emphasis on collaboration and cooperation among stakeholders is now evident in both federal and state
innovation policies.23
While the need for collaboration to solve scientific and technological problems is not new,24 what is new is the scale and complexity of
the problems that need to be solved and the large and diverse group
of people who can come together to solve them using decentralized,
20. I am not arguing that patents are always bad for cooperation. Patents may be
useful, even essential, in supporting certain kinds of coordination and cooperation in
innovation. See, e.g., Jonathan M. Barnett, The Illusion of the Commons, 25 BERKELEY
TECH. L.J. 1751 (2010) [hereinafter Barnett, The Illusion of the Commons]; Paul J. Heald,
A Transaction Costs Theory of Patent Law, 66 OHIO ST. L.J. 473 (2005); F. Scott Kieff,
Coordination, Property, and Intellectual Property: An Unconventional Approach to
Anticompetitive Effects and Downstream Access, 56 EMORY L.J. 327 (2006). Moreover, firms
have alternative appropriability strategies that they can use in place of patents, so
weakening patents may not necessarily mean greater access. See, e.g., Jonathan M.
Barnett, Is Intellectual Property Trivial?, 157 U. PA. L. REV. 1691 (2009). I argue only that
patents create systematic disadvantages for certain kinds of cooperative innovation that
provide social benefits such as lower cost innovation, competition with existing proprietary
systems, or the creation of new products and ideas that might otherwise be unavailable.
21. This speech now has its own spot on Wikipedia. See a full discussion of this and
related speeches at You Didn’t Build That, WIKIPEDIA.ORG, http://en.wikipedia.org/wiki/
You_didn’t_ build_that (last modified Jan. 19, 2014).
22. SEE BARACK OBAMA, U.S. PRESIDENT, REMARKS AT A CAMPAIGN EVENT IN ROANOKE,
VIRGINIA (JULY 13, 2012) (TRANSCRIPT AVAILABLE AT HTTP://WWW.WHITEHOUSE.GOV/THE-PRESSOFFICE/2012/07/13/REMARKS-PRESIDENT-CAMPAIGN-EVENT-ROANOKE-VIRGINIA).
23. See, e.g., Steven C. Currall & Ed Frauenheim, How the U.S. Can Lead on
Technological Innovation, SEATTLE TIMES (Mar. 7, 2014, 5:14 PM), http://seattle
times.com/html/opinion/2023076667_stevencurralledfrauenheimopedinnovationresearchxx
xml.html; Bruce Katz & Judith Rodin, Innovative State and City Government Solutions to
Watch in 2012, THE ATLANTIC (Jan. 17, 2012), http://www.theatlanticcities.com/jobs-andeconomy/ 2012/01/innovative-local-government-solutions-watch-2012/951/.
24. See generally Ernan McMullin, Openness and Secrecy in Science: Some Notes on
Early History, 10 SCI. TECH. & HUMAN VALUES 14 (1985) (examining the history of the
ideal and reality of open science).
2014]
COOPERATIVE INNOVATION
1073
low-cost, web-based technologies. Emerging forms of cooperative
innovation can harness new resources, bring multiple disciplines and
perspectives to bear on previously intractable scientific problems,
and increase competition in key areas of intellectual production.
I argue that innovation policies need to respond to the opportunities
of these cooperative innovation systems. To do so, however, requires
an understanding of how these systems diverge from traditional
market-driven modes of innovation in ways that leave them vulnerable to patents. I therefore explore the mechanisms on which cooperative systems of innovation rely and the potential effects of patents
on these mechanisms. Two examples of cooperative innovation help
to illustrate both the power of these models and their vulnerability
to patents.
A. Distinct Features of Cooperative Innovation
“[W]hat really distinguishes open source is not just source, but an
‘architecture of participation’ . . . .”—Tim O’Reilly
Cooperative innovation is used in this article to refer to open access innovation processes fueled by the voluntary participation of
large numbers of people who share their ideas and discoveries freely
and openly with other members of the innovation community. Entry
is free, or close to free, and the cost to participate is low. Progress is
highly sequential and dependent upon large numbers of small steps
that build upon each other. This kind of innovation diverges from
traditional market-based modes of innovation in its reliance on nonmarket mechanisms for sustaining cooperation.
Cooperation involves working together to advance common goals
or obtain mutual benefits. It can occur among even purely selfish
economically rational actors where individual economic interests
align with the collective interest. Sustaining cooperation when individual economic interests diverge from the interests of the collective
is more challenging. A number of studies have tried to identify the
factors that are needed to sustain cooperation beyond situations in
which rational economic actors have aligned interests.25 Case studies
25. See, e.g., YOCHAI BENKLER, THE PENGUIN AND THE LEVIATHAN (2011); SOCIAL
DILEMMAS: PERSPECTIVES ON INDIVIDUALS AND GROUPS (David A. Schroeder ed., 1995)
(giving an overview of different perspectives on decisions made by those in social dilemmas
and the factors that influence choices to act in the interests of the group); Benkler, supra
note 1; see also ROBERT C. ELLICKSON, ORDER WITHOUT LAW: HOW NEIGHBORS SETTLE
DISPUTES (1991); ELINOR OSTROM, GOVERNING THE COMMONS: THE EVOLUTION OF
INSTITUTIONS FOR COLLECTIVE ACTION (1990); Elinor Ostrom, A General Framework for
Analyzing Sustainability of Social-Ecological Systems, 325 SCI. 419 (2009). For an
organizational approach to understanding how patents influence knowledge communities,
see, for example, Fiona Murray & Scott Stern, Learning To Live with Patents: A Dynamic
Model of a Knowledge Community’s Response to Legal Institutional Change (Nov. 2008)
(unpublished manuscript) (on file with author).
1074
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
of open source software communities have been particularly helpful
in identifying the mechanisms that support cooperative intellectual
production.26 They have illustrated the importance of trust and reciprocity, and shown how these mechanisms can work when groups
become large and relationships attenuated.
In open source software communities, people cooperate with each
other even when they do not know each other and their interactions
are limited. Team members must decide how much they can trust
each other before joining the team and without the chance to establish personal relationships or forms of mutual control.27 Instead of
personal relationships, the decision to join the team is based on beliefs about the motivation of other team members to adhere to the
mutual norms of the open source software community. Participants
must believe that other members of the community will continue to
act in ways that are consistent with the open source model of free
sharing. This is an example of swift trust, a unique form of trust that
occurs between people or groups who come together in temporary
teams to solve collaborative tasks.28 The maintenance of these communities depends on “their ability to a) develop and enforce rules of
cooperation in a self-organized manner, and to b) develop selfenforcing swift trust which is based on generalized reciprocity between group members.”29 Keeping the cost of participation low is also
essential to the sustainability of these groups.30
A growing body of empirical work in other areas involving selfgoverning systems of cooperative resource management and use supports these findings, emphasizing the importance of trust—
particularly swift trust, benefit-sharing, and reciprocity in attracting
26. See, e.g., Greg R. Vetter, Commercial Free and Open Source Software: Knowledge
Production, Hybrid Appropriability, and Patents, 77 FORDHAM L. REV. 2087 (2009).
27. See Margit Osterloh & Sandra Rota, Trust and Community in Open Source Software Production, 26 ANALYSE & KRITIK 279 (2004).
28. For early development and use of “swift trust,” see, for example, Debra Meyerson
et al., Swift Trust and Temporary Groups, in TRUST IN ORGANIZATIONS 166, 168 (Roderick
M. Kramer & Tom R. Tyler eds., 1996). For an interesting discussion of swift trust as it
arises in different contexts and the factors affecting its formation, see Michael J. Fahy,
Understanding “Swift Trust” To Improve Interagency Collaboration in New York City
(Sept. 2012) (unpublished M.A. thesis, Naval Postgraduate School), available at
http://hdl.handle.net/10945/17362.
29. Osterloh & Rota, supra note 27, at 280; see also C. Brad Crisp & Sirkaa L. Jarvenpaa, Swift Trust in Global Virtual Teams: Trusting Beliefs and Normative Actions, 12 J.
PERSONNEL PSYCHOL. 45 (2013) (examining the role and vulnerabilities of swift trust in ad
hoc global virtual teams).
30. Copyright law has provided a limited enforcement mechanism, through the use of
licenses that impose varying commitments on users of open source software to make their
own contributions open and accessible. Even with these licenses in place, however, trust
remains an important part of sustaining open source software systems.
2014]
COOPERATIVE INNOVATION
1075
and sustaining participation.31 This work also suggests that to sustain cooperation, laws need to be applied in ways that support, or at
the very least minimize interference with, the norms, customs, and
organizational structures that support these critical non-market
mechanisms.32 Where members of a group are contributing their time
and ideas freely and voluntarily, their willingness to participate will
be influenced by the behavior of other group members and even, to a
lesser extent, by the behavior of non-group members.33 Their willingness to make contributions will depend on the extent to which other
group members act in accordance with shared group norms and reciprocate with their own contributions and the extent to which the contributions that group members freely provide remain free. Free riding
and private appropriation of the benefits produced in the group by
people outside of the group may also undermine internal motivations
to contribute. Problems may arise for the group where legal rules or
the incentives that these rules create undermine or conflict with the
shared understandings and commitments of the group, leading to deviations from accepted and expected cooperative behavior.
These insights are making their way into at least some areas of
the law. In contract law, for example, relational contract theory provides a view of contracts as relations rather than discrete transactions, with many of the contract terms left implicit and governed by
trust between the parties.34 Recent work on contracts and innovation
illustrates how firms use incomplete contracts to sustain cooperative
relationships in the face of imperfect information.35 In corporate law,
31. See, e.g., TRUST AND RECIPROCITY: INTERDISCIPLINARY LESSONS FROM EXPERIMENTAL
RESEARCH (Elinor Ostrom & James Walker eds., 2003); UNDERSTANDING KNOWLEDGE AS A
COMMONS (Charlotte Hess & Elinor Ostrom eds., 2007); STEVEN WEBER, THE SUCCESS OF
OPEN SOURCE (2004); Benkler, supra note 1; Dan M. Kahan, The Logic of Reciprocity: Trust,
Collective Action, and Law, 102 MICH. L. REV. 71 (2003); Osterloh & Rota, supra note 27; Margit Osterloh, Sandra Rota & Bernhard Kuster, Trust and Commerce in Open Source—A Contradiction? (Nov. 22, 2002) (unpublished manuscript), available at http://www.research
gate.net/publication/226907215_Trust_and_Commerce_in_Open_Source_A_Contradiction.
32. See, e.g., TRUST AND RECIPROCITY: INTERDISCIPLINARY LESSONS FROM EXPERIMENTAL
RESEARCH, supra note 31; UNDERSTANDING KNOWLEDGE AS A COMMONS, supra note 31.
33. See, e.g., Osterloh, Rota & Kuster, supra note 31, at 9.
34. See, e.g., Ian R. Macneil, The Many Futures of Contract, 47 S. CAL. L. REV. 691 (1974).
35. See, e.g., Ronald J. Gilson et al., Contracting for Innovation: Vertical
Disintegration and Interfirm Collaboration, 109 COLUM. L. REV. 431 (2009). Contracts
allow parties to structure their relationships with each other in ways that foster trust and
facilitate relationship-specific investments where the payoffs from the relationship cannot
be predicted and contracted for in advance. In some cases the terms governing the division
of payoffs from a shared innovation project are left open, the parties relying instead on the
braiding of formal terms for sharing information about the progress and prospects of their
joint activities with informal terms governing subsequent outcomes of the joint work. This
information-sharing regime “braids” together the formal and informal elements of the
contract in a way that endogenizes the growth of trust between the participants. See
Ronald J. Gilson et al., Braiding: The Interaction of Formal and Informal Contracting in
Theory, Practice, and Doctrine, 110 COLUM. L. REV. 1377 (2010).
1076
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
behavioral theories based on trust and trustworthiness challenge
conventional views of the firm.36 This work cautions against excessive
reliance on external sanctions that may undermine internal trust.37
As new forms of collective production emerge, there have even been
suggestions for creating an entirely new field of law, cooperation law,
to reflect arrangements between people that are based on a variety of
modes of sharing, cooperation, and collaboration—such as co-housing,
barter, and community-financed businesses.38 These legal responses
aim to modify existing formal rules where needed to support and protect beneficial informal rules and understandings. I suggest that internalizing the negative effects that patents may have on non-market
mechanisms of trust, benefit-sharing, and reciprocity in cooperative
systems will begin to pull similar considerations into patent law in
contexts, such as those described below, where they are most important.39 The following two examples illustrate both the power of
cooperative innovation and the vulnerabilities of such systems to the
incentives and associated transaction costs that patents create.
B. Two Examples
1. Crowd Science: The Example of Foldit
The power of crowd science, the use of many volunteers working
together to solve complex data intensive problems, has been illustrated in areas as diverse as finding planets, deciphering ancient
texts, and building climate models.40 Crowd science is increasingly
informing what we know about diverse natural phenomena such as
bird populations and their distributions,41 the pollination habits of
36. See, e.g., Margaret M. Blair & Lynn A. Stout, A Team Production Theory of
Corporate Law, 85 VA. L. REV. 247 (1999); Margaret M. Blair & Lynn A. Stout, Trust,
Trustworthiness, and the Behavioral Foundations of Corporate Law, 149 U. PA. L. REV.
1735 (2001) [hereinafter Blair & Stout, Trust] (arguing that the behavioral phenomena of
internalized trust and trustworthiness play important roles in encouraging cooperation
within firms); Donald C. Langevoort, Behavioral Approaches to Corporate Law, in
RESEARCH HANDBOOK ON THE ECONOMICS OF CORPORATE LAW 442 (Claire A. Hill & Brett
H. McDonnell eds., 2012).
37. See Blair & Stout, Trust, supra note 36.
38. See, e.g., Janelle Orsi, Cooperation Law for a Sharing Economy, YES! MAG. (Sept. 23,
2010), http://www.yesmagazine.org/new-economy/cooperation-law-for-a-sharing-economy (“A
new sharing economy is emerging—but how does it fit within our legal system? Time for a
whole new field of cooperation law.”).
39. For a discussion of how IP operates at the boundaries between groups, see, for
example, Strandburg, supra note 14.
40. GAMING FOR THE GREATER GOOD, http://gamingforthegreatergood.com (last visited
Mar. 22, 2014).
41. The Audubon Society’s Christmas Bird Count began in 1900. A transformed
version, eBird, was launched in 2002 by Cornell Lab of Ornithology in collaboration with
the National Audubon Society. About eBird, EBIRD, http://ebird.org/content/ebird/about/
2014]
COOPERATIVE INNOVATION
1077
bees,42 and the surface of the moon,43 and is also helping to solve puzzles in quantum physics,44 the anthropology of microbes,45 and the
early development stages of dinosaurs.46 These divergent crowd science projects share a common foundation built upon open access,
sharing, and cooperation between many heterogeneous participants
with diverse motivations that may not be consistent with a pricebased system. They illustrate the opportunities that cooperation outside of the marketplace offer for scaling up the collection and analysis
of large amounts of data and for building knowledge through the accumulation of many incremental contributions of information and ideas.47 Unfortunately they also share common vulnerabilities to a patent
system that is not designed with them in mind. The story of Foldit, a
particularly promising citizen science project, highlights both the risks
and opportunities of these kinds of cooperative innovation.48
This story starts with the creation of a video game called Foldit
that anybody can play.49 Within ten days of their start on December
16, 2010, players of this new online protein folding game were able to
solve the protein structure of a retrovirus similar to HIV, a structure
that had eluded scientists for over ten years.50 The results provided
scientists with important insights into the treatment of AIDS and
were published in a scientific journal with the video game team players as co-authors.51 This is only one of a number of scientific advances
made by Foldit players. Other contributions by Foldit players include
(last visited Mar. 22, 2014). eBird collects bird abundance and distribution data and enlists
public participation in analyzing the over eighty million observations. See id.
42. The Great Sunflower Project was started by a single academic researcher
interested in examining the pollination activities of bees, and it now has 90,000 registered
volunteers planting sunflowers and taking observation samples. See GREAT SUNFLOWER
PROJECT, http://www.greatsunflower.org (last visited Mar. 22, 2014).
43. See, e.g., Moon Zoo, ZOONIVERSE, https://www.zooniverse.org/project/moonzoo (last
visited Mar. 22, 2014).
44. See, e.g., The Story Behind “Quantum Moves,” SCI. HOME, http://www.science
athome.org (last visited Mar. 22, 2014) (describing Quantum Moves, a game that uses
community efforts to help build a quantum computer).
45. See American Gut, supra note 5.
46. See OPEN DINOSAUR PROJECT, http://opendino.wordpress.com/ (last modified
May 17, 2013).
47. See, e.g., Clay Shirky, How the Internet Will (One Day) Transform Government,
Address at TED Conference (June 2012), available at http://www.ted.com/talks/clay_
shirky_how_the_internet_will_one_day_transform_government.
48. See, e.g., Eric Hand, People Power, 466 NATURE 685 (2010) (discussing projects
involving the concept of distributed thinking, featuring Foldit); Collins Kilgore, Gaming for the
Greater Good, VAND. J. ENT. & TECH. L. (Sept. 27, 2011), http://www.jetlaw.org/?p=8381.
49. See FOLDIT, http://fold.it/portal/ (last visited Mar. 22, 2014).
50. See Firas Khatib et al., Crystal Structure of a Monomeric Retroviral Protease
Solved by Protein Folding Game Players, 18 NATURE STRUCTURAL & MOLECULAR
BIOLOGY 1175, 1177 (2011).
51. Id.
1078
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
the discovery of a unique enzyme “backbone” configuration for the
development of novel enzymes, a discovery described by scientists as
the most detailed remodeling of a protein structure by humans working through a computer-based process.52 This work moves beyond
protein folding, which is critical for understanding how the human
body works, and into the realm of protein design, which holds promise for advancing drug discovery. The protein design results were
published in Nature Biotechnology, a prestigious scientific journal,
with Foldit Players again included in the author list.53
Foldit was developed by David Baker, a protein research scientist
at the University of Washington, together with Zoran Povic and Seth
Cooper, computer scientists at the same university. The program is
supported through a collaboration between the Biochemistry Department and the Center for Game Science at the University of
Washington.54 The vision behind Foldit is one of enabling public participation in large-scale distributed science. The project “aims to predict, design, and understand biochemical structures, and study how
humans and computers can best work together to further these aims”
through a growing community of game-developed expert volunteers.55
Anybody can play the game simply by visiting the Foldit website,
consenting to user-friendly terms of use that focus primarily on explaining how information is collected and shared, and downloading
the free software needed to play the game.56
Participants in the game, most of who have little or no background
in biochemistry, are introduced to some basic concepts of protein folding and then engaged in solving “puzzles” designed around specific
protein structure problems that have been identified but not solved
by scientists. The players collaborate with teammates while competing against other players to obtain the highest-scoring (lowestenergy) models. The scoring system includes different categories
based on levels and types of contributions. The website identifies the
52. See Christopher B. Eiben et al., Increased Diels-Alderase Activity Through Backbone Remodeling Guided by Foldit Players, 30 NATURE BIOTECHNOLOGY 190, 192 (2012);
see also David Baker, More Amazing FoldIt Results and New Flu Virus Challenges, FOLDIT
(May 27, 2011, 6:41 PM), http://fold.it/portal/node/989769; Zoran, Recent Exciting Discoveries by Foldit, FOLDIT (Apr. 19, 2011, 5:15 PM), http://fold.it/portal/node/989576.
53. See Eiben et al., supra note 52 (showing that Foldit Players are included on the
publication list; reporting the use of game-driven crowdsourcing to enhance the activity of
a computationally designed enzyme through the functional remodeling of its structure).
54. See CENTER GAME SCI., http://centerforgamescience.com/site/ (last visited
Mar. 22, 2014).
55. See Foldit Terms of Service and Consent, FOLDIT, http://fold.it/portal/legal (last
updated Mar. 8, 2013).
56. New players begin by loading the game and watching tutorials that explain the
game and the scientific concepts needed to understand it. See How To Download and Start
Playing Foldit!, FOLDIT (Nov. 9, 2010, 11:47 PM), http://fold.it/portal/node/988864.
2014]
COOPERATIVE INNOVATION
1079
top players in each category and for each puzzle. It also provides
rankings based on global scores and a hall of fame for the very top
scorers. The game has separate rankings for soloists, who work on
their projects alone, and evolvers, who work on and improve solutions
that have been shared with other people. A high ranking is valued
and sought after by the active players, and many of them carefully
watch the scores and study the ways in which the rankings are calculated. Players are quick to point out areas where they believe the
scoring is not accurately reflecting good performance or otherwise
seems unfair.57
Three years into the life of the Foldit games, there is clear evidence that this massively multiplayer online game, which enlists
players worldwide to solve challenging protein-structure prediction
problems, offers improvements over the computational models currently being used by scientists.58 “Obsessive gamers’ hours at the
computer have now topped scientists’ efforts to improve a model enzyme, in what researchers say is the first crowd sourced redesign of a
protein.”59 The use of games offers valuable ways for engaging citizen
scientists, drawing on the vast amounts of free time that people are
willing to devote to problem solving just because they want to. “[T]he
average young person today in a country with a strong gamer culture
will have spent 10,000 hours playing online games by the age of
twenty-one. . . . [C]ollectively all the World of Warcraft gamers have
spent 5.93 million years solving the virtual problems of [that particular game world].”60 Foldit harnesses this energy to solve difficult and
data-intensive problems, serving as the flagship game for the concept
of using crowds to solve complex scientific problems.61
This system of intellectual production relies upon non-market
mechanisms of trust, benefit-sharing, and reciprocity. While the
game is structured as a competition to get the highest score, competi-
57. See, e.g., Bruno Kestemont, Actual Competitors for the Global Score, FOLDIT (Oct.
25, 2013, 9:46 AM), http://fold.it/portal/node/996160.
58. See Seth Cooper et al., Predicting Protein Structures with a Multiplayer Online
Game, 466 NATURE 756, 756 (2010).
59. Jessica Marshall, Victory for Crowdsourced Biomolecule Design: Players of the
Online Game Foldit Guide Researchers to a Better Enzyme, NATURE (Jan. 22, 2012),
http://www.nature.com/news/victory-for-crowdsourced-biomolecule-design-1.9872.
60. See Jane McGonigal, Gaming Can Make a Better World, Address at TED Conference
(Feb. 2010) (emphasis added), available at http://www.ted.com/talks/jane_mcgonigal_gaming_
can_make_a_better_world.
61. See, e.g., John Walker, The Inaugural Horace Awards for Forgotten IGF Entrants,
ROCK, PAPER, SHOTGUN (Jan. 8, 2013, 9:00 PM), http://www.rockpapershotgun.com/2013/
01/08/the-inaugural-horace-awards-for-forgotten-igf-entrants/ (discussing the Foldit game in
the wake of results on the winners of the Fifteenth Annual Independent Games Festival,
which is well known in the video game industry, and suggesting it should be recognized with
the Horace Award for Actually Advancing Science).
1080
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
tion among the players is moderated by the fact that there are no
monetary consequences attached to having a winning score. Moreover, the results of gameplay are portrayed as contributions to the
public domain of scientific knowledge. The Foldit site explains that
diversity advances this knowledge, so even those with lower scores
feel they are doing something important. The low-level incentives to
win are combined with norms of sharing and open communication to
foster a collaborative environment.62 Members are encouraged to
communicate, share ideas, and troubleshoot through the Foldit blog,
forum, wiki, or through sharing “recipes” useful in solving various
puzzles. The importance of sharing as a fundamental value for the
community is highlighted in the Foldit Terms of Service, which remind members that: “We are sharing what we learn with others from
all over the world. That is how science is done. We have to share so
others can learn, too.”63 Reciprocity in the exchange of ideas is also
encouraged. It is encouraged through member interests in sustaining
a robust forum for troubleshooting and a library of shared recipes to
advance gameplay. There is also a built-in reciprocal exchange between the players and Foldit. With the consent of the players, the
Foldit project continuously gathers and analyzes gameplay data such
as biochemical structures, algorithms, and tool and algorithm usage
that may have research benefits, and in return the project agrees to
give attribution to the players who make discoveries and to make
these discoveries publicly available.
Out of this collective effort has come publishable results and novel
discoveries.64 While scientific publications co-authored by video game
players may seem unusual, this could become the norm in at least
some areas of science if crowd science continues to progress at its
current rate. The Foldit community is large and growing. In January
2012, the game had 240,000 registered players and approximately
2200 active players.65 In January 2014 the game appears to have almost doubled this number of registered players and more than doubled the number of active players.66 The game continues to expand
not just its number of players, but also the reach of its problemsolving power. Most recently, the creators of the game have turned
their eyes to the design of new therapeutic enzymes and even small
62. Foldit community rules can be found online at Community Rules, FOLDIT,
http://fold.it/portal/communityrules (last updated Apr. 18, 2013).
63. Foldit Terms of Service and Consent, supra note 55.
64. See, e.g., Khatib et al., supra note 49, at 1175 (listing authors as including the
“Foldit Contenders Group” and the “Foldit Void Crushers Group”).
65. Marshall, supra note 59 (noting that active players are players who have logged in
and been active on the web site within the last week as measured on a particular day).
66. See Top Evolvers, FOLDIT, http://fold.it/portal/players (last visited Mar. 22, 2014).
2014]
COOPERATIVE INNOVATION
1081
molecule design, taking the crowdsourcing approach into the realm of
drug development.67
As puzzles move into areas with more direct commercial potential,
however, tensions between the non-market mechanisms on which
Foldit relies and proprietary interests enabled by patenting are likely
to emerge. Analogies can be drawn to the tensions that have emerged
in many areas of academic science as the distinction between research and commercial application has blurred.68 During its first few
years the Foldit website provided little information on intellectual
property ownership, not seeing it as something that was needed. In
response to queries by its participants, the site administrators/founders stated that the “Foldit project was initiated with the
goal of democratizing science, and we stand behind that. [T]he process of discovery and the eventual results of game play will all be
open domain.”69 In line with the original game philosophy of democratic science, the founders of the game then asked the game participants for their views on intellectual property ownership. Developer
chat discussions in 2012 were used to flesh out what a Foldit intellectual property ownership policy should look like.70
The community discussions about intellectual property ownership
raised a number of important issues about how patents might impact
Foldit. Some Foldit members focused on negative effects that patenting of results might have on the intrinsic motivations and norms of
sharing that motivate the game. They expressed the view that Foldit
is a public, volunteer-driven process, and therefore all ideas and contributions should remain public.71 Others focused more on the challenges of benefit-sharing in a system that works best when it is collaborative and invites cumulative refinements of ideas by competing
members of the game. Many of the contributions take the form of
computer recipes for improving game play, contributions which may
67. Rebecca Hersher, FoldIt Game’s Next Play: Crowdsourcing Better Drug Design,
SPOONFUL OF MEDICINE, NATURE MED. (Apr. 13, 2012, 4:35 PM), http://blogs.nature.com/
spoonful/2012/04/foldit-games-next-play-crowdsourcing-better-drug-design.html?WT.mc_id=
TWT_NatureBlogs.
68. See, e.g., Robert P. Merges, Property Rights Theory and the Commons: The Case of
Scientific Research, in SCIENTIFIC INNOVATION, PHILOSOPHY, AND PUBLIC POLICY 145
(Ellen Frankel Paul, Fred D. Miller, Jr. & Jeffrey Paul eds., 1996). I have explored the
challenges that patents create in biomedical research at greater length. Liza Vertinsky,
Universities as Guardians of Their Inventions, 2012 UTAH L. REV. 1949.
69. Zoran, Comment to Set My Mind at Ease, FOLDIT (Mar. 22, 2009, 3:31 AM),
http://fold.it/portal/node/267249.
70. See Developer Chat, FOLDIT (June 1, 2012), http://fold.it/portal/node/992849.
71. See, e.g., GlaciusCool, Comment to Foldit Ownership Policy, FOLDIT (May 31, 2012,
1:25 AM), http://fold.it/portal/node/992792 (“That which is created by the public should remain
in the hands of the public for the benefit of all human kind. Scientific discovery—even if it’s
just a molecular recipe—should be used by and available to all who seek the knowledge.”).
1082
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
not themselves be patentable or serve as patent-invalidating prior art
but which could lead to the discovery of something that is. Participants might not be willing to share their recipes freely if they
thought that someone else could use them to make and appropriate
results for private commercial gain. Other issues raised by Foldit
members focused on the practical challenges, transaction costs, and
administrative headaches that patents might create. Participants
worried about what formula would be used to allocate rights in a way
that fairly rewarded performance, and whether it would be possible
to discern who the co-inventors of an invention are in the context of
cumulative contributions. The possibility that some contributions
might come from employees with preexisting obligations to their
companies created a further layer of complexity and concern.
Ultimately, the Foldit administrators proposed the following ownership provision, guided no doubt by the University of Washington
Center for Commercialization:
Scientific discoveries will be made publicly available and the University of Washington will handle ownership of discoveries. All
significant scientific discoveries (such as structures, algorithms,
etc) made in game will be made publicly available. In the event
that some discoveries may warrant patent protection, University of
Washington will handle the patent application process. US patent
law will govern IP attribution for each discovery. Individual players who contributed to the discovery will be considered coinventors for any discovery produced through play.72
A few things are notable about this policy. First, the University of
Washington plays a direct role in managing patent choices. The University of Washington has an active technology transfer center, the
Center for Commercialization, and its intellectual property policies
and licensing practices reflect traditional assumptions about patenting and technology transfer. The University’s intellectual property
policies are “intended to show the University’s positive attitude toward transfer of results of its research to the private sector” and are
based on the assumption that “it is generally in the best interests of
the University and the public that patents be obtained and/or licenses granted.”73 This approach conflicts with one focused on the free
and open sharing and use of ideas.
72. Foldit Terms of Service and Consent, supra note 55; see also Seth Cooper, Comment to
Foldit Ownership Policy, FOLDIT (June 4, 2012, 5:48 PM), http://fold.it/portal/node/992792
(announcing a draft ownership policy).
73. PRESIDENT OF THE UNIV. OF WASH., EXEC. ORD. NO. 37: PATENT, INVENTION, AND
COPYRIGHT POLICY (last updated Oct. 27, 2003), available at http://www.washington.edu/
admin/rules/policies/PO/EO36.html.
2014]
COOPERATIVE INNOVATION
1083
Second, while the implication is that inventors will be the owners
of their inventions, the policy is not clear about what rights the University will have, what rights the inventors will have, and what restrictions will be imposed on participant inventors who end up with
ownership rights over a patented invention. The policy indicates only
that the University will handle ownership and that significant scientific discoveries will be made publicly available. Perhaps in response
to concerns about the sustainability of a public domain approach, the
Foldit developers, who are University of Washington employees, have
announced to the Foldit community that they have committed to assign all proceeds from any patents on inventions that they discover
back to the development of the Foldit community.
Third, although the policy provides that “[i]ndividual players who
contributed to the discovery will be considered co-inventors for any
discovery produced through play,”74 joint inventorship is determined
solely by patent law.75 In some cases, individuals might think that
they are co-inventors when they are not, since the statutory definition of joint inventorship is narrower than the policy suggests.76
Finally, the Foldit Terms of Service can be changed at any time,
leaving the University to reconsider its stance on ownership should
the Foldit game start to yield results that are commercially lucrative.
While the concerns of the Foldit members have yet to be realized,
the potential of the game to aid in areas of commercial interest, such
as drug discovery, will inevitably increase tensions between the system of volunteer contributions and sharing on the one hand and patenting and revenue sharing on the other hand.77 One of the remark74. Foldit Terms of Service and Consent, supra note 55.
75. Joint inventorship is legally determined under patent law based on involvement
by each inventor in the original conception of the invention. See, e.g., Burroughs Wellcome
Co. v. Barr Labs. Inc., 40 F.3d 1223, 1227 (Fed. Cir. 1994). The contributor must have in
mind and be contributing to the idea of the complete and operative invention, which would
require knowledge of the entire invention and not just an understanding that some kind of
discovery is likely to result. Reducing the invention to practice, while a valuable
contribution, is not an inventorship contribution. Id. at 1228. While each joint inventor
must contribute to the conception of the invention, it is enough to contribute to one claim.
35 U.S.C. § 116 (2012). While there must be some element of joint behavior involved, a
loose collaboration, “working under common direction,” or even “one inventor seeing a
relevant report and building upon it,” may be enough. Kimberly-Clark Corp. v. Procter &
Gamble Distrib. Co., 973 F.2d 911, 916-17 (Fed. Cir. 1992). If joint inventorship has
occurred, the U.S. patent application must name all of the inventors. In the absence of any
agreement to the contrary, the joint inventors are joint owners of the patent and can
independently exploit the benefits of the patent. 35 U.S.C. § 262 (2012).
76. For an interesting discussion of some of the problems caused by this disconnect
between common understandings and legal definitions of joint inventorship, see Aaron
Fellmeth, Conception and Misconception in Joint Inventorship, 2 N.Y.U. J. INTELL. PROP. &
ENT. L. 73, 77-78 (2012).
77. See, e.g., Merges, supra note 68 (noting the tensions that patents create in
academic science).
1084
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
able features of the Foldit community is the willingness of game
players to share the recipes that they develop to solve protein structure puzzles.78 This sharing helps to sustain the evolvers, who copy
and build on the game strategies used by others, and also aids the
general population of game players. These recipes function like research tools, serving as critical inputs into the processes that lead to
discoveries about protein folding without describing such discoveries
or being independently patentable.
The ability to patent a discovery that may have commercial value
will make participants less willing to share their game strategies for
a number of reasons. First, participants will be less willing to reveal
game strategies, including the software code that enables these
strategies, if there is a chance that a third party will take this information and use it to make proprietary discoveries that are not in
turn shared with the group. The nature of the information that is
shared in the exchange of recipes is unlikely to give these participants any claim as a joint inventor and is unlikely to constitute prior
art, leaving them with little ability to either block the patent or share
in its control.
Second, patenting will shift attention from the intrinsic motivation of obtaining a high score to the extrinsic motivation of obtaining
commercially valuable intellectual property. Those who have contributed significant ideas that are not patentable may have to sit by
and watch as their contributions lead to commercially lucrative patentable inventions that benefit others. The rules around coauthorship and the ability to be expansive in attributing coauthorship at low cost make joint authorship a useful way of respecting group contributions. The approach does not translate well into
situations of joint inventorship, since contributions will typically take
the form of tools that make discovery more likely rather than contributions to the conception of the invention.79 Members may even have
incentives to drop out of the game and seek patent protection for discoveries that they would not have made but for access to the ideas
contributed by other players.
78. As one high-scoring player of Foldit explained during an interview, “I shared BF
[blue fuse tool] fully because Foldit is so much more than a game—the competition is serious and fierce, but we are also trying to improve the understanding of huge biological proteins. We collaborate and compete at the same time.” Researchers Uncover Foldit Gamers’
Strategies, DRUG DISCOVERY & DEV. (Nov. 8, 2011, 11:16 AM), http://www.dddmag.com/
news/2011/11/researchers-uncover-foldit-gamers’-strategies.
79. In the scientific world, journal articles often include a variety of contributors,
including those who have performed experiments or contributed hard work to generate the
results. In contrast, joint inventorship is legally determined under patent law based on
involvement by each inventor in the original conception of the invention. For further discussion
of joint inventorship, see supra notes 75-76.
2014]
COOPERATIVE INNOVATION
1085
Third, if and when Foldit moves into areas where it competes with
commercial players, particularly if it partners with pharmaceutical
companies or other private companies, there will be stronger pressures on the administrators of the Foldit game to limit the public use
of the discoveries generated. Where discoveries have identifiable
commercial applications, the financial incentives to limit information
sharing and to restrict use of the discoveries will be larger. In addition, if commercial partners play a bigger role in the game, the very
different norms and rules governing commercial drug development
will influence the ways in which the game is organized and the results are shared.
While these harms remain primarily prospective for now, the risks
that market-based incentives may crowd out non-market mechanisms of cooperation have been well documented in other similar areas and should not be ignored.80 Crowding out occurs when one approach to incentivizing participation has a negative impact on another.81 It can occur in a variety of different ways, including internal
displacement of intrinsic motivations and disruptions of the system
through the actions and pressures of participants from outside of the
system.82 In projects like Foldit, the social value of contributing to the
project and the enjoyment derived from playing games in an open
science context, as well as the commitment of Foldit administrators
to open science, may eventually be overshadowed by financial incen-
80. As described by Yochai Benkler, “[w]e have now . . . almost two decades of literature in experimental economics, game theory, anthropology, political science field studies,
that shows that cooperation in fact does happen much more often than the standard economics textbooks predict, and that under certain structural conditions non-price-based
production is extraordinarily robust. The same literature also suggests that there is crowding-out, or displacement, between monetary and non-monetary motivations as well as between different institutional systems: [sic] social, as opposed to market, as opposed to
state.” Yochai Benkler, Comment to Calacanis’s Wallet and the Web 2.0 Dream, ROUGH
TYPE (July 28, 2006, 11:22 AM), http://www.roughtype.com/?p=466; see also Yochai
Benkler, Coase’s Penguin, or, Linux and The Nature of the Firm, 112 YALE L.J. 369 (2002)
[hereinafter Benkler, Coase’s Penguin].
81. “Crowding out, or the non-separability of social preferences from the introduction
of explicit extrinsic motivation, poses a systemic challenge to using traditional, incentivesbased mechanisms, both private and public, for eliciting desirable behavior.” Benkler,
supra note 1, at 307; see also Samuel Bowles, Policies Designed for Self-Interested Citizens
May Undermine “The Moral Sentiments”: Evidence from Economic Experiments, 320 SCI.
1605 (2008). One of the most widely debated examples of crowding out is the use of
payments to encourage blood donations, which pits an economic framework for blood
donations against a system of altruistic unpaid donors. See, e.g., RICHARD M. TITMUSS, THE
GIFT RELATIONSHIP 158-59 (1971); Alena M. Buyx, Blood Donation, Payment, and NonCash Incentives: Classical Questions Drawing Renewed Interest, 36 TRANSFUSION MED. &
HEMOTHERAPY 329, 329-30 (2009) (exploring strategies for creating well-designed non-cash
incentives which cut across the rigid dichotomy of altruistic donation versus payment).
82. See, e.g., OSTROM, supra note 25; SUCCESSFUL MANAGEMENT BY MOTIVATION:
BALANCING INTRINSIC AND EXTRINSIC INCENTIVES (Bruno S. Frey & Margit Osterloh
eds., 2002).
1086
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
tives, such as patent royalties. The financial incentives might thus
crowd out social motivations. This kind of displacement poses an undeniable risk for open source, volunteer-driven models of innovation
like Foldit.83 Sustaining the kinds of non-market mechanisms that
support Foldit, including the important mechanisms of trust, benefitsharing, and reciprocity, will undoubtedly become harder in the presence of patents.84 Ironically, the more successful the non-market
mechanisms are in generating inventions, the more likely it is that
they will come under challenge.
2. Open Source Software: The Example of Linux
“No one ‘owns’ the software in the traditional sense . . . . The result
has been the emergence of a vibrant, innovative and productive collaboration, whose participants are not organized in firms and do
not choose their projects in response to price signals.”—Yochai
Benkler 85
The second example is drawn from open source software.86 Open
source software projects have proven to be a significant economic and
social phenomenon, particularly in the context of software development.87 Sourceforge, which is one of the main internet sites hosting
open source software projects, lists more than 4.8 million daily downloads, more than 430,000 open source software projects involving
more than 3.7 million developers. Its directory connects more than
41.8 million consumers with these open source projects.88 While most
of these projects are small, some, such as the Linux operating system, the Apache web server software, the MySQL database, and the
Firefox web browser, are massive and compete with established proprietary software products. The open source smartphone operating
83. For discussions of intrinsic versus extrinsic motivations and the crowding-out effects
of carrot and stick approaches, see DANIEL H. PINK, DRIVE: THE SURPRISING TRUTH ABOUT
WHAT MOTIVATES US 3-4, 8-9, 17-21 (2009).
84. As noted earlier, there are many reasons why people become involved in projects like
Foldit, and the three mechanisms described here are meant simply to capture the ways in
which these cooperative systems sustain the involvement of these diverse participants with
their diverse motivations and interests.
85. Benkler, Coase’s Penguin, supra note 80.
86. Much has been written about the free and open source software movement. For a
good overview from an intellectual property perspective, see Vetter, supra note 26.
87. See, e.g., Eric von Hippel & Georg von Krogh, Open Source Software and the “Private-Collective” Innovation Model: Issues for Organization Science, 14 ORG. SCI. 209, 209
(2003) (discussing history of open source software and its emergence as major cultural and
economic phenomenon); see also ERIC S. RAYMOND, THE CATHEDRAL AND THE BAZAAR:
MUSINGS ON LINUX AND OPEN SOURCE BY AN ACCIDENTAL REVOLUTIONARY (1999).
88. SOURCEFORGE, http://sourceforge.net/about (last visited Mar. 22, 2014).
2014]
COOPERATIVE INNOVATION
1087
system Android, for example, which is closely tied to Linux, was estimated to have seventy percent of the smartphone market in 2012.89
Open source software projects deviate from private proprietary
models of software development in at least two important ways.90
First, most truly open source software projects are fueled at least in
part by software developers who are primarily intrinsically rather
than extrinsically motivated to participate. People contribute to the
project because they want to, whether to solve their own problems,
contribute to a community that they have benefited from in the past,
gain a reputation, or as an outlet for creativity.91 Second, participants
freely reveal the software that they have developed in ways that allow other participants not only to use it, but also to modify and build
upon it.92 These characteristics are combined with organizational innovations that allow people to contribute to the software project in a
massively distributed, decentralized way, and with very low transaction costs. These open source software systems show how valuable
products can be developed in systems that are based upon the free
contribution and sharing of ideas.93
While a few open source software systems, such as Linux and its
stepchild Android, have been able to compete with proprietary products, the costs of doing so in the presence of an increasing number of
software patents have been large. These costs include millions of dollars spent by Linux and Android users in obtaining patents solely for
defensive purposes, the development and refinement of defensive patent pooling and licensing organizations and strategies, time spent
trying to invalidate patents that might impede use of Linux, and
foreclosure of potentially valuable development paths for Linux
89. See, e.g., Nirav Patel, Apple’s iPhone Is Losing Market Share as Android Gets
70% of the Smartphone Market, GADGET MASTERS (Jan. 28, 2013), http://www.thegadget
masters.com/2013/01/28/apple-iphone-is-losing-market-share-as-android-gets-70-of-the-sma
rtphone-market/.
90. For an overview of open source software and its implications and intersections
with intellectual property, see, for example, Maurer & Scotchmer, supra note 15.
91. See, e.g., Jürgen Bitzer et al., Intrinsic Motivation in Open Source Software Development, 35 J. COMP. ECON. 160, 162 (2007); Georg von Krogh et al., Carrots and Rainbows:
Motivation and Social Practice in Open Source Software Development, 36 MIS Q. 649, 652
(2012); Chorng-Guang Wu et al., An Empirical Analysis of Open Source Software Developers’ Motivations and Continuance Intentions, 44 INFO. & MGMT. 253, 254-55 (2007). But see,
e.g., Sharon Belenzon & Mark Schankerman, Motivation and Sorting in Open Source Software Innovation (Nov. 5, 2012) (unpublished manuscript), available at https://faculty.
fuqua.duke.edu/~sb135/bio/Belenzon%20Schankerman%20OSS%20July%202012.pdf (examining observed pattern of contributions to open source software projects to extract revealed preferences of developers and illustrating the importance of understanding incentives heterogeneity in open source contributions).
92. See von Hippel & von Krogh, supra note 87 (describing open source software
development systems, referring to them as illustrations of a private-collective model of
innovation that occupies the middle ground between private investment and collective action).
93. Id.
1088
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
where such paths fall outside of the protective boundaries established
by large corporate users of open source.94 Moreover, other promising
open source systems switch course or even shut down when confronted with real or imagined patent threats.95 I focus on the story of
Linux and the patent shadows created by Microsoft’s patents for two
reasons. First, this story illustrates the susceptibility of even the
most successful open source models to patent threats. Second, it
demonstrates the difficulties that open source software supporters
face in reducing patent threats.96
Linux was one of the pioneering free and open source software
projects. It has its roots in the intertwined ideologies and development models generated by early free software and open source software movements, which are collectively referred to as Free and Open
Source Software (FOSS).97 At a general level, FOSS is software which
is made freely available, in both object and source code form, for anyone to use, copy, modify, and distribute, thus enabling people to voluntarily improve the design of the software.98 On a closer look, however, FOSS refers both to an ideology about the freedom to use, modify, and share this resource, which lies at the core of the Free Software Movement, and to a methodology for peer-to-peer development,
which is the focus of the Open Source Movement.99
94. For a discussion of some of these costs, see, for example, Deborah Nicholson, Open
Invention Network: A Defensive Patent Pool for Open Source Projects and Businesses, 2012
TECH. INNOVATION MGMT. REV. 12. For a discussion of recent efforts by Microsoft to raise the
costs for a number of Android manufacturers through threats of patent infringement suits and
licensing deals with large royalties and very restrictive licensing terms, see, for example,
Michele Boldrin & David K. Levine, The Case Against Patents (Fed. Reserve Bank of St. Louis,
Working Paper No. 2012-035A, 2012), available at http://research.stlouisfed.org/wp/2012/2012035.pdf (discussing impact of patents in Android context); Tom Warren, Barnes & Noble
“Exposes” Microsoft’s Android Patent Fees and Strategy, WINRUMORS (Nov. 14, 2011, 6:53 PM),
http://www.winrumors.com/barnes-noble-exposes-microsofts-android-patent-fees-and-strategy/.
95. See, e.g., Nicholson, supra note 94, at 14 (discussing chilling effect of potential
patent threats on small companies and smaller projects even when no actual patent is
asserted); Wen Wen et al., Patent Commons, Thickets, and Open Source Software Entry by
Start-Up Firms, (Nat’l Bureau of Econ. Res., Working Paper No. 19394, 2013), available at
http://www.nber.org/papers/w19394 (empirical study showing that rate of entry of start-up
software firms increases with the size of open source software communities).
96. For an example of the power of open source software models as a substitute for
traditional standard-setting approaches, see Greg R. Vetter, Open Source Licensing and
Scattering Opportunism in Software Standards, 48 B.C. L. REV. 225 (2007).
97. See, e.g., Vetter, supra note 26; see also WOLFGANG LEISTER ET AL., OPEN SOURCE,
OPEN COLLABORATION AND INNOVATION 15-26, 47-78 (Wolfgang Leister & Nils
Christophersen eds., 2012).
98. See, e.g., The Open Source Definition, OPEN SOURCE INITIATIVE,
http://opensource.org/osd (last visited Mar. 22, 2014); What Is Free Software and Why Is It
So Important for Society?, FREE SOFTWARE FOUND., http://www.fsf.org/about/what-is-freesoftware (last visited Mar. 22, 2014).
99. For a related discussion of the causal factors underlying FOSS, see Greg R. Vetter,
Exit and Voice in Free and Open Source Software Licensing: Moderating the Rein over Software
Users, 85 OR. L. REV. 183 (2006). For the Free Software Movement’s description of how they
2014]
COOPERATIVE INNOVATION
1089
While these two camps share many features, they have different
views about what FOSS requires—disparities that quickly become
apparent in their responses to proprietary uses of software.100 The
Free Software Movement believes that software should be free. To
promote this goal and to protect the free use of software, it advocates
that uses of and improvements to open source software should be
made available on open source terms. Beliefs about freedom, fairness,
and reciprocity lie at the center of the free software movement. The
Open Source Software movement is less worried about proprietary
software and proprietary use of FOSS, viewing open source more as a
development methodology and less as a social movement.101 Both
groups share a reliance on swift trust and reciprocity, however, and
beliefs about the value of open source software help to sustain norms
of open access and sharing that are critical to open source software.102
The entanglement of ideology and development methodology is important in explaining both the strengths and the vulnerabilities of
even large and successful open source projects like Linux.103
Linux was developed pursuant to a unique collaborative development project,104 and it remains one of the largest systems of collaborative development in the history of computing. As described by Eric
Raymond in his seminal article The Cathedral and the Bazaar, “Who
would have thought even five years ago (1991) that a world-class operating system could coalesce as if by magic out of part-time hacking
by several thousand developers scattered all over the planet, connected only by the tenuous strands of the Internet?”105 It encapsudiffer from the Open Source Movement, see Why “Free Software” Is Better Than Open Source,
GNU OPERATING SYS., https://www.gnu.org/philosophy/free-software-for-freedom.html (last
visited Mar. 22, 2014) (noting that the movements “disagree on the basic principles, but agree
more or less on the practical recommendations”).
100. See, e.g., Richard Stallman, Free Software as a Social Movement, ZNET (Dec. 18, 2005),
http://www.zcommunications.org/free-software-as-a-social-movement-by-richard-stallman.
101. See, e.g., Anne Barron, Free Software Production as Critical Social Practice, 42
ECON. & SOC’Y 597 (2013) (describing the contrast between the ideological approach of the
free software movement and the pragmatic approach of the subsequent open source
software movement).
102. See, e.g., Osterloh & Rota, supra note 27; Osterloh, Rota & Kuster, supra note 31,
at 2 (“The fragile balance between intrinsically and extrinsically motivated developers may
be disturbed by the entrance of extrinsically motivated commercial firms into the world of
open source.”).
103. See, e.g., Vetter, supra note 99 (exploring important aspects of this entanglement
between ideology and methodology).
104. See, e.g., LINUX FOUNDATION, http://www.linuxfoundation.org (last visited Mar. 22,
2014) (describing how Linux is built).
105. See Eric Steven Raymond, The Cathedral and the Bazaar, http://www.catb.org/~esr/
writings/cathedral-bazaar/cathedral-bazaar/index.html (last updated Aug. 2, 2002)
(contrasting two forms of software development, the hierarchical cathedral approach used in
proprietary software and the decentralized, open, Internet-based development style used
by Linux).
1090
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
lates the powerful idea of software as a modular and communal Internet-based effort.106 The source code of Linux is made available
without charge, and with the ability to freely modify and use this
code, pursuant to version two of a well-known free and open source
copyright license called the General Public License (GPL).107 The licensing model is a legal innovation, what has been referred to as legal jujitsu, designed to use intellectual property rights to preserve
the open source nature of the project.108 The GPL embodies principles
of open access, free sharing, and reciprocity in its license restrictions
and contractual obligations as a way of preserving and expanding the
domain of open source software. While open source licensing strategies
such as this one have helped to sustain open source efforts, the need for
additional strategies to preserve open source has been widely acknowledged by proponents of open source.109 In the case of Linux, backing by
large corporate players was an essential part of its survival.
Linux rapidly gained popularity as an alternative to Microsoft’s
proprietary software operating system. In response, Microsoft began
a concerted strategy to disadvantage Linux by encouraging companies to trade open source participation for financial gain and by getting the users of Linux to pay royalties to Microsoft. This included a
controversial joint patent agreement between Microsoft and a company called Novell, Inc., a software company that had initially set
itself up to compete with Microsoft using Linux-based open source
106. See, e.g., LEISTER ET AL., supra note 97, at 47-78.
107. The GPL provides that software licensed under the GPL can be freely used, copied
and modified, provided that any modifications and improvements to this software are also
made available to the public in source code form under the same license terms. The first
two versions of the GPL focused primarily on copyright, but a third version has been
developed to address the significant challenges that software patents create for the open
source process. See GNU General Public License, Version 2, GNU OPERATING SYS. (June,
1991), http://www.gnu.org/licenses/gpl-2.0.html; GNU General Public License, Version 3,
GNU OPERATING SYS. (June 29, 2007), http://www.gnu.org/licenses/gpl.html; see also
Robert W. Gomulkiewicz, A First Look at General Public License 3.0, 24 COMPUTER &
INTERNET L. 15, 15 (2007). For critiques of GPL version 3 see, for example, James E.J.
Bottomley et al., Kernel Developers’ Position on GPLv3: The Dangers and Problems with
GPLv3, LWN.NET (Sept. 15, 2006), http://lwn.net/Articles/200422/; John Carroll, The Crux
of the GPL Problem, ZDNET (June 4, 2007, 7:42 AM), http://www.zdnet.com/blog/
carroll/the-crux-of-the-gpl-problem/1707.
108. The GNU/Linux computer operating system includes the Linux kernel, an open
source project initiated by Linus Torvalds in 1991, and GNU software emerging from the free
software movement founded by Richard Stallman in 1985. Richard Stallman, Linux and the
GNU System, GNU OPERATING SYS., http://www.gnu.org/gnu/linux-and-gnu.html (last updated
Apr. 12, 2014). I focus here on the development of the Linux kernel, although the issues extend
to the broader GNU/Linux operating system.
109. See, e.g., Clark D. Asay, A Case for the Public Domain, 74 OHIO ST. L.J. 753
(2013); Jason Schultz & Jennifer M. Urban, Protecting Open Innovation: The Defensive
Patent License as a New Approach to Patent Threats, Transaction Costs, and Tactical Disarmament, 26 HARV. J.L. & TECH. 1 (2012); Greg R. Vetter, A Public Domain Approach to
Free and Open Source Software?, 75 OHIO ST. L.J. FURTHERMORE (forthcoming 2014).
2014]
COOPERATIVE INNOVATION
1091
software products.110 Novell received large payments pursuant to its
deal with Microsoft, and when the deal ended, a technology consortium led by Microsoft acquired key intellectual property assets from
Novell for $450 million.111 Members of the FOSS community criticized
Novell for what they saw as a defection from group norms and a compromise of the interests of the free software community. Novell was
seen as free-riding on the value created by the open source community and capitalizing on the holdup value of patents covering inventions
used by the open source community. More importantly, the deal was
seen as providing unwarranted legitimacy for patent claims made
against Linux, creating a cloud of uncertainty and fear for users of
Linux and deterring open source developers.112
Shortly after the first agreement with Novell was signed, Microsoft began making public claims that Linux violated more than
200 of Microsoft’s patents—no fewer than 235 according to Microsoft’s general counsel at that time.113 The fact that the Linux
source code is freely available allows patent holders to scrutinize the
code for areas that might implicate their patents. The fact that the
system involves combining and building on incremental contributions
means that patents covering even small parts of the system may
have tremendous holdup power. Since its initial claims of infringement, Microsoft has focused on systematically securing patent licensing deals from Linux users, although it has also relied on patent infringement litigation against select users of Linux.114 Patent assertion entities such as IP Innovations, a subsidiary of Acacia Technologies, have also brought patent suits against prominent users of Linux
such as Red Hat. Some speculate that Microsoft backed the Acacia
litigation against Red Hat.115 Red Hat, which provides services based
110. See, e.g., Elizabeth Montalbano, Update: Microsoft, Novell Strike Linux Deal,
COMPUTERWORLD (Nov. 2, 2006, 12:00 PM), http://www.computerworld.com/s/article/
9004723/Update_Microsoft_Novell_strike_Linux_deal; Paul Krill, The Microsoft-Novell Linux
Deal: Two Years Later, INFOWORLD (Nov. 18, 2008), http://www.infoworld.com/d/opensource/microsoft-novell-linux-deal-two-years-later-858.
111. See sources cited supra note 104.
112. See, e.g., Katherine Noyes, Microsoft’s Hand in Novell Deal Bodes Ill for Linux,
PCWORLD (Nov. 22, 2010, 3:12 PM), http://www.pcworld.com/article/211414/microsofts_
hand_in_novell_deal_bodes_ill_for_linux.html; TECHRIGHTS, http://www.techrights.org/home/
(last visited Mar. 22, 2014) (showing Roy Schestowitz is editor of the site).
113. See, e.g., Mary Jo Foley, Microsoft: Free and Open Source Software Violates 235
Microsoft Patents, ZDNET (May 13, 2007, 4:46 PM), http://www.zdnet.com/blog/microsoft/
microsoft-free-and-open-source-software-violates-235-microsoft-patents/436; Peter Lattman,
Patent Litigation’s Battle Royale: Microsoft v. Open Source, WALL ST. J., LAW BLOG (May
15, 2007. 9:21 AM), http://blogs.wsj.com/law/2007/05/15/patent-litigations-battle-royalemicrosoft-v-open-source/.
114. See, e.g., John C. Dvorak, Microsoft’s Nuisance Suit Strategy, PC MAG. (Sept. 21,
2011), http://www.pcmag.com/article2/0,2817,2393361,00.asp.
115. See, e.g., J. Nicholas Hoover, Acacia Research, Linux Patent Adversary, Has Long
Litigation History, INFORMATIONWEEK (Oct. 12, 2007, 4:32 PM), http://www.information
1092
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
on Linux, is often heralded as one of the open source software company success stories, making it a natural target for opponents of
Linux. The company complains of having to routinely address “attempts to impede the innovative forces of open source via allegations
of patent infringement.”116 General and widespread threats like those
made by Microsoft increase the real and perceived risks and the
transaction costs of contributing to and using open source software
for all participants in the open source system, with the largest impact
on developers and users who are not affiliated with and protected by
large companies. Even small changes in cost and risk may deter
many users of the resulting software, chill development efforts, and
alter otherwise promising development paths.
The availability of software patents increased the commercial attractiveness of defecting from the FOSS community for Novell and
increased the risks of third party patent infringement suits against
community members.117 These risks ultimately forced community
members to rely heavily on the resources of large commercial participants such as IBM to protect Linux through defensive patenting
strategies. Linux survived despite patent risks and costs largely by
becoming an established part of the business model of major participants in the software industry.118 By some estimates more than half
of the companies in the Fortune 500 are using Linux in their data
centers.119 These large corporate users have invested significant resources both in the development and the protection of the Linux project.120 In response to patent infringement threats made by Microsoft
against Linux, large corporate users of Linux—i.e. IBM, NEC, Novell,
Phillips, Red Hat, and Sony—set up the Open Invention Network
week.com/acacia-research-linux-patent-adversary-has-long-litigation-history/d/d-id/106 0255?;
Roy Schestowitz, Red Hat Pays Microsoft-Linked Patent Troll Again, Refuses to Provide
Details, TECHRIGHTS (May 25, 2013, 7:16 AM), http://techrights.org/2013/05/25/acacia-red-hat/.
116. See, e.g., Soulskill, Red Hat Settles Patent Case, SLASHDOT (Oct. 4, 2010, 6:45 PM),
http://linux.slashdot.org/story/10/10/04/2148218/red-hat-settles-patent-case?sdsrc=rel.
117. See, e.g., Krill, supra note 110.
118. See, e.g., Oliver Alexy & Markus Reitzig, Private-Collective Innovation, Competition,
and Firms’ Counterintuitive Appropriation Strategies, 42 RES. POL’Y 895 (2013) (examining
role of exclusion rights for technology in competition between private-collective and other
innovators); Joel West, How Open Is Open Enough? Melding Proprietary and Open Source
Platform Strategies, 32 RES. POL’Y 1259 (2003) (examining hybrid strategies and contrasting
them with purely open and purely proprietary software alternatives).
119. Parloff, supra note 11.
120. See, e.g., Dan Woods, Can Intel Heal the Hadoop Open Source Ecosystem?, FORBES
(Feb. 26, 2013, 9:22 AM), http://www.forbes.com/sites/danwoods/2013/02/26/can-intel-heal-thehadoop-open-source-ecosystem/2/ (“In the Linux community the primary contributors are those
who benefit from using Linux in their businesses. IBM, Intel, Google, HP, Oracle all make a
pile of money because Linux solves a variety of problems for them. The amount of value that
they receive from this use dwarfs that captured by Red Hat or Suse, the Linux distributors.
Linux thrives because the big players take part of the massive revenue from the use value and
invest heavily in large development teams.”).
2014]
COOPERATIVE INNOVATION
1093
(OIN) in 2005 to acquire a portfolio of patents that could create problems for companies like Microsoft should they create problems for
Linux users.121 OIN is self-described as “an intellectual property company that was formed to promote the Linux system by using patents to
create a collaborative ecosystem.”122 “Patents owned by OIN are available royalty-free to any [entity] or individual that agrees not to assert
its patents against [the] Linux” system.123 In addition, Open Source
Development Labs, which is the consortium that promotes and coordinates Linux development, established its own patent commons to accept donations of rights to use patents.124 The Patent Commons Project
is self-described as creating an area of safety, “a preserve where developers and users of software can innovate, collaborate, and access patent resources in an environment of enhanced safety, protected by
pledges of support made by holders of software patents.”125
These patent pooling efforts can be seen as a way of trying to preserve the trust and reciprocity on which open source systems depend.
Compliance with group norms is achieved by providing a patent
shield for members who adhere to group norms. Reciprocity is enforced through contingent protections that are available only to participants who do not themselves assert their patents against other
group members or protected open source projects. Other community
responses to the patent threats have included Linux Defenders, an
online clearinghouse for prior art designed to invalidate poor-quality
software patents,126 and a project called A Patent a Day, with its goal
of identifying one Microsoft-owned patent every day that Linux potentially infringes with the goal of helping to get rid of the dependency on these patents and/or get rid of the patents.127 These efforts illustrate the ways in which open source communities must participate
in the patent system simply as a way of protecting their nonproprietary software systems from extinction.
Such efforts come with a very steep price tag both in terms of cost
and in terms of increasingly centralized control over open source.
Monitoring and acquiring patents are ironic ways for an open source
121. Nicholson, supra note 94, at 16; see also OPEN INVENTION NETWORK,
http://www.openinventionnetwork.com (last visited Mar. 22, 2014).
122. Nicholson, supra note 94, at 16.
123. OPEN INVENTION NETWORK, supra note 121.
124. See, e.g., Steve Hamm, Linux Marches On, BLOOMBERG BUSINESSWEEK (Nov. 16,
2005),
http://www.businessweek.com/the_thread/techbeat/archives/2005/11/linux_marches_
o.html; see also PATENT COMMONS, http://www.patentcommons.org (last visited Mar. 22, 2014).
125. PATENT COMMONS, supra note 124.
126. LINUX DEFENDERS, http://linuxdefenders.org (last visited Mar. 22, 2014).
127. See PatentSleuth, Introduction Post: A Patent a Day Project, LINUX PATENTS (July
4, 2009, 8:26 AM), http://linuxpatents.blogspot.com/2009/07/introduction-post-patent-dayproject.html.
1094
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
community to spend its funds. Reliance on defensive patenting and
pooling strategies also changes the balance of power within the open
source community, giving corporate participants with large patent
holdings control over the directions that open source projects take.
One of the largest corporate supporters of Linux, IBM, also regularly
tops the list of patents filed at the U.S. Patent and Trademark Office.128 Corporate patent holders like IBM have significant power in
shaping the boundaries of protected open source software spaces,
aligning the boundaries with their own private business interests
and strategies. In addition to favoring some open source software projects but not others, commercial partners of the open source community tend to favor hybrid models in which proprietary and open source
software development coexists.129 While not everybody sees this shift
towards hybrid models as a problem,130 many members of the open
source software community are skeptical that a balance of closed and
open source software will survive.131 Their concerns include the negative impact of hybrid models on the motivations supporting open
source software communities, the shift in control over development
paths for software that impacts private business interests, the high
direct and indirect costs of defensive patents strategies that pervade
this hybrid world, and the threats that patents will continue to pose
for open source models not backed by large corporate users.132
Thus, while private orderings of intellectual property rights such
as defensive patent pooling may serve to mitigate patent threats for
open source software, they do so in a limited and costly way. Patents
continue to challenge the viability of FOSS systems that rely on
128. See, e.g., IBM Earns Most U.S. Patents for 17th Consecutive Year; Will Offer
Licenses to Patent Portfolio Management Know-How, IBM (Jan. 12, 2010), http://www03.ibm.com/press/us/en/pressrelease/29168.wss.
129. See, e.g., Open-Source Software: Going Hybrid, ECONOMIST, (July 25, 2002), available
at http://www.economist.com/node/1251254; Rosa Maria Ballardini, Proprietary Software vs.
FOSS: Challenges with Hybrid Protection Models (May 2012) (Ph.D. dissertation, IPR Univ.
Ctr.), available at http://www.iprinfo.com/julkaisut/verkkojulkaisut/ipr-series-b/en_GB/
proprietary-software-vs-foss/_files/88742337887406149/default/B4_Ballardini_eng.pdf.
130. See, e.g., Dreyfuss, supra note 9; Ronald J. Mann, Commercializing Open Source
Software: Do Property Rights Still Matter?, 20 HARV. J.L. & TECH. 1, 3 (2006) (exploring the
symbiotic relationship between proprietary and open source software); Jay P. Kesan, The
Fallacy of OSS Discrimination by FRAND Licensing: An Empirical Analysis (Univ. of Ill.
Coll. of Law Pub. Law & Legal Theory Res. Paper Series, Paper No. 10-14, 2011), available
at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1767083 (examining compatibility of
OSS licenses with proprietary uses).
131. See, e.g., Sonali K. Shah, Motivation, Governance, and the Viability of Hybrid
Forms in Open Source Software Development, 52 MGMT. SCI. 1000 (2006). For a discussion
of hybrid models and the issues that may arise, see, for example, Vetter, supra note 26.
132. See, e.g., Arnold Polanski, Is the General Public License a Rational Choice?, 55 J.
INDUS. ECON. 691 (2007) (showing that proprietary licensing can lead to a holdup problem
which may terminate a sequence of innovation prematurely, and that free open source
licensing may be able to avoid this).
2014]
COOPERATIVE INNOVATION
1095
keeping costs and risks low and adherence to community norms
based on openness, trust, and reciprocity high.133 I suggest that in the
absence of larger patent reforms addressing concerns with software
patents,134 systematic efforts by courts to take into account the costs
of patents in cooperative contexts may help to limit these costs and to
support a greater diversity of open source software projects.
III. PATENT ROADBLOCKS
As cooperative systems such as crowd science and free and open
source software develop, they must confront patent laws that are designed with a very different model of cooperation in mind. In some cases the incentives that patents create may interfere with the mechanisms of trust, benefit-sharing, and reciprocity on which these and
other cooperative innovation systems rely, increasing collective action
problems. In other cases, patent laws may help to solve collective action problems. Fashioning the right policy response requires closer
scrutiny of the relationship between patent laws and intellectual production when private incentives and public interests diverge.135
Patents may impede the collective action needed to sustain cooperative innovation in two different ways. First, patents increase the
incentives for individual members of a cooperative system to defect
from group norms of open access and sharing. Where a member of a
group has the opportunity to patent and privately exploit an invention that is useful to the group, this member will be less willing to
share her ideas with the group and may be more likely to free ride on
133. For studies of the survival factors of open source software, see, for example, Kevin
Crowston et al., Free/Libre Open-Source Software Development: What We Know and What
We Do Not Know, 44 ACM COMPUTING SURV. 7:1 (2012); Vishal Midha & Prashant Palvia,
Factors Affecting the Success of Open Source Software, 85 J. SYS. & SOFTWARE 895 (2012);
Jing Wang, Survival Factors for Free Open Source Software Projects: A Multi-Stage Perspective, 30 EUR. MGMT. J. 352 (2012).
134. There is a rich literature identifying and responding to problems with software
patents. See, e.g., Kevin Emerson Collins, Patent Law’s Functionality Malfunction and
the Problem of Overbroad, Functional Software Patents, 90 WASH. U. L. REV. 1399 (2013);
Mark A. Lemley, Software Patents and the Return of Functional Claiming, 2013 WIS. L.
REV. 905 (2013); James Bessen, A Generation of Software Patents, (Bos. Univ. Sch. of Law,
Working Paper No. 11-31, 2011) (critically examining software patenting practices and impact
on social welfare).
135. For use of game-theoretic models to explore opportunities for using patents to
improve cooperative outcomes, see, for example, Shubha Ghosh, Patent Law and the Assurance
Game: Refitting Intellectual Property in the Box of Regulation, 18 CAN. J.L. & JURISPRUDENCE
307 (2005) (referencing the assurance game as an alternative way of thinking about the role of
patent law in regulating innovation). Under the assurance game, more commonly known as
the stag hunt, two hunters can jointly hunt a stag for high payoffs or individually hunt rabbits
for smaller payoffs. If either hunts a stag alone, the chance of success is minimal. Hunting
stags is most beneficial for the group but requires significant trust among its members. See
also Ted Sichelman, Quantum Game Theory and Coordination in Intellectual Property (San
Diego Legal Studs., Paper No. 10-035, 2010).
1096
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
and exploit the ideas generated by the group. Additionally, a member
may be driven against her interest to seek patent protection for protective reasons, fearing that others will appropriate the benefits from her
contribution without sharing in return. Patents thus increase the individual payoffs from defecting from group norms and may reduce the
payoffs from adhering to group norms. As a result, members of the
group will have reduced expectations that other members will behave
in accordance with group norms of open and free sharing, further
weakening these norms. Members of the group might also place less
value on non-market rewards within the group such as reputation.
Second, patents increase the risks to group members posed by
third parties. The open and transparent nature of intellectual production in contexts of cooperative innovation makes it easy for third
parties to assert patents against the group, and where third parties
have patents covering technologies that are widely used by the group
the third parties will have the ability to hold up the group and collect
more than the incremental value of their technological contribution.
Third parties may also be able to free ride on the intellectual production of the group without any reciprocating contributions. Even generalized threats of third party litigation increase transaction costs for
developers and users of open source software by forcing them into
defensive patent licensing and pooling arrangements which need to
be maintained and enforced. This requires them to spend time examining potential patent risks and planning for potential patent infringement claims, and in some cases induces them to enter licenses
and make royalty payments simply as a way of averting threats from
aggressive patent holders.136 Cooperative systems are particularly
sensitive to these kinds of transaction costs, since they rely on large
numbers of volunteers freely sharing ideas. Even small changes in
the cost of participation—such as the need to reserve funds to respond to cease-and-desist letters from third parties, the learning
costs associated with participating in defensive patent pools, the administrative costs associated with making and using software covered by a growing family of different open source software license
agreements, and small changes in the risk of participation—may
therefore interfere with the continuing operations of the group. In
some cases, these groups survive only by aligning the activities of the
group with the private interests of one or a few large players with
deep pockets and vested interests in the innovation system. By increasing defections from group norms, and by increasing threats from
136. See, e.g., Schultz & Urban, supra note 109 (surveying options and some of the costs
involved with these options); see also Nicholson, supra note 94, at 14 (discussing some of the
transaction costs and chilling effects of patents for open source software participants).
2014]
COOPERATIVE INNOVATION
1097
outside the groups, patents may sometimes threaten or distort valuable processes of cooperative innovation.
The disconnect between the needs of cooperative innovation and
the incentives that patents create can be explained in part by patent
law’s attachment to an outdated model of innovation. The U.S. patent
statute, as well as the broader international patent law framework
now in place, is anchored on a paradigm of market-based, producerdriven innovation.137 As a result, the patent statute and implementing laws and regulations deal primarily with the creation, definition,
and enforcement of ownership rights over inventions and the administration of this process.138 The right of the patent owner to exclude
others from use of the invention forms the backbone of patent law,
and most of the statute is devoted to defining and policing these
rights. Patent remedies are designed to restore the patent owner to
the status quo before infringement through the award of reasonable
royalties, lost profits resulting from lost sales or price erosion,139
and/or injunctive relief.140 The statute pays much less attention to the
source and nature of the inputs leading to invention,141 and to how
the subsequent patents are managed and used,142 or, even more importantly, not used.143
Patent law does address issues of team production, but it does so
largely with a focus on identifying the true inventors and protecting
their collective rights through rules governing derivation and joint
137. See, e.g., Rochelle Cooper Dreyfuss, Commodifying Collaborative Research, in THE
COMMODIFICATION OF INFORMATION 397 (Niva Elkin-Koren & Neil Weinstock Netanel eds.,
2002) (examining myth of the individual inventor and disjuncture between current IP law and
issues of importance to collaborators).
138. See 35 U.S.C. (2012). Part I of the statute deals with the establishment and operation
of the USPTO, and Part II focuses primarily on how to obtain a patent, including requirements
for patentability. Part III focuses on protection of patent rights. Part IV deals with the Patent
Cooperation Treaty, addressing international issues relevant to the Act.
139. Compensation to the patent owner extends even to harm from offers to sell. See, e.g.,
Timothy R. Holbrook, Liability for the “Threat of a Sale”: Assessing Patent Infringement for
Offering To Sell an Invention and Implications for the On-Sale Patentability Bar and Other
Forms of Infringement, 43 SANTA CLARA L. REV. 751, 756-64 (2003).
140. For a description and critique of this private law focus of patent remedies, see
Sichelman, supra note 19.
141. See, e.g., Mark A. Lemley, Point of Novelty, 105 NW. U. L. REV. 1253 (2011); Lemley,
supra note 9; Sean B. Seymore, Rethinking Novelty in Patent Law, 60 DUKE L. J. 919, 929
(2011) (exploring complexity of ensuring that a patent does not issue if the public already
possesses the invention, and joining “a larger project to bridge the disconnect between patent
law and the norms of science”).
142. See, e.g., Herbert Hovenkamp, Innovation and the Domain of Competition Policy, 60
ALA. L. REV. 103 (2008) (discussing rules dealing with patent management confined to defining
what constitutes infringement, remedies, and patent misuse).
143. See, e.g., Oskar Liivak & Eduardo M. Peñalver, The Right Not to Use in Property and
Patent Law (Cornell Law Sch. Legal Studs. Res. Paper Series, Research Paper No. 12-62,
2012), available at http://ssrn.com/abstract=2162667.
1098
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
inventorship. Doctrines of derivation144 and the shared ownership
rules for co-inventors145 protect the rights of those team members who
have either made an invention that is misappropriated or made inventorship contributions to a collective invention. Pursuing derivation or joint inventorship claims is costly, however, and involves evidentiary burdens that may be difficult to satisfy in contexts of massively distributed innovation. In addition, these doctrines do little to
recognize or protect many kinds of valuable contributions to the collective process of intellectual production. To be a joint inventor, an
individual must contribute to the conception of the invention. This
leaves out individuals who make contributions that increase the likelihood of discovery but do not contemplate the discovery and individuals who work hard to reduce the invention to practice. Derivation
proceedings address only situations in which it can be shown that
members of the group actually made an invention that the patent
applicant took without authorization.146 Again, this leaves out many
situations in which the group performs the hard work needed to
make discovery more likely through contributions such as research
tools and ideas about paths not to take, as well as situations in which
the group contributes to the reduction of the invention to practice.
Moreover, both doctrines will be of limited practical relevance for
many cooperative innovation projects that have limited budgets and
144. Derivation refers to situations in which an alleged inventor has derived the claimed
invention from another. See 35 U.S.C. § 135 (2012). Derivation proceedings are costly and
evidence intensive, requiring the challenging inventor(s) to file their own patent application,
file a derivation petition within one year of the first publication of the claimed invention, show
that the invention is “the same or substantially the same” as the earlier claim to the invention,
provided a detailed explanation for claiming unauthorized derivation, and provide substantial
evidence to support the petition. See, e.g., Derivation Proceeding: Overview, AM. INVENTS ACT,
http://www.aiarulemaking.com/rulemaking-topics/group-3/derivation-proceedings.php
(last
visited Mar. 22, 2014). For concerns about the limits of derivation proceedings, see, for
example, N. Scott Pierce, The Effects of the Leahy-Smith America Invents Act on Collaborative
Research, 94 J. PAT. & TRADEMARK OFF. SOC’Y 133 (2012); Joshua D. Sarnoff, Derivation and
Prior Art Problems with the New Patent Act, 2011 PATENTLY-O PAT. L.J. 12; Dennis Crouch,
With 102(f) Eliminated, Is Inventorship Now Codified in 35 U.S.C. 101? Maybe, but not
Restrictions on Patenting Obvious Variants of Derived Information, PATENTLY-O (Oct. 4, 2012),
http://www.patentlyo.com/patent/2012/10/with-102f-eliminated-is-inventorship-now-codifiedin-35-usc-101.html.
145. Where an invention is made by two or more persons jointly, they are considered joint
inventors even if they did not work together or make the same kinds or amount of contribution.
35 U.S.C. § 116 (2006). Unless otherwise agreed by contract, they are joint owners of the
resulting patent and are free to use and authorize others to use the invention without the
consent of or accounting to the other owners and all must join in an infringement suit based on
the patent. 35 U.S.C. § 262 (2006).
146. See supra note 138; see also Gene Quinn, First Inventor to File: USPTO Derivation
Proceedings Go Final, IPWATCHDOG (Sept. 10, 2012, 11:26 AM), http://www.ipwatchdog.com/
2012/09/10/first-inventor-to-file-uspto-derivation-proceedings-go-final/id=27986/
(describing
derivation proceedings).
2014]
COOPERATIVE INNOVATION
1099
a decentralized, volunteer-based project design that is poorly suited
to the pursuit of legal measures.147
In addition to the doctrines relevant to team production discussed
above, patent law has been amended in response to the special needs
of collaborations among different entities, but once again its response
has been largely confined to ensuring that inventions can be patented
by at least one of the collaborators. When different entities collaborate to innovate, patent law has been adjusted to limit the barriers
that sharing information may pose for patenting inventions that
emerge from the collaboration.148
With respect to how subsequent patents are managed and used, or
not used, the focus in patent law remains on the patent owner’s
rights to exclude others from using the patented invention. Little effort has been made to distinguish between socially beneficial and detrimental kinds of unauthorized patent use.149 Defenses to infringement based on special circumstances of creation and use, such as independent discovery or experimental use, remain narrow even after
changes introduced to patent law by the America Invents Act (AIA)
to expand protections for prior inventors.150 Third parties are generally not free to use the patented invention for any purposes, not even
for experimentation or to confirm that the invention actually works
as disclosed in the patent. Concepts of protecting certain kinds of
publicly beneficial uses, such as the fair use found in copyright law,
are missing altogether from the patent statute.151
147. See Aaron X. Fellmeth, Conception and Misconception in Joint Inventorship, 2 N.Y.U.
J. INTELL. PROP. & ENT. L. 73 (2012) (discussing problems with claim fixation in inventorship
and consequences for collaborative research).
148. See Cooperative Research and Technology Enhancement (CREATE) Act of 2004, Pub.
L. No. 108-453, 118 Stat. 3596 (amending Section 103 of the Patent Act). For a discussion of
the CREATE Act, including some potential concerns that it may pose for participants in
collaborations, see Liza S. Vertinsky, Understanding and Applying the CREATE Act in
Collaborations, in 3 AUTM TECHNOLOGY TRANSFER PRACTICE MANUAL (3d ed. 2008); see also
Pierce, supra note 144.
149. This may be changing, albeit in a limited way, as courts and perhaps even Congress
respond to the challenges that patents covering standards essential to critical technologies
have created. See Part IV.A.
150. See, e.g., Lemley, supra note 9 (examining the disconnect between traditional theories
of patent law and real-world experience, particularly in the context of independent invention).
The AIA expands the protection for prior inventors. It provides a “prior use defense” to patent
infringement that protects parties who can establish that they have in good faith commercially
used a product or process covered by a patent at least one year before the earlier of the public
disclosure or the effective filling date of the patent disclosing the invention. Although more
robust than the one it replaced, it includes a number of limitations. See U.S. PAT. &
TRADEMARK OFF., REPORT ON THE PRIOR USER RIGHTS DEFENSE (2012), available at
http://www.uspto.gov/aia_implementation/20120113-pur_report.pdf.
151. See, e.g., Maureen A. O’Rourke, Toward a Doctrine of Fair Use in Patent Law, 100
COLUM. L. REV. 1177 (2000) (arguing that new technology has put pressure on patent laws that
increasingly interfere with follow on innovation, and proposes a doctrine of fair use in patent
law to relieve some of this pressure); Katherine J. Strandburg, Patent Fair Use 2.0, 1 U.C.
1100
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
Members of a community that contribute valuable ideas to the inventor are simply out of luck unless they have the resources and the
ability to show that they are joint inventors, that their invention was
derived from them, or that the patented invention would be obvious
to individual community members in light of the combined
knowledge and effort of the community.152 Determining who contributed what to an invention in a collaborative effort is difficult, particularly where innovation takes the form of cumulative, incremental
contributions by many participants.153 Showing that an invention was
derived from the cumulative, widely-shared discoveries of the group
is likely to be both costly and challenging. The limits that nonobviousness imposes on patentability are also unlikely to provide adequate protection for collective intellectual production.154 Even if a
group can find and present prior art challenging the novelty of a patent, they must overcome the legal presumption that the patent is
valid. This presumption holds regardless of whether the U.S. Patent
and Trademark Office considered the prior art when granting the
patent.155 Patent law, both on the books and as applied, remains too
heavily oriented around the patent ownership rights of the pioneering lone inventors and their assignees. In doing so, patent law neglects the needs of other very different forms of intellectual production such as cooperative innovation.156
This is not to suggest that patents are always or even mostly a
constraint on innovation or that innovation would increase in the ab-
IRVINE L. REV. 265 (2011) (arguing that there should be greater use of defenses and
exemptions to patent infringement to respond to the different contexts in which inventions are
used); see also Lorelei Ritchie de Larena, What Copyright Teaches Patent Law About “Fair Use”
and Why Universities Are Ignoring the Lesson, 84 OR. L. REV. 779 (2005) (arguing that a
doctrine of fair use will help reduce negative externalities and clarify expectations on what
type of infringement is actionable).
152. See 35 U.S.C. § 103 (2012) (addressing conditions for patentability and nonobvious subject matter); supra note 75 (discussing joint inventorship); supra note 144
(discussing derivation).
153. See also Dennis Crouch, Inventorship: Limits of Collaboration, PATENTLY-O (Apr.
1, 2013), http://patentlyo.com/patent/2013/04/inventorship-limits-of-collaboration.html (discussing outcome of Rubin v. Gen. Hosp. Corp., 523 F. App’x 719 (Fed. Cir. 2013)).
154. See, e.g., Shawn P. Miller, Where’s the Innovation: An Analysis of the Quantity and
Qualities of Anticipated and Obvious Patents, 18 VA. J.L. & TECH. 1 (2012) (suggesting patents
are being awarded with negligible innovation value); see also Gregory Mandel, The NonObvious Problem: How the Indeterminate Nonobviousness Standard Produces Excessive Patent
Grants, 42 U.C. DAVIS L. REV. 57 (2008).
155. See, e.g., Sciele Pharma Inc. v. Lupin Ltd., 684 F.3d 1253, 1260 (Fed. Cir. 2012).
156. While the requirement to disclose the invention could be seen as facilitating
sharing, it is at best a limited form of sharing. See Timothy R. Holbrook, Possession in
Patent Law, 59 SMU L. REV. 123, 132-34 (pointing out the disconnect between role of
disclosure as teaching and theory of combating free-riding, suggests that we think of the
function of disclosure as possession).
2014]
COOPERATIVE INNOVATION
1101
sence of patents.157 Patents play important roles in attracting resources and providing market incentives for proprietary, producerdriven innovation.158 Patents can also enable limited departures from
closed, producer-driven systems of innovation by supporting open innovation models in which companies sell or license out unused technologies and acquire or license in third party discoveries.159 Where
patents interfere with cooperation that patent owners are interested
in preserving, the patent owners can create their own contract-based
forms of sharing and pooling patent rights. Patent pools,160 licensing
strategies that incorporate broader public objectives and public uses
into the license terms,161 and patent licensing strategies modeled on
those employed by open source software,162 can and do operate to preserve areas of cooperation.163 Moreover, existing rules and doctrines
such as joint inventorship and the limiting effects of prior art and
derivation on patentability in contexts of cumulative intellectual production mitigate some of the incentive problems that patents might
otherwise create for cooperative innovation.
While these legal tools and private arrangements are helpful in
making room for cooperative innovation, they do not adequately address the harmful incentive effects of patents on non-market mecha157. See, e.g., Jonathan M. Barnett, Intellectual Property as a Law of Organization,
84 S. CAL. L. REV. 785 (2011) [hereinafter Barnett, Law of Organization] (discussing
the role of patents in supporting disaggregation of innovation process); Barnett, The
Illusion of the Commons, supra note 20 (examining the importance of property in
supporting sharing regimes).
158. Strong patent rights may facilitate coordination among many different market
players by reducing transaction costs, solving problems of incomplete contracting, signaling
firm value, or facilitating bargains over use of the patent rights. See, e.g., Ashish Arora &
Robert P. Merges, Specialized Supply Firms, Property Rights and Firm Boundaries, 13
INDUS. & CORP. CHANGE 451 (2004); Barnett, Law of Organization, supra note 157 (patents
as facilitating specialization); Heald, supra note 20 (discussing the role of patents in
solving team production problems and facilitating technology transfer through transaction
cost savings); Kieff, supra note 20 (focusing on the role of patents in facilitating
coordination among many diverse complementary users of an asset in a way that increases
competition and access).
159. See, e.g., Henry Chesbrough, Open Innovation: A New Paradigm for
Understanding Industrial Innovation, in OPEN INNOVATION: RESEARCHING A NEW
PARADIGM (Henry Chesbrough, Wim Vanhaverbeke & Joel West eds., 2006) (defining open
innovation as “a paradigm that assumes that firms can and should use external ideas as
well as internal ideas, and internal and external paths to market, as they look to advance
their technology”); see also Barnett, Law of Organization, supra note 157.
160. See, e.g., Robert P. Merges, Institutions for Intellectual Property Transactions: The
Case of Patent Pools, BERKELEY L. (1999), https://2048.berkeley.edu/files/pools.pdf.
161. See, e.g., Amy Kapczynski et al., Addressing Global Health Inequities: An Open
Licensing Approach for University Innovations, 20 BERKELEY TECH. L.J. 1031 (2005); Peter
Lee, Contracting To Preserve Open Science: Consideration-Based Regulation in Patent Law,
58 EMORY L.J. 889, 915 (2009).
162. See, e.g., Schultz & Urban, supra note 109.
163. See, e.g., Robert P. Merges, A New Dynamism in the Public Domain, 71 U. CHI. L.
REV. 183 (2004) (examining private efforts to self-correct for excesses of patents).
1102
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
nisms important to some forms of cooperative innovation. The doctrinal tools available to support collaboration remain tethered to a market-based, producer-driven model of innovation. They do little to support the kinds of exchanges that are needed in a volunteer-driven
system of combined intellectual production. In many cases, the private measures are inherently incomplete, primarily or even solely
defensive in nature, and often extremely expensive.164 Some of the
strategies used to preserve openness may even backfire if patents
obtained for defensive purposes later become litigation tools.
Despite mitigating doctrines and private market responses,165
patents as currently enforced can and do systematically and significantly disadvantage some forms of cooperative innovation.166 Since
it is effectively impossible to opt out of the patent system, change
may be needed from within patent law to give forms of cooperative
innovation that are vulnerable to patents a chance to compete.167 In
addition, the public nature of a change in the law might have expressive effects missing from private efforts to circumvent the law, effects
that would help to strengthen group norms and values important to
cooperative systems.168
In Part IV, I propose one way for courts to respond to the vulnerabilities of cooperative innovation systems without radically changing
164. See, e.g., Asay, supra note 109, at 805 (describing the costs involved in trying to
preserve open systems of knowledge exchange through private orderings); Mattioli, supra
note 17, at 108 (examining question of whether patent sharing reflects a form of market
self-regulation, and critiquing view that private ordering can always correct for excessive
apportionment of patent rights).
165. See, e.g., U.S. DEP’T OF JUST. & FED. TRADE COMM’N, ANTITRUST GUIDELINES FOR
THE LICENSING OF INTELLECTUAL PROPERTY §§ 3.4, 4, and 5.6 (1995), available at
http://www.justice.gov/atr/public/guidelines/0558.pdf (discussing the use of the rule of
reason to evaluate various kinds of IP agreements and arrangements); Sheila F. Anthony,
Antitrust and Intellectual Property Law: From Adversaries to Partners, 28 AIPLA Q.J. 1
(2000) (discussing intersection of IP and antitrust, key focus of antitrust on precompetitive
nature of various agreements and arrangements involving IP).
166. For example, as work by Nobel Prize winner Elinor Ostrom and others on the
management of common pool resources has shown, external legal rights that control allocations within the group may interfere with valuable non-market forms of cooperation that
could achieve better collective outcomes. See, e.g., UNDERSTANDING KNOWLEDGE AS A
COMMONS, supra note 31, at 7 (including design principles for managing the production of
knowledge, viewed as a common pool resource; emphasizing importance of trust and reciprocity in sustaining cooperative systems; and emphasizing role of locally designed rules).
167. See, e.g., Schultz & Urban, supra note 109, at 10.
168. See, e.g., Jeanne C. Fromer, Expressive Incentives in Intellectual Property, 98 VA. L.
REV. 1745, 1746 (2012) (exploring the role of expressive incentives, those that express
solicitude for and protect a creator’s strong personhood and labor interests, in patent law);
Timothy R. Holbrook, The Expressive Impact of Patents, 84 WASH. U. L. REV. 573, 613 (2006)
(exploring the expressive impact that patent law can have, looking at how the grant of a
patent could communicate a message of inferiority to groups whose identity is tied to their
biology); William Hubbard, Inventing Norms, 44 CONN. L. REV. 369, 391 (2011) (arguing that
patent law has an important role to play in supporting “inventing norms” and that these
inventing norms should be incorporated into traditional patent law analysis).
2014]
COOPERATIVE INNOVATION
1103
the patent system. Recent pressures on the patent system fueled by
patent wars over the use of critical technology standards have created opportunities to make room for cooperative innovation in discussions about patent remedies. Rather than adding to the costs and uncertainties associated with broader judicial discretion in patent remedies, the principles I suggest might even act to limit existing uncertainties by providing a focal point for how patent remedies should be
adjusted and limiting the areas in which they are to be adjusted.
IV. PATENT REMEDIES WITH COOPERATION IN MIND
Patent remedies have become the topic of increasing public interest and debate as the social costs of patent litigation mount and the
divergence of public and private interests in laws governing patent
remedies become more apparent.169 In the pages that follow I show
why, when, and how we should make patent remedies more responsive to the needs of cooperative innovation.
A. Remedies Without Context
Patent remedies have historically been based on measuring and
awarding reasonable compensation for past infringement to the owner of a valid, infringed patent, generally accompanied by injunctive
relief to preclude future infringement.170 The patent owner has been
entitled to no less than a reasonable royalty, generally defined as the
royalty that a willing licensor and licensee would have agreed to at the
time of the initial infringement.171 Historically, courts would also routinely provide the plaintiff patent owner with injunctive relief preclud-
169. See James Boyle, Open Source Innovation, Patent Injunctions, and the Public Interest, 11 DUKE L. & TECH. REV. 30 (2012); Thomas F. Cotter, Reining in Remedies in Patent
Litigation: Three (Increasingly Immodest) Proposals, 30 SANTA CLARA COMPUTER & HIGH
TECH. L.J. 1 (2014); Ghosh, supra note 135; Golden, supra note 18, at 507 (“In legislative debates, public-choice concerns loom large, as proposed reforms appear commonly to track private, rather than necessarily public, interests.”); Megan M. La Belle, Patent Law As Public
Law, 20 GEO. MASON L. REV. 41 (2012); Rajec, supra note 19; Sichelman, supra note 19.
170. See also 35 U.S.C. § 284 (2006). The Patent Act provides that the owner of a valid,
infringed patent is entitled to damages that are “adequate to compensate for the infringement,
but in no event less than a reasonable royalty for the use made of the invention by the
infringer” with the possibility of punitive damages for willful infringement. Id.
171. See, e.g., Christopher B. Seaman, Reconsidering the Georgia-Pacific Standard for
Reasonable Royalty Patent Damages, 2011 BYU L. REV. 1661, 1676-81 (discussing Federal
Circuit approach to reasonable royalties). Monetary remedies for patent infringement may
take the form of either lost profits or lost royalties. In practice, however, determining what the
appropriate monetary remedies are has been “a complicated and confusing task.” Roger D.
Blair & Thomas F. Cotter, Rethinking Patent Damages, 10 TEX. INTELL. PROP. L. J. 1, 2 (2001)
(providing overview of traditional approaches to lost profits and reasonable royalties and
suggesting economic framework to aid in rethinking how patent damages are calculated).
1104
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
ing further infringing activity by the defendant.172 This approach to
patent remedies aligns roughly with the reward function of patents
within the traditional paradigm of producer-driven, market-based innovation. Focusing on compensation to the patent owner, however,
may result in patent remedies that decrease rather than increase innovation in some contexts of cooperative intellectual production.173
One of the biggest limitations of the traditional approach to patent
remedies has been its focus on the interests of the patent owner, to
the exclusion of the interests of additional participants in the innovation process and the broader public interests in socially optimal levels
of innovation.174 Interests beyond those of the patent owner, including
both the infringer and the public more generally, started to play a
bigger role in the calculation of patent remedies following the Supreme Court decision in eBay v. MercExchange in 2006.175 The Supreme Court held in eBay that injunctive relief in patent cases, just
as in other types of cases, should not be presumed but rather should
be determined based on the conventional four-factor test for injunctive relief.176 The Supreme Court thus shifted the calculus used by
courts when determining whether to award injunctive relief to the
owner of an infringed patent by requiring courts to take the harm of
an injunction to the public and the defendant explicitly into account.177 Justice Kennedy’s influential concurring opinion went on to
172. See, e.g., Richardson v. Suzuki Motor Co., 868 F.2d 1226, 1247 (Fed. Cir. 1989) (“It
is the general rule that an injunction will issue when infringement has been adjudged,
absent a sound reason for denying it.”).
173. See, e.g., Ted Sichelman, Purging Patent Law of “Private Law” Remedies, 92 TEX.
L. REV. 517 (2014); see also Boyle, supra note 169, at 32-34 (outlining some of the ways in
which open source innovation is particularly vulnerable to patent injunctions); Thomas R.
Cotter, Make No Little Plans: Response to Ted Sichelman, Purging Patent Law of “Private
Law” Remedies, 92 TEX. L. REV. 25 (2014); Paul J. Heald, Optimal Remedies for Patent
Infringement: A Transactional Model, 45 HOUS. L. REV. 1165 (2008).
174. Patent remedies take the form of injunctive relief and damages. Damages are
based on calculations of loss to the patent owner, either in the form of a patent owner’s lost
profits, price erosion, or reasonable royalties. See, e.g., THOMAS F. COTTER, COMPARATIVE
PATENT REMEDIES: A LEGAL AND ECONOMIC ANALYSIS (2013) (comprehensively discussing
patent remedies in the United States and selected other countries). The works referenced
in footnote 171 highlight various limitations of the traditional approach to remedies taken
by the courts.
175. eBay Inc. v. MercExchange, LLC, 547 U.S. 388, 394 (2006) (holding that the wellestablished principles of equitable relief apply with equal force to disputes under the
Patent Act).
176. Id. at 391-93. This case replaced the Federal Circuit’s “general rule” in favor of
granting injunctions based on a presumption of irreparable harm with a required
balancing of interests. To obtain injunctive relief, the plaintiff must show that it has
suffered irreparable harm, remedies available under the law are inadequate to
compensate, the balance of hardships associated with injunctive relief favor the plaintiff,
and the public interest will not suffer by the issuance of the requested injunction. Id.
177. See, e.g., FED. TRADE COMM’N, THE EVOLVING IP MARKETPLACE: ALIGNING PATENT
NOTICE AND REMEDIES WITH COMPETITION 253 (2011) [hereinafter FTC REPORT]; see also
2014]
COOPERATIVE INNOVATION
1105
suggest specific situations in which district courts might find injunctive relief inappropriate, such as patent assertion by non-practicing
entities and situations of patent holdup.178
While eBay has made the role of the public interest explicit in determinations of injunctive relief, the case fails to provide principles to
guide courts in their determinations of what the public interest is or
how it should be measured.179 Courts have taken into account a
greater variety of factors when determining whether an injunction
should be awarded, although many of these considerations remain
focused on whether the patent owner will suffer irreparable harm
due to the infringement through loss of market share, price erosion,
difficulty in calculating damages, or harm to reputation or goodwill.180 Whether the patent owner and infringer compete and whether
the patent owner is a practicing entity appear to be important but
not determining factors in predicting the availability of injunctive
relief, and factors such as market structure and the relationship of
the patented invention to the infringing product increasingly inform
court decisions.181 Despite the more contextualized analysis of the
effects of infringement and injunctive relief, however, detailed analysis of the effects of injunctive relief on the broader public interest
Mark P. Gergen, John M. Golden & Henry E. Smith, The Supreme Court’s Accidental
Revolution? The Test for Permanent Injunctions, 112 COLUM. L. REV. 203, 204-05 (2012)
(suggesting that the eBay case has launched a revolution in the law of equitable remedies
beyond patent law).
178. See, e.g., eBay Inc., 547 U.S. at 395-97 (Kennedy, J., concurring).
179. See, e.g., FTC REPORT, supra note 177, at 255 (“After enumerating the four
equitable factors in the eBay decision, the opinion of the full Court gave little guidance on
their application.”); see also Scott A. Allen, “Justifying” the Public Interest in Patent
Litigation, 88 IND. L.J. 1047, 1051 (2013).
180. FTC REPORT, supra note 177, at 215 (discussing the ways in which courts have
considered and applied the four factors in the test for injunctive relief); see also James M.
Fischer, What Hath eBay v. MercExchange Wrought?, 14 LEWIS & CLARK L. REV. 555, 564-65
(2010); Stephen E. Noona, Permanent Injunctions for Patent Infringement in a Post eBay
World, INSIDE BUS. (Dec. 9, 2010, 4:12 AM), http://insidebiz.com/blogs/kaufman-canoles/
permanent-injunctions-patent-infringement-post-ebay-world; Bryan J. Vogel & Shane St.
Hill, IP: Injunctions and Irreparable Harm After eBay, INSIDECOUNSEL (June 19, 2012),
http://www.insidecounsel.com/2012/06/19/ip-injunctions-and-irreparable-harm-after-ebay.
181. See, e.g., FTC REPORT, supra note 177, at 259 (showing results of survey of case
law); see also eBay, 547 U.S. at 393 (Supreme Court’s warning against categorical rules in
injunction analysis); Andrew Beckerman-Rodau, The Aftermath of eBay v. MercExchange,
126 S. Ct. 1837 (2006): A Review of Subsequent Judicial Decisions, 89 J. PAT. &
TRADEMARK OFF. SOC’Y 631 (2007); Bernard H. Chao, After eBay, Inc. v. MercExchange:
The Changing Landscape for Patent Remedies, 9 MINN. J.L. SCI. & TECH. 543 (2008)
(arguing the importance of factors such as whether infringer is a competitor and showing of
irreparable harm in post-eBay cases). For a more recent update that largely affirms the
importance of whether the parties were competitors and the ability to show irreparable
harm, see, for example, Barbara A. Fiacco, The Impact of eBay v. MercExchange,
Presentation at the Duke Pat. Law Inst. (May 16, 2013) (presentation available at
http://law.duke.edu/sites/default/files/centers/judicialstudies/patentlawintensive/Fiacco-Ma
y%2016%20eBay%20v%20MercExchange.pdf).
1106
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
remains rare except in situations in which injunctions would have
severe consequences for public health and safety or would significantly disrupt markets, or more recently in situations where patents
cover important industry standards.182 Instead of conducting a broader inquiry into the effects of injunctions on innovation, some courts
still simply equate the public interest with supporting innovation
through a strong patent system.183 Although injunctive relief is
no longer routine in the wake of eBay, injunctive relief seems to be
based most frequently on a showing of irreparable harm to the patent
owner and continues to be granted in the majority of cases.184 Thus,
while determinations of injunctive relief following eBay have moved
patent remedies in the direction of reflecting a richer range of
entitlements, I suggest that courts still have not adequately captured
the public interest in limiting injunctive relief or given it sufficient
context or weight.
This limited approach to the public interest may be starting to
change, particularly in the context of cases involving patents that are
essential to the use of important industry-wide technology stand182. See, e.g., FTC REPORT, supra note 177, at 232-34 (discussing the role of public
interest, including survey of cases that address public interest explicitly as part of decision
to award or deny injunctive relief). Greater attention to the public interest does seem to
occur in cases involving substantial network effects that are threatened by patents. See,
e.g., Apple, Inc. v. Samsung Elecs. Co., 909 F. Supp. 2d 1147, 1163 (N.D. Cal. 2012) (taking
impact on consumers into account and denying injunctive relief despite status of parties as
direct competitors); Apple Inc. v. Samsung Elecs. Co., 735 F.3d 1352, 1375 (Fed. Cir. 2013)
(affirming District Court’s denial of injunctive relief despite status of parties as direct competitors for design patents, but vacating denial of injunctive relief for utility patents).
183. FTC REPORT, supra note 177, at 270-271 (“Only a small number of post-eBay cases
have provided an extended discussion of this factor in deciding whether to grant an
injunction. In the majority of cases, courts simply recognize that the ‘public has an interest
in maintaining a strong patent system. This interest is served by enforcing an adequate
remedy for patent infringement.’ ” (quoting TiVo Inc. v. Echostar Commc’ns Corp., 446 F.
Supp. 2d 664, 670 (E.D. Tex. 2006)); see also, e.g., Pozen Inc. v. Par Pharm., Inc., 800 F.
Supp. 2d 789, 825 (E.D. Tex. 2011) (reasoning that an injunction served the public interest
because it encouraged innovation by upholding patent owner’s “right to exclude”); Zen
Design Grp., Ltd. v. Clint, No. 08-cv-14309, 2009 WL 4050247, at *6 (E.D. Mich. Nov. 23,
2009) (reasoning that a permanent injunction served the public interest because, without
it, the patent’s actual value would be reduced to “a fraction of its intended value”); Funai
Elecs. Co., Ltd. v. Daewoo Elecs. Corp., 593 F. Supp. 2d 1088, 1111 (N.D. Cal. 2009)
(“Protection of the rights of patent holders is generally in the public interest.”).
184. See, e.g., FTC REPORT, supra note 177, at 217-218 (“Surveys of post-eBay cases
reveal that district courts have granted approximately 72%–77% of permanent injunction
requests.”); see also Ernest Grumbles III et al., The Three Year Anniversary of eBay v.
MercExchange: A Statistical Analysis of Permanent Injunctions, INTELL. PROP. TODAY, Nov.
2009, at 25, 26 (noting that seventy two percent of requests were granted through May 1,
2009, based on review of decisions available through Lexis); Erin Coe, Injunctions Harder To
Win in Post-EBay Courts, LAW360 (Oct. 30, 2008, 12:00 AM), http://www.law360.com/
articles/74829/injunctions-harder-to-win-in-post-ebay-courts; Post-eBay Permanent Injunction
Rulings in Patent Cases, PATSTATS.ORG, http://www.patstats.org/Patstats2.html (last
updated May 26, 2013) (showing that injunctions were granted in 167 cases and denied in
55 cases, that is, granted seventy-five percent of the time).
2014]
COOPERATIVE INNOVATION
1107
ards.185 In a series of high-profile patent infringement cases involving
patents covering technology standards used in the smartphone industry, courts have had to consider whether injunctive relief should
be available where patent owners have previously agreed to license
patents essential to the use of these standards on reasonable and
non-discriminatory terms. In one of the first court decisions to weigh
in on this matter, Judge Posner suggested that injunctive relief
should not be available, emphasizing the harm to the public interest
that such relief would impose.186 Judge Posner went on to suggest
that if the patent owner and infringer cannot agree on licensing
terms, compulsory licensing with ongoing royalties should be used to
resolve the dispute in a way that appropriately balances the harm to
the patentee from infringement with the harm to the infringer and to
the public from an injunction.187 This approach focuses explicitly on
what the public interest is and how it should be reflected in determinations of patent remedies in standard setting contexts.188 While
Judge Posner’s willingness to radically reshape how remedies are
calculated diverges from the mainstream,189 his underlying message
about the need to reconsider patent remedies in contexts where they
may threaten important forms of coordination and cooperation is being taken seriously by policymakers. So far, however, this reconsid185. For an optimistic view of this change more generally, see, for example, Boyle,
supra note 169. On the other hand, one fragmented opinion from the Federal Circuit
illustrates the divergence of views held by judges in that court about when injunctive
relieve should be available in the context of patents covering important industry standards.
See Apple Inc. v. Motorola, Inc., Nos. 12-1548, 12-1549, 2014 WL 1646435, at *30-35 (Fed.
Cir. April 25, 2014); id. at *36-37 (Rader, C.J., dissenting in part); id. at *43-46 (Prost, J.,
concurring in part and dissenting in part).
186. See Apple, Inc. v. Motorola, Inc. 869 F. Supp. 2d 901, 913-14 (N.D. Ill. 2012), aff’d
in part, rev’d in part, Apple Inc., 2014 WL 1646435. (suggesting that the purpose of the
fair, reasonable, and non-discriminatory terms (FRAND) requirement is “to confine the
patentee’s royalty demand to the value conferred by the patent itself as distinct from the
additional value—the hold-up value—conferred by the patent’s being designated as
standard-essential”). But see Apple, Inc., 2014 WL 1646435, at *35 (affirming Judge
Posner’s ruling that this particular holder of the patent subject to FRAND requirements
was not entitled to injunctive relief, but insisting that eBay analysis of factors, with
particular attention to whether patent holder will suffer irreparable harm as a result of the
infringement, is required).
187. Apple Inc., 869 F. Supp. 2d at 918.
188. See eBay Inc. v. MercExchange, LLC, 547 U.S. 388, 395-97 (2006) (Kennedy,
J., concurring).
189. See, e.g., Gene Quinn, Chief Judge Rader on the Supreme Court and Judge Posner,
IPWATCHDOG (Mar. 17, 2013, 12:07 PM), http://www.ipwatchdog.com/2013/03/17/chiefjudge-rader-on-the-supreme-court-and-judge-posner/id=37620/; see also Apple, Inc., 2014
WL 1646435, at *35 (holding that Judge Posner erred in his district court opinion by
applying “a per se rule that injunctions are unavailable for [standard-essential patents]”);
id. at *36-37 (Rader, C.J., dissenting in part) (reasoning that the district court’s failure to
consider the appropriate factors left it “adrift without a map”); id. at *45 (Prost, J.,
concurring in part and dissenting in part) (rejecting the district court’s “categorical rule
that a patentee can never obtain an injunction on a FRAND-committed patent”).
1108
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
eration has been limited primarily to the context of standardessential patents encumbered by contractual commitments to license
on reasonable and non-discriminatory terms.190
Calculations of patent damages have also moved in the direction
of a more contextualized analysis, driven largely by decisions in the
standard setting context. In the first court decision to confront and
calculate reasonable and non-discriminatory royalties for patents essential to the use of industry standards, Judge Robart concluded in
Microsoft v. Motorola that the traditional factors used to determine
reasonable royalties, referred to as the Georgia-Pacific factors, should
be explicitly modified to take the standard-setting context into account.191 Judge Robart, as Judge Posner had before him, emphasized
the importance of the public interest in access to the standards and
190. Judge Posner’s approach in Apple v. Motorola is contrasted with Judge Robart’s
ruling on FRAND damages in Microsoft Corp. v. Motorola, Inc., No. C10-1823JLR, 2013
WL 2111217, at *18-19 (W.D. Wash. April 25, 2013). Areas of divergence include Judge
Robart’s reliance on the traditional Georgia-Pacific factors as a good way of calculating
patent damages, and Judge Posner’s rejection of this framework. Compare id.
(summarizing Judge Robart’s position), with Apple, Inc., 869 F. Supp. 2d at 910-11 (giving
Judge Posner’s take). Judge Posner also insists that the FRAND value of a standardessential patent should be determined ex ante, pre-standardization, as compared to the
traditional approach based on the date that infringement began. See, e.g., Florian Mueller,
A Closer Look at the 207-Page, Landmark FRAND Rate-Setting Decision in Microsoft v.
Motorola, FOSS PATENTS (Apr. 28, 2013, 11:49 PM), http://www.fosspatents.com/2013/04/acloser-look-at-207-page-landmark.html#judgesrobartandposner. Both decisions, however,
point to the importance of the public interest in access to standard-essential patents
and the need to exclude the holdup value of patents covering standards from
royalty determinations.
191. See, e.g., Microsoft Corp., 2013 WL 2111217, at *1, 18 (deciding in contract dispute
over whether Motorola had breached contract to offer patents on FRAND terms by asking
for royalties that were unreasonably high). Judge Robart’s decision seems to explicitly
modify the traditional Georgia-Pacific factors to account for the standard-setting context.
Id. at *18-20. He recognizes that the licensing of standard-essential patents takes on a
public character and must be conducted and reviewed with those public benefits in mind.
He emphasizes, for example, that patent royalties should not incorporate the holdup value
that may result after a standard incorporating the patent has been chosen, and attention
to royalty stacking. Id.; see also Thomas F. Cotter, Some Initial Reactions to Judge Robart’s
Opinion in Microsoft v. Motorola, INTELLECTUALIP (May 3, 2013), http://intellectual
ip.com/2013/05/03/some-initial-reactions-to-judge-robarts-opinion-in-microsoft-v-motorola/;
Michael Carrier, A US Court Issues First Analysis of an Appropriate Royalty that a
Patentee Could Obtain After Promising To License Its Patent on Reasonable and
Nondiscriminatory (RAND) Terms (Microsoft v Motorola), E-COMPETITIONS BULL. (Inst. of
Competition Law, New York, N.Y.), May 2013; Jorge L. Contreras, So That’s What “RAND”
Means?: A Brief Report on the Findings of Fact and Conclusions of Law in Microsoft v.
Motorola, PATENTLY-O (Apr. 27, 2013), http://patentlyo.com/patent/2013/04/so-thats-whatrand-means-a-brief-report-on-the-findings-of-fact-and-conclusions-of-law-in-microsoft-vmotorola.html; Thomas F. Cotter, Day 2 of the University of Florida Workshop on Standard
Essential Patents and FRAND: Page on “Judging Monopolistic Pricing: F/RAND and
Antitrust Injury”, COMP. PAT. REMEDIES BLOG (Sept. 12, 2013), http://comparativepat
entremedies.blogspot.com/2013/09/day-2-of-university-of-florida-workshop_12.html (summarizing comments by William Page suggesting that while the form of FRAND royalties was
as a hypothetical bilateral negotiation, in reality it was a calculation in light of the
economic consensus of what FRAND rates should be).
2014]
COOPERATIVE INNOVATION
1109
the need to exclude any patent holdup value arising from the collective adoption of a standard covered by the patent from the calculation
of damages. Judge Robart’s approach, including the emphasis on excluding the value of standardization in royalty determinations, has
been followed in a second court ruling on reasonable and nondiscriminatory royalty rates.192 Recognizing the strong public interest
in using standards as mechanisms for coordinating interoperable
technologies, the Federal Trade Commission, the Department of Justice, the U.S. Patent and Trademark Office, and even the U.S. Trade
Representative have all offered similar opinions about limiting injunctive relief and damages.193 Courts and policymakers alike are
thus emphasizing the public interest in preserving access to innovative standards, and their responses illustrate ways in which the
broader effects of patent remedies on cooperation in innovation can
be incorporated into remedy determinations.194 So far, however, they
remain limited primarily to patents essential to common industry
technology standards.195 They also remain focused primarily on determinations of injunctive relief and damages for past infringement,
leaving open questions about whether the public interest will similar-
192. See In re Innovatio IP Ventures, LLC Patent Litig., 921 F. Supp. 2d 903, 915-17
(N.D. Ill. 2013); see also Michael Carrier, A US Court Issues Second Ruling Determining
RAND Rate for Standard Essential Patent (Innovatio), E-COMPETITIONS BULL. (Inst. of
Competition Law, New York, N.Y.), Oct. 2013.
193. See, e.g., Editorial Board, FTC/DOJ/USPTO Take Action on Standards-Essential
Patents, ANTITRUST & COMPETITION NEWSL., ORRICK (Feb. 4, 2013), http://blogs.orrick.com/
antitrust/2013/02/04/dojftcuspto-take-action-on-standards-essential-patents/ (summarizing
recent actions by FTC, DOJ, and USPTO with regard to standard-essential patents); John
Ribeiro, Google Withdraws Standard-Essential Patent Claims in Xbox Complaint,
COMPUTERWORLD (Jan. 9 2013, 12:20 PM), http://www.computerworld.com/s/article/
9235502/google_withdraws_standard_essential_patent_claims_in_Xbox_complaint; see also
Letter from Michael B.G. Froman, Ambassador, Exec. Off. of the President, to Irving A.
Williamson, Chairman, U.S. Int’l Trade Comm’n (Aug. 3, 2013) available at
http://www.ustr.gov/sites/default/files/08032013%20Letter_1.PDF (discussing a policy
decision to disapprove Commission’s determination to enjoin Apple from importing and
selling infringing devices, after review in light of public interest concerns).
194. See, e.g., Thomas F. Cotter, Remedies for the Infringement of Standard Essential
Patents Subject to a FRAND Commitment, COMP. PAT. REMEDIES BLOG (May 8, 2013, 9:40
AM), http://comparativepatentremedies.blogspot.com/2013/05/remedies-for-infringementof-standard.html.
195. Existing proposals in the literature focus primarily on risks of patent holdup and
related forms of opportunistic behavior by patent holders. See, e.g., Cotter, supra note 19;
Mark A. Lemley & Carl Shapiro, Patent Holdup and Royalty Stacking, 85 TEX. L. REV.
1991 (2007); Mark A. Lemley, Ten Things To Do About Patent Holdup of Standards (and
One Not To), 48 B.C. L. REV. 149 (2007); Robert P. Merges & Jeffrey M. Kuhn, An Estoppel
Doctrine for Patented Standards, 97 CALIF. L. REV. 1 (2009); Joseph Scott Miller, Standard
Setting, Patents, and Access Lock-in: RAND Licensing and the Theory of the Firm, 40 IND.
L. REV. 351 (2007). While I am also concerned with patent holdup, my concerns extend to
other ways in which patents might interfere with cooperative innovation, and my goal is to
limit patent remedies when, and to the extent, that the presence of patents makes useful
cooperative systems of innovation unviable.
1110
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
ly inform determinations of ongoing royalties where injunctive relief
is denied.196
These responses to cooperative contexts, while creating opportunities to inject broader public interests into determinations of patent
remedies, leave open important questions about how tensions between patents and systems that rely more heavily on non-market
mechanisms of cooperation should be handled. They address only
problems arising from commercial participants in standard setting
organizations who do not honor their obligations to each other relating to standard-essential patents. They limit their focus to the potential harms to the public from refusals of both patent-holding members of standard setting and the users of patented standards to negotiate licenses to patents essential to innovative standards on terms
that are reasonable and non-discriminatory. Moreover, they fail to
acknowledge or protect the public interest in sustaining non-market
mechanisms of cooperation that are valuable to innovation. The principles proposed below are largely consistent with but go further than
existing policy responses. They provide the courts with a way of systematically incorporating the harm that patents impose on cooperative systems into determinations of patent remedies.197
B. Principles for the Design of Remedies
Where patent rights impact systems of cooperative innovation,
these negative effects should be included in determinations of patent
remedies.198 As the examples of Foldit and open source software
demonstrated, three key non-market mechanisms play important
roles in sustaining cooperative innovation: trust, benefit-sharing, and
reciprocity.199 Patents under the current system directly threaten
196. See, e.g., Thomas F. Cotter, Four Principles for Calculating Reasonable Royalties
in Patent Infringement Litigation, 27 SANTA CLARA COMPUTER & HIGH TECH. L.J. 725
(2011); Mark A. Lemley, The Ongoing Confusion over Ongoing Royalties, 76 MO. L. REV.
695 (2011); see also Thomas F. Cotter, U.S. District Court Awards Carnegie Mellon $1.5
Billion in Patent Damages, COMP. PAT. REMEDIES BLOG (Apr. 21, 2014), http://comparative
patentremedies.blogspot.com/2014/04/us-district-court-awards-carnegie.html
(suggesting
decision not to award a higher ongoing royalty than the reasonable royalties awarded for
past infringement is consistent with view that courts should take socially optimal
incentives to innovate into account).
197. Ideally this is just a first step in thinking more systematically about where
patent laws need to change in response to changing forms of collaborative and
cooperative innovation.
198. For a related discussion and insights into the challenges of cumulative and
sequential innovation and the disconnect with current approaches to patent remedies, see,
for example, J. H. Reichman, Of Green Tulips and Legal Kudzu: Repackaging Rights in
Subpatentable Innovation, 53 VAND. L. REV. 1743 (2000). For a similar discussion focusing
on open source innovation, see, for example, James Boyle, supra note 169.
199. As noted earlier, there are other important aspects of cooperative innovation
systems that may not be fully captured by these three mechanisms. I focus on them for
2014]
COOPERATIVE INNOVATION
1111
each of these mechanisms, putting such cooperative innovation systems at risk. In response, I provide three principles for courts to use
in determining patent remedies that are designed to protect the operation of these mechanisms in situations of socially beneficial cooperative innovation. The three principles are: (1) protect reliance interests in norms of open access and sharing;200 (2) limit the private
appropriation of collective value;201 and (3) reinforce reciprocity in
free, open systems of innovation. Where the negative effects of patents on cooperative mechanisms are likely to be strong, the principles provide a reasoned way of incorporating these patent harms into
determinations of patent remedies. This approach would lead not only to changes in how courts determine remedies for patents that arise
from or cover the activities of cooperative systems of innovation, but
also, and more importantly, to changes in ex ante decisions by members of a cooperative innovation system about whether to defect from
the system and decisions by third parties about whether and when to
obtain and enforce patents against members of the cooperative system in the first place.202 It is these ex ante effects on decisions to de-
the purpose of this analysis because they seem to play an important role in many of
the cooperative systems and also seem to capture many of the effects that patents
might have on the rich variety of non-economic factors and motivations that drive
cooperative innovation.
200. The idea of respecting the reliance interests of firms in standards that are adopted
by the industry has been suggested in Merges & Kuhn, supra note 195. This principle goes
further, extending to any situation in which multiple participants work collectively to
advance a particular product or field and either must coordinate their activities through
standards and/or find it necessary to use certain core technologies as research tools or
platform technologies on which to build their contributions. There are also similarities here
to an essential-facilities doctrine approach to intellectual property. See, e.g., M. Elaine
Johnston, Intellectual Property As an “Essential Facility”, 22 COMPUTER & INTERNET L. 17
(2005) (summarizing case law and trends in applying essential-facilities doctrine to
intellectual property contexts). Jorge Contreras proposes a “market reliance” approach in
the context of promises not to assert patents on FRAND terms, focusing on the importance
of protecting the reasonable reliance of third-party market participants on such promises.
His work provides additional ideas about how this kind of principle might work in the
context of cooperative innovation. See Jorge L. Contreras, Market Reliance and Patent
Pledges, UTAH L. REV. (forthcoming Spring 2015), available at http://ssrn.com/
abstract=2309023.
201. See, e.g., Cotter, supra note 169 (arguing that damages should be apportioned
according to relative value of patent to the whole product); Lemley & Shapiro, supra
note 195 (exploring problems of patent holdup, royalty-stacking, and consequent
royalty overcharges).
202. The principles will have the effect of limiting injunctive relief and patent damages
in ways that reduce the payoffs that group members might expect from defecting and their
expectations that others might defect. This will, in turn, reduce the incentives of group
members to defect from the group in the first place, and will also strengthen the
motivations of group members to continue to adhere to group norms of open access and
sharing. The principles will also limit the ability of outside parties to appropriate the value
arising from collective efforts and their ability to threaten the activities of the group. Third
parties will have lower expected payoffs from asserting or threatening to assert patents
against the group in contexts where they are seeking to tax or free ride on the collective
1112
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
fect and decisions to threaten and enforce patents against the group
that are most important to the survival of cooperative innovation.
1. First Principle
The first principle requires courts to protect reliance interests in
norms of open access and sharing. Where the norms are supporting
socially valuable innovation, reliance interests in continued access
to an invention based on this norm should be treated as part of the
public interest in continued access, to be balanced against the
interests of a patent owner in restricting access when fashioning patent remedies.203 To receive protection, the reliance must be reasonable in light of widely adopted and publicly known norms governing
when and how knowledge will be shared and used. It must be reliance on a norm of open access and sharing that is relevant to the cooperative process of innovation. Defining what is reasonable reliance
would thus be context-specific, depending on the scope and nature of
the norms and customs of the innovation community, the extent to
which these norms are publicly known, the relationship of the inventor and patent owner to the community, and the balancing of collective interests in access and private incentives to make and develop
the patented technology.
This approach provides courts with a way of respecting and reinforcing informal rules that establish socially beneficial uses of
knowledge by giving them weight when balancing public and private
interests. Pursuant to this principle, reasonable reliance on norms of
open access and sharing would become an important factor that
weighs against injunctive relief. This reliance would also become a
factor limiting damages for infringement of patented inventions that
are used in the context of cooperative innovation.204 Calculations of
reasonable royalties should be lower where reliance interests are
efforts of the group and less bargaining power in situations where injunctive relief is likely
to be unavailable. This may discourage them from asserting claims or making threats and
will also limit the deterrent effects of potential third-party claims on participation rates in
cooperative-innovation projects. Where licenses are required, third parties will settle for
lower royalties based on expectations that courts will limit injunctive relief and damages.
203. This principle finds support in studies of the role that informal norms play in
knowledge communities. See, e.g., Robert P. Merges, From Medieval Guilds to Open Source
Software: Informal Norms, Appropriability Institutions, and Innovation (Univ. of Wis. Law
Sch. Conf. on the Legal Hist. of Intell. Prop., Working Paper, 2004).
204. For discussions of the effects of patents on norms, see, for example, Merges, supra
note 68; Katherine J. Strandburg, User Innovator Community Norms: At the Boundary
Between Academic and Industry Research, 77 FORDHAM L. REV. 2237 (2009) (exploring
implications of convergence of academic research with commercial interests and
implications for norms of sharing research tools and materials and suggesting need for
policies to enhance sharing); Murray & Stern, supra note 25; Fiona Murray, The
Oncomouse that Roared: Resistance & Accommodation to Patenting in Academic Science
(March 2006) (unpublished manuscript) (on file with author).
2014]
COOPERATIVE INNOVATION
1113
higher and royalties should reflect the economic constraints of the
group. Courts would apply this principle with the goal of reducing the
attractiveness for members of a cooperative system to deviate from
sharing norms, and to limit at least some of the third-party actions
that increase the cost of adhering to group norms of sharing. While
finding a way to measure and account for this reliance interest will
be difficult, courts, as a starting point, can look to the emerging
framework for determining when injunctive relief should be available
and what reasonable and nondiscriminatory royalties are in the
standard-setting context.
While existing doctrines of implied license and estoppel, as well as
the shop rights that a company may have over its employees’ inventions, go part of the way towards recognizing and protecting reasonable reliance in continued group use of inventions, this first principle
goes beyond the limits of these doctrines.205 The principle would encompass the reliance interests of participants who are not in contractual privity, or even in direct or indirect communication, with the patent owner, as long as the norms of open access and sharing that the
group relies upon are widely adopted and publicly known and are associated with a socially valuable system of innovation. Protection of
the reliance interest in continued use will be strongest for inventions
discovered by members of the group that are used by the group. It
will extend in a more limited way to inventions made by group members working on independent projects entirely outside of the group,
and in an even more limited way to inventions owned by third parties
that the group uses. In these latter two cases, the reliance interest
will play a role only for inventions that are both closely related and
important to the activities of the group and additionally made accessible to the group in ways that encourage, whether directly or indirectly, widespread use by the group. If a software company benefits
from Linux and makes available software tools with a reasonable expectation that members of the Linux community will pick up and
widely use the tools, for example, the principle would limit the remedies that the software company might expect from asserting patents
covering this software. The principle would also weigh against injunctive relief and limit damages in situations such as the generalized patent threats made by Microsoft to Linux users. Knowing that
205. See, e.g., Barnes & Noble, Inc. v. LSI Corp., 849 F. Supp. 2d 925, 939-40 (N.D. Cal.
2012) (holding that a competitor’s adoption of set standards can constitute reasonable reliance); Lautzenhiser Techs., LLC v. Sunrise Med. HHG, Inc., 752 F. Supp. 2d 988, 1009
(S.D. Ind. 2010) (holding that amicable dealings over a period of years may be enough to
constitute misleading conduct that induces reasonable reliance). Similar trends are evident
in contract law, where courts are increasingly willing to protect reasonable reliance interests of negotiating parties prior to or in the absence of a final contract. See, e.g., Alan
Schwartz & Robert E. Scott, Precontractual Liability and Preliminary Agreements, 120
HARV. L. REV. 661 (2007).
1114
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
injunctive relief and remedies will be limited, Linux users would feel
more comfortable continuing to participate in open source software,
and Microsoft would find its divide-and-conquer strategies less effective. In contrast, this principle would not block injunctive relief or
limit damages in situations of blatant copying of an independent proprietary product for the purpose of making it open source.
Where established norms of open access and sharing govern intellectual production and members have expectations of continued access to and use of inventions based on these norms, protecting their
reliance interests in this way makes it easier to maintain the swift
trust that both open source software systems and crowd science systems rely on.206 Participants in cooperative innovation make contributions and adhere to group rules with the expectation that others
will do the same. Sustaining this kind of trust in the behavior of the
group requires participation by most, if not almost all, members of
the group. Anything that increases the ability and incentives of group
members to defect from group norms will threaten this equilibrium
state of general trust and make it harder to sustain norms of open
access and free sharing. In the Linux example, one of the reasons
that the license between Novell and Microsoft was so troubling to the
open source community was its negative impact on expectations that
open source norms would continue to govern open source software.
Patents covering open source software provided Novell with an opportunity for significant private commercial gain, leading to a defection from open access norms. Not surprisingly, additional license
deals between Microsoft and other open source software companies
followed in the wake of the Novell deal.207 Without the backing and
defensive patenting of large companies like IBM, such defections
might have led to unraveling of cooperation and the future of Linux
might have been questionable. While Linux survived, this survival
has come at a high cost, both in terms of centralized corporate control
over development paths and the costs of defensive patenting. Application of this proposed principle would have limited the ability of
Novell or any third party acquirer to obtain injunctive relief or damages from the assertion of Novell’s open source patents against the
open source community. This would have made the patents less
commercially attractive and reduced Novell’s incentives to defect.
In the context of Foldit, the commitment of the Foldit administrators and Foldit players to ensuring that scientific discoveries and the
206. For a discussion of swift trust and its role in open source software communities,
see Part II.A.
207. See, e.g., Mary Jo Foley, Microsoft and SUSE Extend Microsoft’s Controversial
Novell Linux Pact, ZDNET (July 25, 2011), http://www.zdnet.com/blog/microsoft/
microsoft-and-suse-extend-microsofts-controversial-novell-linux-pact/10164; Paul Krill,
supra note 110.
2014]
COOPERATIVE INNOVATION
1115
research tools that enable such discoveries are made publicly available is critical to its success. Members of Foldit are willing to freely
contribute their energy and ideas because they expect and believe
both that other participants will behave similarly and that the results of their collective efforts will be freely shared with each other
and with the scientific community. While defensive publishing of results may limit the ability of Foldit players to patent their discoveries, and claims of joint inventorship may limit the ability of any individual to obtain exclusive control, much of what the game members
contribute and share are software tools and problem solving techniques rather than the ultimate discovery. Moreover, the fact that
participants make their results public now does not mean that they
will continue to do so if commercial considerations play a larger
role.208 This additional protection is therefore necessary to protect
Foldit norms of open and free sharing.
2. Second Principle
The second principle is to limit the private appropriation of collective value by a patent owner, with the goal of supporting benefitsharing in contexts of cooperative innovation. The principle would
limit patent remedies where patents are used to hold up group production or extract rents from the group based on the group’s adoption
and collective use of the invention.209 While this is analogous to the
patent holdup concerns that arise when patents cover the use of
standards adopted by industry members, it is broader, encompassing
situations in which a group widely adopts and uses inventions that
further the innovation objectives of the group. Concerns about the
appropriation of collective value through patent holdup will be
strongest in situations where the cost of switching to another technology would be high and where the fact that the group has used the
patented invention has made it more valuable to the group and to
people outside the group. This principle would weigh strongly against
208. Similar considerations and concerns about the effects of patents on sharing have
arisen in academic science. Academic science is characterized by norms that support open
dissemination and use of research results. The more scientists can rely on these norms, the
more willing they will be to continue to share their own discoveries and to use and
experiment with the discoveries of other scientists, encouraging trust and reciprocity. These
norms are public knowledge, there are social benefits from encouraging reliance on these
norms, and it appears that many scientists do indeed rely on these norms. See, e.g., Mark A.
Lemley, Ignoring Patents, 2008 MICH. ST. L. REV. 19. But see Merges, supra note 68.
209. There is an extensive literature on patent holdup, and this principle adopts and
extends the recommendations that many patent scholars have made in the context of
patents covering one component of a larger product. See, e.g., Lemley, supra note 195. But
see F. Scott Kieff & Anne Layne-Farrar, Incentive Effects from Different Approaches to
Holdup Mitigation Surrounding Patent Remedies and Standard-Setting Organizations, 9 J.
COMPETITION L. & ECON. 1091 (2013) (emphasizing importance of context but in this case
to limit the situations of patent holdup that deserve a patent remedy response).
1116
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
awarding injunctive relief where the value of a patented invention is
largely attributable to its widespread adoption and use by the
group.210 The interests of the group would nevertheless still be balanced against the interests of the patent owner, and the principle
will have most application for patents covering inventions that are
made by members of the group or third-party patent owners who collaborate with or directly benefit from the activities of the group. This
principle would also factor into the determination of damages by excluding the value that the patented invention has as a result of its
widespread adoption and use by the group.211
The principle will also limit patent remedies to account for the
benefits, if any, received by the patent owner from the group—
primarily but not exclusively patent owners who are members of the
group or who obtained their rights to the invention from a member of
the group. First, the group will have limited ability to patent and restrict use of inventions discovered by the group, making bargaining
with third-party patent owners and members of the group who defect
and patent their own inventions more difficult.212 Determining when
an invention has emerged, who the inventors are, and whether and
how patenting should be pursued, as well as handling negotiations
with a third party regarding use of their patented technology, can be
extremely difficult in contexts of massively distributed collective intellectual production. In addition, many systems of cooperative innovation rely on a set of beliefs that are not consistent with patenting
and enforcing patents. This puts the group at a disadvantage when
210. For a similar line of reasoning in the context of FRAND promises, see Jorge
Contrerras, supra note 200 (manuscript at 11-12).
211. This approach was adopted by Judge Robart in his modification of the GeorgiaPacific factors in the Microsoft v. Motorola decision. See, e.g., Contreras, supra note 191.
Thomas Cotter and Norman Siebrasse have suggested an alternative way of thinking
about the ex ante hypothetical negotiation, focusing on the ex ante contingent value of
patented technologies that might be adopted as part of a standard and, if adopted, would
have higher value. This approach provides a way of thinking about a reasonable ex ante
negotiation that is not based on holdup value, but rather on the incremental contribution
that the patented technology makes to the standard as compared with the next best
alternative technology. See, e.g., Norman Siebrasse, Comments on Sidak Part 2: The Ex
Ante Contingent Value Approach (Siebrasse), COMP. PAT. REMEDIES BLOG (Dec. 6, 2013),
http://comparativepatentremedies.blogspot.com/2013/12/comments-on-sidak-part-2-ex-an
te.html; Thomas F. Cotter, Comments on Sidak, Part 3: Should a FRAND Royalty be
Higher than a Reasonable Royalty? (Cotter), COMP. PAT. REMEDIES BLOG (Dec. 18, 2013),
http://comparativepatentremedies.blogspot.com/2013/12/comments-on-sidak-part-3-shouldfrand.html. Comparing and evaluating alternative approaches and their implications for
contexts of cooperative innovation is left for further discussion.
212. This principle may actually make it even harder for the group to obtain their own
patents, making defensive patenting and licensing strategies harder. While I think that
the benefits of the principle will likely outweigh the costs in the contexts I am focusing on,
this potential cost should not be ignored.
2014]
COOPERATIVE INNOVATION
1117
they are forced to pay for their use of a third party’s technology but
cannot charge for the use of their technology.
Second, the group will have limited ability to extract payment for
the use of its non-patentable contributions by either a group member
or a third party who obtains a patent covering an invention that was
discovered through the use of the group’s knowledge. As a result of
these limits, third parties or defectors from the group may be able to
capture the benefits of group production for free while extracting
payment for their own incremental contributions. Where a member of
the group is able to defect from the group, that is, patent and privately benefit from an invention that builds on the efforts of the group,
this erodes the group members’ trust that other members of the
group will adhere to its norms of sharing and reciprocity. Where the
incremental improvement is made and patented by a third party, the
result is less harmful to group dynamics but still increases the cost of
innovation for the group and reduces the incentives of the group
members to freely and openly share their ideas. If the patent can be
used to block the future efforts of the group or to extract rents from
the group through licensing, benefit-sharing is even more skewed,
and the transaction costs for participants in the cooperative project
increase. This problem is not unique to systems of cooperative innovation, but it is particularly harmful to them because it disrupts the
non-market mechanisms upon which they rely.
While existing patent law doctrines of derivation, joint inventorship and prior art significantly limit the ability of an individual to
patent inventions that utilize, borrow from, or build on group discoveries, these protections do not adequately address the following problems that patents create for group investments in intellectual production.213 One of the biggest limits of these existing patent law doctrines is their failure to remove the ex ante incentive problems that
are created when individual members of the group see chances to
limit their sharing of knowledge so as to increase the chances of their
own individual discovery and patenting for monetary gain. These
doctrines also do not address and protect many of the valuable contributions made by group members. Member contributions often take
the form of tools that increase the chances of invention but are not
themselves either patentable or prior art for the resulting invention.
In Foldit, for example, participants develop software programs that
help to increase scores in the game, and these programs, along with
observations about how to solve problems in game play, are some-
213. See supra Part II (discussion of limits of these doctrines in cooperative
innovation contexts).
1118
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
times shared among the players.214 Although the software programs
and strategies may increase the likelihood that a patentable discovery about proteins will be made, they do not themselves describe such
a discovery and therefore do not constitute prior art for such a discovery. The people contributing the programs are not joint inventors
of the resulting discovery. Moreover, although providing valuable information that increases the chances of a patentable discovery, it is
unlikely that the software programs or strategies for gameplay will
themselves be patentable discoveries.
Even where the doctrines do apply, they may be difficult and costly for the group to utilize. Where the group wants to challenge thirdparty inventorship claims on the grounds of joint inventorship or derivation, it is forced to undertake the costs of pursuing these claims
and may find it difficult to satisfy the evidentiary requirements under patent law despite the value of its contributions. The massively
distributed nature of the group and the importance of keeping transaction costs low similarly limit the usefulness of these doctrines. In
addition, where the group wants the value of its contributions reflected in negotiations for the use of third-party patents, it may be
forced to explore patenting, and this can be inconsistent with and potentially undermine group beliefs and values. Additional help from
the law is needed to fill these gaps and address these limits. In response to some of these gaps, I am suggesting that where a patented
invention benefits from the knowledge provided by the group, the
benefits received from the group should be considered and sometimes
limit the remedy that a patent owner receives.
This principle, like the first, builds on and finds support in existing proposals to address patent holdup by limiting the ability of
patent holders to obtain injunctive relief and by restricting what
can be included in the calculation of damages.215 It is also consistent
with recent court decisions in the context of standard-essential
patents that seek to limit patent damages to the value of the technology and exclude the value conferred by the standardization of this
technology.216 The principle goes further than these existing ap214. See Public-Shared Recipes, FOLDIT, https://fold.it/portal/recipes/public (last visited
Mar. 22, 2014).
215. See, e.g., Cotter, supra note 19, at 1174-87 (explaining how patent law should
play a role in responding to, or enabling private parties to avoid, patent holdup); Lemley
& Shapiro, supra note 195 (explaining how patent law should be modified to respond
to problems of patent holdup and royalty-stacking, including limits on injunctive relief
and damages).
216. See Microsoft Corp. v. Motorola, Inc., No. C10-1823JLR, 2013 WL 2111217, at *98101 (W.D. Wash. Apr. 25, 2013). There are also proposals designed to address patent
holdup in context of standard-setting. Jay P. Kesan & Carol M. Hayes, FRAND’s Forever:
Standards, Patent Transfers, and Licensing Commitments, 89 IND. L.J. 231 (2014) (suggesting combination of contract and property principles to govern FRAND commitments);
2014]
COOPERATIVE INNOVATION
1119
proaches, however. It is not limited to technology standards and patents covering those standards, but instead applies when a group
adopts an invention as part of its research platform and this adoption
and use confers value on the invention. Where the value arising from
widespread adoption and use is large, this should be a factor weighing against injunctive relief, and this value should be excluded from
royalty calculations to the extent that the value reflects simply its
use as a standard. The principle also requires courts to look at
whether knowledge contributions from the group were instrumental
in leading to the invention. If the knowledge contribution was substantial, this should weigh against injunctive relief and should be
reflected in reduced royalties. In this way courts will play a role in
enforcing benefit-sharing and promoting fairness, making members
of the group more willing to participate and freely contribute their
work to the group.
This principle would play the strongest role in situations where
members of a cooperative innovation project, such as Linux or Foldit,
defect with an invention, patent it, and either seek to assert it
against the group or to assign it to a third party who then asserts it
against the group. In this case the combination of benefits from group
production and public interest in continued group use of the invention would weigh against injunctive relief and would substantially
limit royalties where the invention relates closely to the intellectual
production of the group. This principle would also extend, although
with less force, to limit patent damages where a third party who is
not a member of the group benefits from the intellectual production
of the group and/or seeks to hold up the production of the group. In
the Linux example, Novell and its assignees would have found it harder to enforce their patents against the open source community, and
knowing this, the open source community would have been less worried, and the patents would have been less valuable. Novell would
therefore have had less incentive to defect from the community in the
Merges & Kuhn, supra note 195 (arguing for a standards estoppel doctrine to check for
good-faith behavior by the patentee). As well, there are proposals designed to protect areas
of innovation that rely on free sharing of information. Strandburg, supra note 150. Alternative approaches have included improving incentives of private parties to agree through
limits on injunctive relief, mandatory arbitration, and other mechanisms for improving
private ordering. See, e.g., Jorge L. Contreras, Fixing FRAND: A Pseudo-Pool Approach to
Standards-Based Patent Licensing, 79 ANTITRUST L.J. 47 (2013) (adapting patent pool approach to standard-setting organizations); Mark A. Lemley & Carl Shapiro, A Simple Approach to Setting Reasonable Royalties for Standard-Essential Patents, 28 BERKELEY TECH.
L.J. 1135 (2013) (discussing a mandatory arbitration mechanism where owner of standardessential patent and standard implementer do not agree on FRAND terms); Timothy
Simcoe, Governing the Anti-Commons: Institutional Design for Standard Setting Organizations (July 1, 2013) (unpublished manuscript) (on file with the National Bureau of Economic Research) (examining SSO practices and debates using Elinor Ostrom’s selfgoverning common pool framework).
1120
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
first place. Microsoft might have found its patent threats and licensing
strategies less effective, reducing the need for defensive patenting.
This principle could limit incentive problems in Foldit as well.
Foldit members will have less incentive to use patents opportunistically and will be more willing to continue to volunteer their efforts
without compensation if they all know that the contribution of ideas
by the group will factor heavily in determining remedies for any invention made and patented by a group member. This principle will
also limit the ability of third parties with inventions that build upon
or are heavily utilized by the group to the benefit of the patent owner
to obtain injunctive relief and damages from these infringing uses.217
The key challenge in employing this principle is to provide the
courts with tools or specific guidelines for how to identify and value
these contributions. Indeed, concerns with the ability of the courts to
measure the contribution of an invention to a larger product or project have stalled changes in patent remedies in the past. Courts are
now confronting and finding ways to navigate analogous measurement problems, however, in the context of determining reasonable
royalties for standard-essential patents.218 This provides a starting
point for courts seeking to take into account the value of group
knowledge contributions and group use when assessing patent damages.219 In the hypothetical negotiation of a license to a patented invention, the court would consider not just the value of the patented
technology but also the value of the intellectual contributions of the
cooperative innovation community. In this hypothetical exchange, the
contributions from the group would reduce the royalty owed to the
patent owner. In addition, royalties to the patent owner would exclude any value that the invention has as a result of its widespread
adoption by the group. This hypothetical licensing negotiation should
reflect the reduced bargaining power of the patent owner where injunctive relief is unlikely, and the increased bargaining power of the
group where they are providing value through their activities.
217. This approach could also address the practical limits of prior art and derivation
doctrines by presuming that the community has some rights over the invention, leaving the
patent owner with the burden of showing that the invention did not benefit from
community production and did not derive its value from adoption by the community as a
standard technology.
218. See, e.g., Cotter, supra note 19, at 1180-88; Mark A. Lemley & Carl Shapiro,
Patent Holdup and Royalty Stacking, 85 TEX. L. REV. 1991, 1994 (2007); Mark A. Lemley,
Ten Things To Do About Patent Holdup of Standards (and One Not To), 48 B.C. L. REV.
149, 153 (2007).
219. How royalties should be determined in this context is an area in which the law
and commentary are in flux. See, e.g., J. Gregory Sidak, The Meaning of FRAND, Part I:
Royalties, 9 J. COMPETITION L. & ECON. 931 (2013); Cotter, supra note 211.
2014]
COOPERATIVE INNOVATION
1121
3. Third Principle
The third principle is to reinforce reciprocity in free, open systems
of innovation.220 This principle would treat both formal and informal
rules governing open access and sharing of ideas and discoveries as
commitments by participants to share their own contributions with
other participants in the project. Where discoveries relevant to free
and open source projects are made by members of the project, free
and open access and use of the discoveries by the group should be
presumed unless the members of the group have explicitly agreed
otherwise. In this way the law reinforces informal rules and norms of
reciprocity and reduces the payoffs from defecting from group
norms.221 Provided that people know about this change in the law in
advance, it makes it easier for members of the group to commit to
behavior in compliance with these rules and norms, in addition to
facilitating ex ante bargaining. The principle would also apply to limit relief for industry members who know about and maybe even benefit from the use of their invention by the cooperative innovation project unless these parties have taken reasonable precautions to put
the community on notice of their proprietary rights before the invention is in use. Where third parties develop inventions that rely on
the work done by an open source project and relate to the open source
project, this principle would weigh heavily against allowing this third
party to block the use of the discovery by the group, either through
an injunction or royalties that are not feasible for group participants
to pay. The impact of this principle should become larger the closer
the relationship of the patent owner to the group and the greater the
benefit to the patent owner of the group’s activities.
Open access and reciprocity play critical roles in both open source
software and crowd science projects.222 In cooperative systems such as
these, discoveries that benefit from the productive efforts of the
group should be available for use by the group at either no cost or,
where the discovery is made outside of the project, at a cost that is
reasonable in light of the competing interests and investments of the
group and the patent owner. Efforts have been made to build reciprocity into open source communities through the use of licenses such
220. For thoughts on motivating reciprocity, see, for example, Stephen Leider
et al., Directed Altruism and Enforced Reciprocity in Social Networks, 124 Q.J. ECON.
1815 (2009).
221. For a related idea on how open source systems can reduce opportunism in the
context of open standards, see, for example, Vetter, supra note 96.
222. See, e.g., Jane Kaye et al, Data Sharing in Genomics—Re-Shaping Scientific
Practices, 10 NATURE REVS. GENETICS 331 (2009); Jennifer C. Molloy, The Open Knowledge
Foundation: Open Data Means Better Science, 9 PUB. LIBR. SCI. BIOLOGY 1.
1122
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
as the General Public License and the Reciprocal Public License.223
These contracting approaches are generally limited to the world of
free and open source software, however, and questions about the enforceability and transaction costs associated with these licenses remain.224 Large public funding agencies such as the National Institute
of Health have sought to build reciprocity into public funding requirements, but again with high transaction costs and questionable
enforceability.225 The principle that I propose can complement these
efforts. It can also expand upon them, applying in situations where
there is no contractual privity and without depending upon the organizational structure that underpins open source software production or publicly funded research.
In the Foldit context, this principle would limit the ability of both
members of Foldit and third parties to impede the free use of scientific discoveries by Foldit members for the non-commercial purposes
of solving Foldit puzzles. It would thus operate in ways that are
analogous both to proposed concepts of patent fair use and proposals
for research use exemptions, but in both cases limited to the context
of crowd science.226 This principle would be most valuable in sustaining non-mainstream, volunteer-based open source projects, which are
among the most vulnerable to transaction costs and the least able to
engage in defensive patenting activities.
All three of these principles draw from a rich body of research that
critiques the lack of safe harbors within patent law for publicly beneficial uses of patented technology, particularly where the patented
technologies have been created through the use of public funds.227
The patent literature includes a number of carefully constructed proposals for research use exemptions and patent fair use.228 Instead of
223. See, e.g., Reciprocal Public License 1.5 (RPL-1.5), OPEN SOURCE INITIATIVE (July
15, 2007), http://opensource.org/licenses/rpl-1.5; see also Tom Hall, Open Source—
Reciprocal Licenses, TECH L. GUY BLOG (Sept. 18, 2013, 12:21 PM), http://techlawguy.blog
spot.com/2013/09/p-margin-bottom-0.html (including a list of reciprocal open source
licenses as determined by GNU).
224. See, e.g., Gomulkiewicz, supra note 107 (examining the most recent version of
general public license that tries to deal with issues such as patents and digital rights
management); Sapna Kumar, Enforcing the GNU GPL, 2006 U. ILL. J.L. TECH. & POL’Y 1.
225. See, e.g., Arti K. Rai & Bhaven N. Sampat, Accountability in Patenting of
Federally Funded Research, 30 NATURE BIOTECH. 953 (2012).
226. See, e.g., Joshua I. Miller, Towards a Doctrine of Fair Use in Some of Patent Law,
2 INTELL. PROP. BRIEF 56 (2011); O’Rourke, supra note 151; Strandburg, supra note 150.
227. See, e.g., Strandburg, supra note 150 (arguing for the use of defenses and
exemptions from infringement as a way of responding to the fact that different uses of
patented technology can have different social costs and benefits).
228. See id.; see also Rochelle Dreyfuss, Protecting the Public Domain of Science: Has
the Time for an Experimental Use Defense Arrived?, 46 ARIZ. L. REV. 457, 464 (2004);
Rebecca S. Eisenberg, Patents and the Progress of Science: Exclusive Rights and
Experimental Use, 56 U. CHI. L. REV. 1017 (1989) (proposing research-use defense that
2014]
COOPERATIVE INNOVATION
1123
focusing on the types of uses, however, the focus here is on preserving non-market mechanisms for cooperative innovation. Moreover,
the principles are not intended as rules that dictate particular patent
remedies, but rather as mechanisms for expanding the range of interests that courts are required to think about when fashioning patent remedies.229 These principles require courts to systematically
take into account the broader costs of patents on socially beneficial
cooperative innovation when fashioning patent remedies.230 Ideally,
the principles would also be adopted by agencies such as the Federal
Trade Commission, the Department of Justice and the International
Trade Commission, and used to guide future changes in the patent
statute made by legislators.231 The principles provide a focal point for
all of these patent policymakers to use when fashioning their responses to patent problems in contexts of cooperative innovation.
V. CONCLUSION
“Many ideas grow better when transplanted into another mind than
the one where they sprang up.”—Oliver Wendell Holmes
Cooperative innovation can bring together diverse perspectives
and ideas and harness underutilized human resources in new ways
to solve previously intractable scientific problems. In some cases, systems of cooperative innovation that rely at least partially on nonmarket mechanisms may complement existing modes of marketdriven innovation. In other cases, they may challenge incumbent systems of intellectual production and intensify competition in ways
that accelerate scientific and technological progress. The potential of
distinguishes between situations requiring payment and those not requiring payment, with
a focus on protecting robust domain for basic research uses); O’Rourke, supra note 151, at
1204-05; Katherine J. Strandburg, What Does the Public Get? Experimental Use and the
Patent Bargain, 2004 WIS. L. REV. 81.
229. But see, e.g., Michael A. Carrier, Cabining Intellectual Property Through a
Property Paradigm, 54 DUKE L.J. 1 (2004) (stating the idea that propertization of patent
rights can lead to the narrowing of IP, providing for limits based on development,
necessity, and equity).
230. Efforts have been made to provide general principles or principles for determining
patent remedies. See, e.g., Thomas F. Cotter, Patent Remedies and Practical Reason, 88
TEX. L. REV. SEE ALSO 125 (2010); Golden, supra note 18. I am not providing principles
that govern the application of remedies generally, but rather specific principles that
require policymakers to consider the effects of patent remedies on the viability of
cooperative innovation.
231. The patent statute provides little guidance for courts in how they determine
patent remedies, leaving the implementation primarily to common law. I focus on the role
of the courts in responding to cooperation through their approach to patent remedies,
utilizing the ability of courts to operate through what Burk and Lemley have termed
“policy levers.” See Dan L. Burk & Mark A. Lemley, Policy Levers in Patent Law, 89 VA. L.
REV. 1575 (2003).
1124
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1067
these emerging systems of innovation may be limited, however, by
patent roadblocks.
This Article has shown that patents may sometimes interfere with
important non-market mechanisms that sustain systems of cooperative innovation by increasing both the costs of participating and the
benefits of defecting from these systems. While some forms of open,
collaborative innovation persist in the face of patent threats, they do
so only at great cost and with uncertainty about their future sustainability. To these costs and uncertainties we must add the social loss
from potentially valuable innovation paths foreclosed.
In response to these concerns, I suggest that courts and other patent policymakers need to pay more attention to the ways in which
patent law may systematically disadvantage cooperative innovation.
As a starting point, I take advantage of current areas of opportunity
in judicial thinking about patent remedies to propose modest changes
in the ways that courts implement patent remedies where nonmarket mechanisms of cooperation are important. Three principles
are provided to guide courts in their determinations of patent remedies with cooperative contexts in mind. The principles are intended to
support cooperation where patents are not already accomplishing
this goal, and only to the extent that patents are impeding this goal.
Limiting the negative incentive effects of patents on non-market
mechanisms in this way may give cooperative innovation systems
chances to complement and to compete against market-driven systems of innovation. In addition, a greater legal commitment to balancing individual interests with broader public interests in cooperative innovation may improve the relationships that cooperative innovative communities have to patent law. Implementing patent law in
ways that take the interests and needs of these communities into account may move us closer towards a patent system that can accommodate alternative paradigms of innovation.
AGGREGATING DEFENDANTS
GREG REILLY*
ABSTRACT
No procedural topic has garnered more attention in the past fifty years than the class
action and aggregation of plaintiffs. Yet, almost nothing has been written about aggregating
defendants. This topic is of increasing importance. Recent efforts by patent “trolls” and BitTorrent copyright plaintiffs to aggregate unrelated defendants for similar but independent
acts of infringement have provoked strong opposition from defendants, courts, and even
Congress. The visceral resistance to defendant aggregation is puzzling. The aggregation of
similarly situated plaintiffs is seen as creating benefits for both plaintiffs and the judicial
system. The benefits that justify plaintiff aggregation also seem to exist for defendant aggregation—avoiding duplicative litigation, making feasible negative-value claims/defenses,
and allowing the aggregated parties to mimic the non-aggregated party’s inherent ability to
spread costs. If so, why is there such resistance to defendant aggregation?
Perhaps, contrary to theoretical predictions, defendant aggregation is against defendants’ self-interest. This may be true in certain types of cases, particularly where the plaintiff’s claims would not be viable individually, but does not apply to other types of cases, particularly where the defendants’ defenses would not be viable individually. These latter cases
are explained, if at all, by defendants’ cognitive limitations. In any event, defendant selfinterest does not justify systemic resistance to defendant aggregation. Likewise, systemic
resistance is not warranted because of concerns of weak claims or unsympathetic plaintiffs,
the self-interest of individual judges handling aggregated cases, or capture by defendant
interests. This Article proposes that to obtain the systemic benefits of defendant aggregation
and overcome the obstacles created by defendant and judicial self-interest, cognitive limitations, and capture, defendant aggregation procedures should use non-representative actions,
provide centralized neutral control over aggregation, and limit aggregation to common issues. This Article concludes with a modified procedure to implement these principles: interdistrict related case coordination.
I. INTRODUCTION ..................................................................................................
II. OVERVIEW OF DEFENDANT AGGREGATION ........................................................
A. Aggregation Procedures .............................................................................
B. Two Types of Defendant Aggregation ........................................................
C. The Argument for Aggregating Similarly Situated Defendants................
D. Resistance to Aggregating Similarly Situated Defendants .......................
1. Patent Litigation ..................................................................................
2. Trademark and Copyright Litigation ..................................................
3. Telecommunications Cases ..................................................................
4. Other Examples....................................................................................
E. The Puzzle of Defendant Aggregation ........................................................
III. DEFENDANTS’ OPPOSITION TO DEFENDANT AGGREGATION...............................
A. Creating Additional Litigation for Defendants .........................................
B. Cost-Differentials .......................................................................................
C. Substantive Effects ....................................................................................
1012
1015
1017
1020
1021
1023
1023
1025
1028
1029
1029
1030
1030
1032
1033
* Harry A. Bigelow Teaching Fellow and Lecturer in Law, University of Chicago
Law School. Thanks to Daniel Abebe, Douglas Baird, Omri Ben-Shahar, Lisa Bernstein,
Bob Bone, Vince Buccola, Tony Casey, Adam Chilton, Zachary Clopton, Lee Fennell, Brian
Fitzpatrick, Roger Ford, Brad Greenberg, James Grimmelmann, Vinay Harpalani, William
H.J. Hubbard, Aziz Huq, Ken Klein, Genevieve Lakier, Saul Levmore, Jonathan Masur,
Arthur R. Miller, Ariel Porat, John Rappaport, David Schwartz, Allan Stein, Crystal Yang,
Steve Yelderman, and participants at the Sixth Annual Junior Federal Courts Workshop,
the Chicago Junior Faculty Workshop, the 2013 Intellectual Property Scholars Conference,
and faculty workshops at the University of Chicago Law School, California Western School
of Law, and Valparaiso University Law School, for helpful discussions and comments.
1012
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
1. Asymmetric Preclusion.........................................................................
2. Jury and Judicial Confusion ...............................................................
3. Autonomy .............................................................................................
4. Tool to Achieve Other Substantive Objectives ......................................
D. Cognitive Limitations ................................................................................
1. Involuntary Groups ..............................................................................
2. Self-Serving Bias .................................................................................
3. Loss Aversion .......................................................................................
E. Summary and the Role of Selection Effects ...............................................
IV. DEFENDANT AGGREGATION FROM THE SYSTEMIC PERSPECTIVE ......................
A. Cost-Differentials, Weak Claims, and Unsympathetic Plaintiffs ..............
B. Substantive Effects, Fairness, Autonomy, and Due Process ......................
C. Capture ......................................................................................................
D. Judicial Self-Interest .................................................................................
V. OVERCOMING RESISTANCE TO DEFENDANT AGGREGATION ...............................
A. Principles of a Defendant Aggregation Procedure .....................................
1. Non-Representative Aggregation..........................................................
2. Centralized Control Over Aggregation.................................................
3. Issue-Only Aggregation ........................................................................
B. Inter-District Related Case Coordination ..................................................
1. Overview of Inter-District Coordination ..............................................
2. Mechanics of Inter-District Coordination ............................................
C. Potential Obstacles to Inter-District Coordination ....................................
1. Practical Obstacles: Free Riding ..........................................................
2. Statutory and Constitutional Obstacles ..............................................
VI. CONCLUSION .....................................................................................................
1033
1035
1036
1037
1039
1039
1041
1042
1043
1045
1046
1048
1050
1051
1052
1053
1053
1054
1055
1056
1056
1058
1061
1061
1063
1065
I. INTRODUCTION
Suppose that Netflix offered a one-month free trial before a customer was billed,1 but actually added a small additional charge to the
first bill to cover tax owed for this “free” trial. If customers wanted to
sue Netflix for deceptive trade practices, it would be unsurprising if
they sought to sue collectively via voluntary joinder, a class action, or
multi-district litigation and, absent a contractual provision, were
permitted to do so.2 But what if Netflix discovered widespread violations of a provision of the Terms of Use prohibiting users from sharing their passwords with people other than “household members?”3
1. See Netflix Terms of Use, NETFLIX, https://signup.netflix.com/TermsOfUse (last
visited July 25, 2014).
2. This is similar to AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740 (2011), where
there was little doubt aggregation would have been possible absent a contractual provision
requiring individual arbitration; indeed, it is exactly the likelihood of aggregation that led
to the inclusion of the contractual provision. See Myriam Gilles, Opting Out of Liability:
The Forthcoming, Near-Total Demise of the Modern Class Action, 104 MICH. L. REV. 373,
414 (2005). See generally Myriam Gilles & Gary Friedman, After Class: Aggregate Litigation in the Wake of AT&T Mobility v Concepcion, 79 U. CHI. L. REV. 623 (2012) (discussing
Concepcion and its aftermath).
3. Netflix’s terms of use permit sharing of passwords with “household members” but
are actually ambiguous about sharing with non-household members. See NETFLIX, supra
note 1. The hypothetical assumes that the Terms of Use include choice of law, choice of
2014]
AGGREGATING DEFENDANTS
1013
Could Netflix aggregate its breach of contract claims and litigate
against the consumers collectively? Both experience and most people’s gut reactions suggest no.
This raises a puzzle. Why are multiple people harmed by the same
defendant in similar ways considered an appropriate litigation group
but multiple people harming the same defendant in similar ways are
not? In theory, aggregation of similarly situated defendants offers the
same benefits as aggregation of similarly situated plaintiffs: avoiding
duplicative and potentially inconsistent litigation; promoting optimal
deterrence by making negative-value claims (for plaintiffs) or defenses (for defendants) viable; and encouraging resolution on the merits,
not litigation costs, by allowing the aggregated party to spread its
costs in the same way that the non-aggregated party inherently can.4
Yet, efforts to aggregate similarly situated defendants across several substantive areas have sparked widespread and vehement opposition from defendants, courts, and policymakers. Courts have rejected joinder where “the plaintiff does no more than assert that the defendants ‘merely commit[ted] the same type of violation in the same
way’ ”5 and have even described efforts to aggregate similarly situated defendants as a “gross abuse of procedure.”6 Even Congress has
weighed in, with one member describing defendant aggregation as an
“abusive practice.”7
Defendant aggregation has been particularly controversial in patent litigation, where courts and defendants have resisted efforts by
patent holders, particularly controversial patent assertion entities or
“trolls,” to join unrelated companies with similar but competing
products alleged to infringe the same patent.8 When the plaintifffriendly Eastern District of Texas permitted it, Congress passed
a special statutory provision prohibiting joinder or consolidation for
trial of unrelated accused infringers.9 Efforts by copyright holders
of pornographic films to join dozens or hundreds of users of the internet file sharing protocol “BitTorrent” have encountered similar
forum, and consent to jurisdiction provisions, though the actual terms of use include an
individual arbitration provision like in Concepcion. See id.
4. See Assaf Hamdani & Alon Klement, The Class Defense, 93 CALIF. L. REV. 685,
696-99 (2005); see also David Rosenberg, Class Actions for Mass Torts: Doing Individual
Justice by Collective Means, 62 IND. L.J. 561, 563-64, 570-72 (1987) (describing rationales
for aggregating plaintiffs).
5. Digital Sins, Inc. v. Does 1–245, No. 11Civ.8170(CM), 2012 WL 1744838, at *2
(S.D.N.Y. May 15, 2012) (alteration in original).
6. Nassau Cnty. Ass’n of Ins. Agents, Inc. v. Aetna Life & Cas. Co., 497 F.2d 1151,
1154 (2d Cir. 1974).
7. David O. Taylor, Patent Misjoinder, 88 N.Y.U. L. REV. 652, 703 (2013) (quoting
157 CONG. REC. H4426 (daily ed. June 22, 2011) (statement of Rep. Bob Goodlatte)).
8. See generally id.
9. See 35 U.S.C. § 299 (Supp. V 2011); see generally Taylor, supra note 7.
1014
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
resistance,10 as has the music industry’s own copyright fight against
internet file sharers and DirecTV’s campaign against individual
signal “pirates.”11
Unlike the exhaustive scholarly consideration of plaintiff aggregation,12 only a handful of articles have addressed defendant aggregation.13 Entirely overlooked in this limited scholarship is the basic
puzzle of defendant aggregation—that is, why aggregation of similarly situated defendants has been so controversial in practice despite
its theoretical benefits.14 This Article tackles this puzzle. The disconnect between the benefits defendant aggregation offers and the opposition it has encountered may result from an assumption that defendant aggregation must take the procedural form most familiar
from plaintiff aggregation—the class action—and a belief that the
class action is problematic when absent defendants, not absent plaintiffs, are to be bound by a class judgment. This Article, however, is
not an analysis of defendant class actions. Rather, it first steps back
from the mechanics of aggregation and asks whether collective resolution of claims against similarly situated defendants is desirable in the
first place. Only then does it address the appropriate procedural mechanism for achieving this collective resolution, whether already existing or
newly developed for the specific needs of multi-defendant cases.
The Article proceeds in four main parts. Part II describes the
background of aggregation generally and the puzzle of defendant aggregation specifically. Part III looks at the puzzle from the defendant
perspective and Part IV from the societal perspective. Defendants’
resistance to being aggregated is the easier of the two to understand.
In many contexts defendants have strategic reasons to oppose being
aggregated, including avoiding litigation that otherwise would not be
brought and resisting the choice of the presumably self-interested
plaintiff. On the other hand, defendants often overestimate the
strength of these strategic concerns, and these strategic reasons do
not even exist in certain contexts—particularly where the stakes are
10. See, e.g., Sanne Specht, Judge Dismisses Film Company's Lawsuit Against Local
Defendants, MAIL TRIBUNE (May 14, 2013, 2:00 AM), http://www.mailtribune.com/apps/
pbcs.dll/article?AID=/20130514/NEWS/305140311; Claire Suddath, Prenda Law, the Porn
Copyright Trolls, BLOOMBERG BUSINESSWEEK (May 30, 2013), http://www.businessweek.
com/articles/2013-05-30/prenda-law-the-porn-copyright-trolls#p1.
11. See infra Part III.D.
12. PRINCIPLES OF THE LAW OF AGGREGATE LITIGATION § 1.02 reporters’ note to cmt. a,
at 15 (2009).
13. But see infra notes 20–23.
14. Prior works either focus on the theoretical benefits to defendants and ignore the
practical opposition from defendants, see Hamdani & Klement, supra note 4, or focus on
the practical opposition from defendants and ignore the theoretical benefits to defendants,
see Sean B. Karunaratne, Note, The Case Against Combating BitTorrent Piracy Through
Mass John Doe Copyright Infringement Lawsuits, 111 MICH. L. REV. 283 (2012).
2014]
AGGREGATING DEFENDANTS
1015
too low to justify an individual defense—which are explained, if at
all, by defendants’ cognitive limitations.
Turning to the more interesting and important question of why
aggregation of similarly situated defendants has faced systemic resistance, Part IV first concludes that any negative effects defendants
suffer from being aggregated should not trouble courts and policymakers, as they neither distort the substantive remedial scheme nor
raise fairness concerns. They explain systemic resistance, if at all,
based on “capture” by defendants’ interests. Although courts and policymakers may use de-aggregation as a way to police weak or disfavored substantive claims, these problems are better addressed
through other procedural mechanisms or substantive reform. Finally,
while managing the complexities created by defendant aggregation
may challenge an individual judge’s self-interest, it is not worse for the
judicial system than repetitive or overlapping dispersed litigation.
Part V identifies three core features of an optimal procedure for
aggregating similarly situated defendants: (1) a non-representative
structure (i.e., not class actions); (2) control by a centralized body, not
the parties or individual judges; and (3) aggregation only of common
issues, not cases. Part V concludes by sketching a mechanism to implement these principles: inter-district related case coordination,
which is a hybrid of existing multi-district litigation and the related
case procedures many federal district courts use to manage cases
filed within a single district.
II. OVERVIEW OF DEFENDANT AGGREGATION
Scholars and policymakers have paid surprisingly little attention
to aggregation of defendants in litigation.15 To some extent, this is
because complex litigation scholarship focuses overwhelmingly on the
class action and largely overlooks other means for aggregating parties, like joinder (Federal Rules of Civil Procedure 19 and 20), consolidation (Federal Rule of Civil Procedure 42), and multi-district litigation (28 U.S.C. § 1407).16 Recent efforts to aggregate defendants
largely have been via these other procedural devices, not defendant
class actions,17 and therefore have escaped the notice of many classaction-focused scholars. Moreover, efforts to aggregate similarly situ15. Hamdani & Klement, supra note 4, at 689, 696.
16. See Robin J. Effron, The Shadow Rules of Joinder, 100 GEO. L.J. 759, 763-64
(2012) (noting that analysis of joinder is mentioned largely as an afterthought to the discussion of mass tort class actions); Emery G. Lee et al., The Expanding Role of Multidistrict Consolidation in Federal Civil Litigation: An Empirical Investigation 5-7 (July 10,
2010) (unpublished manuscript), available at http://papers.ssrn.com/sol3/papers.cfm?
abstract_id=1443375 (summarizing limited literature on multi-district litigation).
17. RICHARD L. MARCUS & EDWARD F. SHERMAN, COMPLEX LITIGATION: CASES AND
MATERIALS ON ADVANCED CIVIL PROCEDURE 382 (3d ed. 1998).
1016
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
ated defendants are uncommon compared to aggregation of similarly
situated plaintiffs,18 though they are becoming more frequent as
mass communications allow dispersed people to cause the same injury to the same person in the same way.19
The leading work on aggregating similarly situated defendants is
a paper by Professors Assaf Hamdani and Alon Klement in which
they argued that the same justifications for plaintiff aggregation apply to defendant aggregation and proposed a procedure by which defendants could initiate a defendant class action.20 A subsequent article by Nelson Netto, largely tracking Hamdani and Klement, contended that defendant classes were an important “functional device
for the defendants” and proposed mandatory defendant class actions
without opt-out.21 More recent work by Francis Shen developed in
more detail the theoretical case for defendant class actions, arguing that
the device can maximize social welfare.22 In contrast to this theoretical
work, several scholars, practitioners, and students have described efforts by plaintiffs in specific substantive contexts to aggregate defendants and the resistance of defendants, courts, and policymakers.23
These conflicting lines of scholarship neatly illustrate the puzzle
of defendant aggregation, which is laid out in more detail in this
Part. Section A provides a background of aggregation procedures
generally and Section B of defendant aggregation specifically. Section
C describes the theoretical argument for aggregation of similarly situated defendants, and Section D describes the widespread resistance
it has faced.
18. PRINCIPLES OF THE LAW OF AGGREGATE LITIGATION § 1.02 reporters’ note to cmt.
b(1)(A), at 17 (2009); id. § 1.02 reporters’ note to cmt. b(1)(B), at 23; STEPHEN C. YEAZELL,
FROM MEDIEVAL GROUP LITIGATION TO THE MODERN CLASS ACTION 58 (1987).
19. Hamdani & Klement, supra note 4, at 741.
20. Hamdani & Klement, supra note 4, at 699 (discussing the similarities between
numerous plaintiffs and numerous defendants and benefits to defendants from being
aggregated). A few previous articles touched on defendant aggregation while focusing on
plaintiff aggregation. Howard M. Erichson, Informal Aggregation: Procedural and Ethical
Implications of Coordination Among Counsel in Related Lawsuits, 50 DUKE L.J. 381,
401-08 (2000). Defendant class actions also have received some attention in student
notes and practitioner articles. See, e.g., Robert R. Simpson & Craig Lyle Perra, Defendant
Class Actions, 32 CONN. L. REV. 1319 (2000); Notes, Defendant Class Actions, 91 HARV. L.
REV. 630 (1978).
21. Nelson Rodrigues Netto, The Optimal Law Enforcement with Mandatory Defendant Class Action, 33 U. DAYTON L. REV. 59, 67, 97-101 (2007).
22. Francis X. Shen, The Overlooked Utility of the Defendant Class Action, 88 DENV.
U. L. REV. 73, 74 (2010).
23. See, e.g., Annemarie Bridy, Is Online Copyright Enforcement Scalable?, 13 VAND.
J. ENT. & TECH. L. 695 (2011) (copyright); Gideon Parchomovsky & Alex Stein, Intellectual
Property Defenses, 113 COLUM. L. REV. 1483 (2013) (intellectual property); Taylor,
supra note 7 (patent); Tracie L. Bryant, Note, The America Invents Act: Slaying Trolls,
Limiting Joinder, 25 HARV. J.L. & TECH. 687 (2012) (patent); Karunaratne, supra note 14,
at 286-87 (copyright).
2014]
AGGREGATING DEFENDANTS
1017
A. Aggregation Procedures
Federal procedural mechanisms have been the focal point for debate over defendant aggregation.24 The most famous of these mechanisms is the class action, which is a representative action by which
one or more people can sue on behalf of a class of similarly situated
individuals, provided the many threshold requirements of Federal
Rule of Civil Procedure 23 are satisfied: sufficient class size; common
question of fact or law; adequacy and typicality of the representative;
etc.25 Class actions bind all individuals that meet the class definition,
even if not actual parties to the litigation.26 Hundreds of class actions
are filed in federal court every month,27 and they have been subject to
substantial political, public, and scholarly scrutiny.28 Although Rule
23 purports to apply equally to plaintiffs and defendants, defendant
classes are virtually non-existent.29
Aggregation occurs less famously but more commonly by joining a
person as a party in the same lawsuit.30 Joinder actions only bind the
individuals to the litigation; at least in theory, each individual party
retains control over its separate claims, defenses, and rights.31 Federal Rule of Civil Procedure 19 describes when a plaintiff must join
other parties,32 while Rule 20 gives the plaintiff discretion to do so in
other circumstances.33 Rule 24 describes when a third-party can “intervene” or join itself to a lawsuit,34 while Rule 14 permits the defendant to “implead” or add a third party that is liable for the claim
made against the defendant.35 Finally, Rule 22 allows a party to “in24. Aside from formal aggregation in federal court, aggregation can occur: (1) in state
court under procedures similar to federal aggregation procedures; (2) inherently when corporations or voluntary associations (e.g., unions) litigate on behalf of their members; (3)
informally among counsel on the same side of related individual cases; and (4) via some
substantive remedial schemes like bankruptcy. See Erichson, supra note 20; Judith Resnik,
From “Cases” to “Litigation”, 54 LAW & CONTEMP. PROBS. 5, 23, 28, 38-39 (1991).
25. FED. R. CIV. P. 23.
26. PRINCIPLES OF THE LAW OF AGGREGATE LITIGATION § 1.02 cmt. b(1)(B), at 11
(2009).
27. See Emery G. Lee III & Thomas E. Willging, The Impact of the Class Action Fairness Act on the Federal Courts: An Empirical Analysis of Filings and Removals, 156 U. PA.
L. REV. 1723, 1750 (2008).
28. See Allan Erbsen, From “Predominance” to “Resolvability”: A New Approach to
Regulating Class Actions, 58 VAND. L. REV. 995, 997 n.1 (2005) (collecting scholarship).
29. See Thomas E. Willging et al., An Empirical Analysis of Rule 23 to Address the
Rulemaking Challenges, 71 N.Y.U. L. REV. 74, 119-20 (1996).
30. PRINCIPLES OF THE LAW OF AGGREGATE LITIGATION § 1.02 reporters’ note to cmt.
b(1)(A), at 16 (2009).
31. Id. § 1.02 cmt. b(1)(A), at 11; id. § 1.02 reporters’ note to cmt. b(1)(A), at 18.
32. FED. R. CIV. P. 19.
33. FED. R. CIV. P. 20.
34. FED. R. CIV. P. 24.
35. FED. R. CIV. P. 14.
1018
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
terplead” other parties that have claims exposing the party to doubleor multiple-liability.36 These joinder rules generally apply to both
plaintiffs and defendants.37 They normally are used to aggregate a
small number of parties; though joinder of tens, hundreds, or thousands is increasingly common, particularly in mass torts.38 Most
mass joinders involve plaintiffs.39
The third type of aggregation is coordination or consolidation of
individual lawsuits, which can occur for hearing, trial, or any other
purpose via Rule 42(a) if all of the cases are pending in the same federal district40 or only for pre-trial proceedings (e.g., discovery, summary judgment, etc.) via the multi-district litigation statute, 28
U.S.C. § 1407, if cases are pending in different federal districts.41
Aggregation occurs in two types of cases.42 First, many of the procedures only apply if individual adjudication could result in a problematic remedy, such as incomplete relief,43 exposure to double, multiple, or inconsistent obligations or liability,44 prejudice to a third
party,45 or effective determination of an entire group’s rights.46 This
type of remedy-driven aggregation has deep historical roots47 and is
generally uncontroversial.48
Second, several procedures permit aggregation based only on various levels of commonality among the claims. Permissive joinder
under Rule 20 requires a common “question of law or fact” and that
the claim “aris[e] out of the same transaction, occurrence, or series of
36. FED. R. CIV. P. 22.
37. PRINCIPLES OF THE LAW OF AGGREGATE LITIGATION § 1.02 reporters’ note to cmt.
b(1)(A), at 16 (2009).
38. Id. § 1.02 reporters’ note to cmt. b(1)(A), at 16-17.
39. Id. § 1.02 reporters’ note to cmt. b(1)(A), at 17.
40. FED. R. CIV. P. 42(a). District courts often have procedures for designating cases as
related and assigning them to the same judge or permitting intra-district transfer. See, e.g.,
U.S. DIST. CT. S.D. CAL. R. 40.1; U.S. DIST. CT. N.D. ILL. R. 40.4.
41. 28 U.S.C. § 1407(a) (2012); see also Lexecon Inc. v. Milberg Weiss Bershad Hynes
& Lerach, 523 U.S. 26 (1998) (requiring remand to original district for trial).
42. See Effron, supra note 16, at 764, 819-21 (proposing similar division).
43. See FED. R. CIV. P. 19(a)(1)(A) (mandatory joinder); see also FED. R. CIV. P.
14(a)(1) (interpleader).
44. See FED. R. CIV. P. 19(a)(1)(B)(ii) (mandatory joinder); FED. R. CIV. P. 22 (interpleader); FED. R. CIV. P. 23(b)(1)(A) (class action).
45. See FED. R. CIV. P. 19(a)(1)(B) (mandatory joinder); FED. R. CIV. P. 23(b)(1)(B)
(class action); FED. R. CIV. P. 24(a)(2) (intervention).
46. See FED. R. CIV. P. 23(b)(2) (class action).
47. See YEAZELL, supra note 18, at 16, 58, 65.
48. But see MARTIN H. REDISH, WHOLESALE JUSTICE: CONSTITUTIONAL DEMOCRACY
AND THE PROBLEM OF THE CLASS ACTION LAWSUIT (2009) (suggesting that mandatory class
treatment is inconsistent with individual autonomy but not challenging non-class aggregation in similar circumstances).
2014]
AGGREGATING DEFENDANTS
1019
transactions or occurrences . . . .”49 A class action under Rule 23(b)(3)
requires only a common question of law or fact but requires that the
common question(s) “predominate over any questions affecting only
individual members . . . .”50 Multi-district litigation under 28
U.S.C. § 1407, consolidation under Rule 42(a), and permissive intervention under Rule 24(b) permit aggregation based only on a common
question,51 though courts tend to require a strong overlap before
ordering aggregation.52
Commonality-driven aggregation was largely an innovation of the
1938 Federal Rules of Civil Procedure and amendments in the
1960s.53 Coordination or consolidation is generally uncontroversial,54
while permissive joinder has been described as both too restrictive for
requiring an overlap in operative facts and too expansive for allowing
mass joinders.55 Common question class actions under Rule 23(b)(3)
are far more controversial: they have been accused of creating substantial exposure that encourages settlement of even meritorious defenses;56 enriching plaintiffs’ lawyers with large fee awards while recouping trivial recoveries for class members;57 being inappropriate for
substantive claims that are positive-value and rife with individual
49. FED. R. CIV. P. 20(a)(1)(A)-(B); 4 JAMES WM. MOORE ET AL., MOORE’S FEDERAL
PRACTICE § 20.05[3] (3d ed. 2013).
50. FED. R. CIV. P. 23(a)(2); FED. R. CIV. P. 23(b)(3). Common-question class actions
must also be superior to other means for resolving the dispute. FED. R. CIV. P. 23(a)(1)-(4).
51. See 28 U.S.C. § 1407(a) (2012); FED. R. CIV. P. 24(b); FED. R. CIV. P. 42(a). The
multi-district litigation statute only expressly includes “common questions of fact,” 28
U.S.C. § 1407(a) (2012), but consolidation has occurred based primarily on common legal
questions or mixed questions of law and fact. See Stanley J. Levy, Complex Multidistrict
Litigation and the Federal Courts, 40 FORDHAM L. REV. 41, 47-48 (1971).
52. PRINCIPLES OF THE LAW: AGGREGATE LITIGATION § 1.02 reporters’ note to cmt.
b(2), at 27-28; Effron, supra note 16, at 789-804 (suggesting courts normally apply predominance standard when evaluating permissive joinder or consolidation).
53. See Robert G. Bone, Mapping the Boundaries of a Dispute: Conceptions of Ideal
Lawsuit Structure from the Field Code to the Federal Rules, 89 COLUM. L. REV. 1, 47-60
(1989); Resnik, supra note 24, at 16, 29-32.
54. See 9A CHARLES ALAN WRIGHT & ARTHUR R. MILLER, FEDERAL PRACTICE AND
PROCEDURE § 2381 (3d ed. 2008) (Rule 42(a)); Resnik, supra note 24, at 35 (multi-district
litigation). Some say that limiting multi-district litigation to pre-trial proceedings is too
constraining. See Resnik, supra note 24, at 35; see also Erichson, supra note 20, at 416.
55. Compare Douglas D. McFarland, Seeing the Forest for the Trees: The Transaction
or Occurrence and the Claim Interlock Civil Procedure, 12 FLA. COASTAL L. REV. 247,
265-66 (2011) (stating that the free joinder of parties is encouraged), with PRINCIPLES OF
THE LAW OF AGGREGATE LITIGATION § 1.02 reporters’ note to cmt. b(1)(A), at 16 (stating
that the rules of joinder “allow[] persons with related claims to join as coparties but [do]
not requiring them to”).
56. Resnik, supra note 24, at 16.
57. See, e.g., Arthur R. Miller, Of Frankenstein Monsters and Shining Knights: Myth,
Reality, and the “Class Action Problem”, 92 HARV. L. REV. 664, 679 (1979).
1020
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
issues, like mass torts;58 and threatening the autonomy of absent
class members.59
B. Two Types of Defendant Aggregation
Though the historical roots of group litigation lie in aggregation of
defendants—often an entire village with a shared obligation of harvest to a landlord or tithe to a church—aggregation of defendants has
been described as a “rarity” in modern times.60 Yet, litigation against
multiple defendants occurs every day in courts across the country:
employers and employees, corporations and their subsidiaries, principals and agents, co-conspirators, manufacturers and distributors,
insured and insurers, etc.61 Two types of defendant aggregation must
be distinguished.
When a plaintiff has a single claim to a single recovery for which
more than one person is potentially liable, whether jointly (e.g., coobligors on a contract), severally (e.g., joint tortfeasors), or in the alternative (e.g., a manufacturer and its component part supplier in a
product defect case), aggregation of defendants is common and uncontroversial.62 Remedy-driven aggregation procedures often apply.
Even if not, aggregation reflects the core purposes for which commonality-driven procedures were added to the federal rules: to prevent division of a single recovery into multiple lawsuits.63 Like the
historical roots of defendant aggregation, the defendants in these
cases tend to have a pre-existing relationship, often direct (e.g., a
manufacturer and its distributor) but at least indirect (e.g., two component part suppliers for the same manufacturer).
On the other hand, when a plaintiff has similar but independent
claims (each entitled to its own recovery) against multiple unrelated
but similarly situated defendants, aggregation is far less common
and more controversial. This Article focuses on these cases and uses
the phrases defendant aggregation or aggregation of similarly situated defendants to refer to them. Notably, aggregation of unrelated
plaintiffs with similar but independent claims is common and widely
accepted.64 This is not to deny the serious debate, extensive commentary, and numerous policy proposals surrounding aggregation of simi58.
59.
60.
61.
See Rosenberg, supra note 4, at 565-66 (summarizing criticism).
See REDISH, supra note 48, at 169-73.
YEAZELL, supra note 18, at 58, 135-36.
See, e.g., 7 CHARLES ALAN WRIGHT, ARTHUR R. MILLER & MARY KAY KANE,
FEDERAL PRACTICE AND PROCEDURE § 1657 (3d ed. 2001) (and notes therein).
62. See FED. R. CIV. P. 20(a)(2)(A); see also WRIGHT, MILLER & KANE, supra note
61, § 1654 (and notes therein).
63. See McFarland, supra note 55, at 260 n.66.
64. See Developments in the Law—Class Actions, 89 HARV. L. REV. 1318, 1322 (1976).
2014]
AGGREGATING DEFENDANTS
1021
larly situated plaintiffs, particularly via the class action. However,
that debate is over the proper scope (e.g., are class actions appropriate for mass tort claim?) or procedures (e.g., how to protect absent
class members or avoid agency problems with plaintiffs’ lawyers). On
the defendant side, the controversy is over the basic availability of
aggregation for similarly situated defendants.
C. The Argument for Aggregating Similarly Situated Defendants
Prior scholarship on defendant aggregation has concluded, “the
fundamental justification for consolidating plaintiff claims applies
with equal force to defendants.”65 Specifically, aggregation of similarly
situated plaintiffs is justified on three primary grounds that also seem
to apply to similarly situated defendants: eliminating duplicative litigation, making otherwise negative-value claims viable, and allowing
plaintiffs to match defendants’ inherent ability to spread costs.
First, a benefit of all aggregative devices is the elimination or reduction of repetitive litigation over the same or similar issues, which
wastes judicial and litigant resources66 and risks inconsistent judgments that could undermine public faith in the administration of justice.67 Multi-defendant cases are as likely to raise the same or similar
issues as multi-plaintiff cases; for example, multiple patent defendants can each challenge the validity of the patent and rely on the
same evidence and arguments.68
Second, plaintiff aggregation makes viable otherwise negativevalue claims, that is, “where the net expected recovery is [individually] small” and would not justify the cost of litigation, “but the total
extent of societal loss is large.”69 Aggregation insures that the defendant “pay[s] an amount equal to the losses caused by its wrong,”
thereby “secur[ing] the practical implementation of the substantive
law” and providing a deterrent that “reduces the willingness of the
defendant to engage in the illegal conduct that caused the harm in
65. Hamdani & Klement, supra note 4, at 689.
66. See WRIGHT & MILLER, supra note 54, R. 42(a) (consolidation); Richard D. Freer,
Avoiding Duplicative Litigation: Rethinking Plaintiff Autonomy and the Court’s Role in
Defining the Litigative Unit, 50 U. PITT. L. REV. 809, 813 (1989); Mary Kay Kane, Original
Sin and the Transaction in Federal Civil Procedure, 76 TEX. L. REV. 1723, 1728-30 (1998)
(permissive joinder); Susan M. Olson, Federal Multidistrict Litigation: Its Impact on Litigants, 13 JUST. SYS. J. 341, 341 (1988-89) (multi-district litigation); Rosenberg, supra note
4, at 563-64 (class actions).
67. See Freer, supra note 66, at 814; Edward F. Sherman, Aggregate Disposition of
Related Cases: The Policy Issues, 10 REV. LITIG. 231, 236-43 (1991).
68. See Bryant, supra note 23, at 704-05.
69. Richard A. Epstein, Commentary, The Consolidation of Complex Litigation: A
Critical Evaluation of the ALI Proposal, 10 J.L. & COM. 1, 6 (1990); see also Rosenberg,
supra note 4, at 563-64, 570-72.
1022
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
the first place.”70 Aggregation of defendants similarly allows the
plaintiff to eliminate some costs that cannot be spread over multiple
individual actions (e.g., filing fees, attending hearings and trial, and
taking and defending depositions). This cost-reduction will make otherwise negative-value claims viable, promoting optimal deterrence.
Moreover, the negative-value defense provides a “mirror-image”
justification for aggregation of similarly situated defendants.71 Even
defendants with meritorious defenses will settle if the cost of litigation exceeds the potential liability, and this threat of overpaying
could over-deter legitimate conduct ex ante.72 For example, many say
patent-assertion entities exploit their cheaper litigation costs to obtain settlement payments less than the expected cost of defense, even
for weak patents.73 Aggregation should allow defendants to reduce
total costs by exploiting economies of scale, making more defenses
positive-value, and moving closer to optimal deterrence.74
Third, even for individually positive-value claims, plaintiff aggregation allows plaintiffs to share and spread costs over the whole set
of their claims, which the defendant naturally will be able to do simply by being involved in many similar cases (e.g., by using the same
expert witnesses, reusing briefing, undertaking a single document
collection, etc.).75 Aggregation thus evens the ability and incentives of
the parties to invest in litigation, making it more likely that the resolution will reflect the merits rather than a “war of attrition” of costly
discovery and motion practice.76 Cost-spreading also applies in reverse—that is, absent aggregation of defendants, plaintiffs have a
greater ability to exploit economies of scale and therefore have greater settlement leverage or ability to engage in a war of attrition.77 For
example, patent assertion entities have been accused of exploiting
their lower cost of litigation and ability to spread costs among many
defendants by making broad and costly discovery requests.78
Thus, in theory, defendant aggregation offers the same benefits as
plaintiff aggregation, and these benefits generally accrue to the
70. Epstein, supra note 69, at 6-7; see also, e.g., Rosenberg, supra note 4, at 563-64
(discussing how aggregation can reduce the costs of individual litigation and promote
optimal deterrence).
71. Hamdani & Klement, supra note 4, at 696-99.
72. Id. at 697.
73. See Ranganath Sudarshan, Nuisance-Value Patent Suits: An Economic Model and
Proposal, 25 SANTA CLARA COMPUTER & HIGH TECH. L.J. 159, 160 (2008); Taylor, supra
note 7, at 674-75; Bryant, supra note 23, at 693.
74. Hamdani & Klement, supra note 4, at 698-99.
75. See Rosenberg, supra note 4, at 570-71.
76. See id.
77. Hamdani & Klement, supra note 4, at 696-99.
78. See Sudarshan, supra note 73, at 166-67; J. Jason Williams et al., Strategies for
Combating Patent Trolls, 17 J. INTELL. PROP. L. 367, 368, 375 (2010).
2014]
AGGREGATING DEFENDANTS
1023
courts and the party being aggregated (the defendant) at the expense
of the non-aggregated party (the plaintiff).79 It also may serve an additional socially beneficial function not shared by plaintiff aggregation, forcing defendants to produce information about relative contributions to harm.80
D. Resistance to Aggregating Similarly Situated Defendants
Recent efforts by plaintiffs to aggregate unrelated but similarly
situated defendants have sparked fierce resistance from defendants,
courts, policymakers, and commentators in a variety of substantive
areas. Although most of the controversy has centered on joinder of
defendants under Rule 20, there also has been resistance to aggregating defendants via defendant class actions,81 consolidation under
Rule 42(a),82 and the multi-district litigation statute.83
1. Patent Litigation
In recent years, patent litigation has become concentrated in districts seen as particularly friendly to patent owners, also known as
“patentees,” especially the Eastern District of Texas.84 Filing in a
plaintiff-friendly district has been popular among what are variably
called patent assertion entities, non-practicing entities, or pejoratively, patent trolls,85 which derive revenue from threatening or filing
patent litigation rather than commercializing the invention.86 Patent
assertion entities are “perhaps the most controversial and least popular group of patent [holders]” accused of delaying litigation to maximize damages and leverage and using weak patents that are likely invalid but costly to invalidate.87
79. See Hamdani & Klement, supra note 4, at 698-99 (describing benefit to defendants); id. at 726-27 (noting that “[f]or plaintiffs, the downsides of the class defense are obvious” and “[t]he upside of the class defense for plaintiffs is unclear”).
80. See Shen, supra note 22, at 98.
81. See Donald E. Burton, The Metes and Bounds of the Defendant Class Action in
Patent Cases, 5 J. MARSHALL REV. INTELL. PROP. L. 292 (2006) (describing disuse of defendant class actions and objections raised to it).
82. See 35 U.S.C. § 299 (Supp. V 2011) (prohibiting consolidation of patent defendants
for trial).
83. See Greg Ryan, JPML Head Judge Explains Why Panel Is More Picky, LAW360
(May 24, 2013, 8:21 PM), http://www.law360.com/articles/442674/jpml-head-judge-explainswhy-panel-is-more-picky (attributing increase in denial of multi-district litigation motions
to increase in motions in patent cases, i.e., where defendants being aggregated).
84. Mark A. Lemley, Where to File Your Patent Case, 38 AIPLA Q.J. 401, 404 (2010).
85. PWC, 2012 PATENT LITIGATION STUDY 24 (showing that 37.4% of patent decisions
in the Eastern District of Texas involved patent assertion entities compared to 20.6% in
all districts).
86. Colleen V. Chien, Of Trolls, Davids, Goliaths, and Kings: Narratives and Evidence
in the Litigation of High-Tech Patents, 87 N.C. L. REV. 1571, 1578 (2009).
87. Id. at 1577-80.
1024
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
Patent assertion entities often filed a single lawsuit against multiple unrelated defendants accused of infringing the same patent(s)
with independent but similar products. For example, a patent assertion entity whose patent purportedly covered MMS messaging in mobile phones might sue LG, Sanyo, Samsung, Research in Motion, and
Apple—each of whom developed and sold MMS-capable phones but
did so independently of, and in competition with, each other. Though
not used exclusively by patent assertion entities, multi-defendant
patent suits came to be associated with their questionable (or to be
fair, questioned) litigation tactics and the plaintiff-friendly Eastern
District of Texas. Patent assertion entities filed nineteen percent of
patent cases but sued twenty-eight percent of defendants,88 and nearly twice as many defendants were sued per case in the Eastern District of Texas than the national average.89 Patent assertion entities
were assumed to file multi-defendant suits to decrease their own
costs, increase the costs of defendants, and decrease the chances of
transfer by naming at least one defendant with a connection to the
plaintiff-friendly district.90
The overwhelming majority of district courts, ultimately endorsed
by the Federal Circuit, held that claims against “separate companies
that independently design, manufacture and sell different products
in competition with each other” did not arise from the same transaction or occurrence for purposes of joinder, even where the products
were accused of infringing the same patent and operated similarly.91
Rather, “an actual link” between the products was required, such as
a relationship between the defendants, the use of identically sourced
components, or overlap in the products’ development or manufacture.92 The minority view, adopted almost exclusively by the Eastern
District of Texas,93 held that patent infringement claims arose from
the same transaction or occurrence “if there is some nucleus of operative facts or law,” such as allegations that the defendants infringed
the same patent or had products that were not “dramatically differ88. Id. at 1603-04 (2000–08 time period).
89. See James C. Pistorino, 2012 Trends in Patent Case Filings and Venue: Eastern
District of Texas Most Popular for Plaintiffs (Again) but 11 Percent Fewer Defendants
Named Nationwide, PAT., TRADEMARK & COPYRIGHT L. DAILY, Feb. 11, 2013, at 3. This
number is derived from the data reported for 2011, before Congress altered the joinder
rules for patent cases.
90. Taylor, supra note 7, at 671-78.
91. Pergo, Inc. v. Alloc, Inc., 262 F. Supp. 2d 122, 128 (S.D.N.Y. 2003); Androphy v.
Smith & Nephew, Inc., 31 F. Supp. 2d 620, 623 (N.D. Ill. 1998); see also In re EMC Corp.,
677 F.3d 1351, 1357 & n.2, 1359 & n.3 (Fed. Cir. 2012) (adopting the overwhelming view
outside the Eastern District of Texas that joinder is inappropriate for independent but
similarly situated defendants).
92. In re EMC Corp., 677 F.3d at 1359-60.
93. Id. at 1357 & n.2.
2014]
AGGREGATING DEFENDANTS
1025
ent.”94 Citing judicial economy, these decisions reasoned that the
most important issues, including the determination of patent scope,
or “claim construction,” and invalidity, would be identical.95
Congress resolved this split as part of a more general overhaul of
the patent system in the 2011 America Invents Act (AIA).96 A new
statutory provision, 35 U.S.C. § 299, provided that joinder, or consolidation for trial, of accused infringers only was permitted if the
claims involved the making, using, importing, offering to sell, or selling “of the same accused product or process” and clarified that “allegations that [the defendants] each have infringed the patent or patents in suit” alone were insufficient for joinder.97 Though the legislative history is sparse, the provision appears to have been motivated
by concern about the Eastern District of Texas and perceived abusive
litigation practices of patent assertion entities.98 As a result of the
AIA’s anti-joinder provision, claims against unrelated defendants
with independent products only can be aggregated for pre-trial purposes and only through Rule 42(a) consolidation (if pending in
the same district) or through multi-district litigation (if pending in
different districts).99
2. Trademark and Copyright Litigation
Trademark holders have similarly sought to join several unrelated
defendants for similar but independent activities alleged to violate
the same rights. Defendants have objected, and courts have held that
these independent activities are not part of the same transaction or
occurrence for purposes of joinder.100
Efforts to aggregate similarly situated copyright defendants have
been even more controversial. The internet allows diffuse individuals
to share copyrighted movies, videos, or music through so-called “peerto-peer” (P2P) networks, the earliest and most famous of which was
94. MyMail, Ltd. v. Am. Online, Inc., 223 F.R.D. 455, 456-57 (E.D. Tex. 2004).
95. See Innovative Global Sys. LLC v. Tpk. Global Techs. LLC, No. 6:09-CV-157, 2009
U.S. Dist. LEXIS 105929, at *5-7 (E.D. Tex. Oct. 20, 2009); Adrain v. Genetec Inc., No.
2:08-CV-423, 2009 U.S. Dist. LEXIS 86855, at *7-10 (E.D. Tex. Sept. 22, 2009); Sprint
Commc’ns Co. v. Theglobe.com, Inc., 233 F.R.D. 615, 617 (D. Kan. 2006); MyMail, 223
F.R.D. at 456-58.
96. See generally Taylor, supra note 7 (discussing background and consequences of
AIA reforms to patent joinder).
97. 35 U.S.C. § 299 (Supp. V 2011).
98. See Taylor, supra note 7, at 700-06.
99. See id. at 719-22.
100. See Bravado Int’l Grp. Merch. Servs. v. Cha, No. CV 09-9066 PSG, 2010 U.S. Dist.
LEXIS 80361, at *11-14 (C.D. Cal. June 30, 2010); Golden Scorpio Corp. v. Steel Horse Bar
& Grill, 596 F. Supp. 2d 1282, 1285 (D. Ariz. 2009); Colt Def. LLC v. Heckler & Koch Def.,
Inc., No. 2:04CV258, 2004 U.S. Dist. LEXIS 28690, at *15-16 (E.D. Va. Oct. 22, 2004); SB
Designs v. Reebok Int’l, Ltd., 305 F. Supp. 2d 888, 892 (N.D. Ill. 2004).
1026
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
Napster.101 After early efforts to hold P2P networks liable ultimately
failed,102 the Recording Industry Association of America (RIAA) and
its major record label members began suing individual file-sharers,
primarily as a deterrent.103 The RIAA’s basic strategy was to search
P2P networks for particular copyrighted materials; collect the IP addresses of uploaders of infringing files; sue numerous “John Does” in
a single lawsuit, with each John Doe representing a different uploader’s IP address; seek court approval for early subpoenas to internet
service providers (ISPs) to determine the identities of the John Does;
and contact the John Does and offer to settle the case for around
$3000, which was less than the cost of defense.104 Overall, the RIAA
and its members sued 30,000 individuals during its 2003 to 2008 enforcement campaign, sometimes in individual lawsuits and sometimes in groups of dozens or hundreds of individuals, most of whom
were John Does.105
Of the many potential problems with the RIAA’s strategy,106 courts
and defendants focused primarily on the propriety of joinder.107 Plaintiffs alleged that joinder was proper because each defendant “committed violations of the same law (e.g., copyright law), by committing
the same acts (e.g., the downloading and distribution of copyrighted
sound recordings owned by Plaintiffs), and by using the same means
101. Bridy, supra note 23, at 698-710; Hamdani & Klement, supra note 4, at 699-700.
102. See Hamdani & Klement, supra note 4, at 700.
103. See id. at 700-01.
104. Karunaratne, supra note 14, at 286-87.
105. Bridy, supra note 23, at 721. Commentators frequently suggest that the RIAA
“usually nam[ed] dozens or hundreds of defendants per suit.” David Kravets, Copyright
Lawsuits Plummet in Aftermath of RIAA Campaign, WIRED (May 18, 2010, 1:24 PM),
http://www.wired.com/threatlevel/2010/05/riaa-bump/. But the RIAA sued the 30,000 individuals in approximately 13,000 lawsuits, an average of only 2.3 defendants per lawsuit.
Bridy, supra note 23, at 721 (noting increase in copyright lawsuits from 2003–2008 and
attributing it to RIAA campaign). The RIAA lawsuits were a mix of suits against a single
individual file-sharer and suits against dozens or hundreds of file-sharers, most of whom
were John Does. See Ray Beckerman, P.C., Index of Litigation Documents Referred to in
“Recording Industry vs. The People”, BECKERMANLEGAL, http://beckermanlegal.com/
Documents.htm (last updated May 17, 2013) (listing exemplary music industry cases). The
RIAA generally used aggregated actions when it did not know the identity of the defendants and individual actions when it did. See Arista Records, LLC v. Does 1–11, No. 1:07–
CV–2828, 2008 WL 4823160, at *3 & n.3, *6 (N.D. Ohio Nov. 3, 2008) (stating that the
RIAA will likely re-file individual actions once the defendants are identified). Individual
suits may have been used to deprive defendants of economies of scale, see Hamdani &
Klement, supra note 4, at 699-702, or it may be that most defendants chose to settle and
those that did not were not subject to personal jurisdiction and venue in the same district.
See Karunaratne, supra note 14, at 287, 298-302.
106. See generally Karunaratne, supra note 14 (noting concerns about abuse of John
Doe procedures, insufficient showings for expedited subpoenas, lack of personal jurisdiction, and lack of a necessary connection between IP address and file-sharer).
107. See id. at 287-88 (noting that courts in RIAA cases responded “with particular
force on the joinder issue” and that courts “were less willing to confront questions of
personal jurisdiction”).
2014]
AGGREGATING DEFENDANTS
1027
(e.g., a file-sharing network) that each Defendant accessed via the
same ISP.”108 The propriety of joinder often was contested by a defendant109 or the court sua sponte,110 and “the majority of district
courts who . . . addressed the issue of joinder . . . concluded that those
allegations were insufficient to satisfy the transactional requirement
of Fed.R.Civ.P. 20(a)(2) and that joinder was therefore improper.”111
Courts held that “merely alleging that the Doe Defendants all used
the same ISP and file-sharing network to conduct copyright infringement without asserting that they acted in concert was not
enough to satisfy the same series of transactions requirement under
the Federal Rules,” and that merely alleging that the defendants
caused “the same type of harm” rather than “the same harm” was insufficient for joinder.112 Some courts even proposed sanctions against
the plaintiffs for attempting joinder.113
With the end of the RIAA’s campaign, mass copyright enforcement
shifted to the movie industry and to the pornographic films industry
in particular. The litigation model essentially was the same as that of
the RIAA, though joinder often was now of hundreds or thousands of
John Does.114 As with the RIAA litigation, defendants115 or courts sua
sponte116 challenged the mass joinder of defendants, and many courts
held that “the fact that a large number of people use the same method to violate the law does not authorize them to be joined as defendants in a single lawsuit” because each defendant independently accessed the P2P network and downloaded the copyrighted file pieces
in “discrete and separate acts that took place at different times”
without concerted action.117
108. Does 1–11, 2008 WL 4823160, at *1 (quoting complaint); Arista Records, LLC v.
Does 1–27, No. 07-162-B-W, 2008 WL 222283, at *1 (D. Me. Jan. 25, 2008) (quoting complaint), report and recommendation adopted, 584 F. Supp. 2d 240 (D. Me. 2008).
109. Does 1–11, 2008 WL 4823160, at *1.
110. See Does 1–27, 2008 WL 222283, at *6 n.5.
111. Does 1–11, 2008 WL 4823160, at *6; see also Bridy, supra note 23, at 722, 723 &
nn.164-69 (explaining the reasoning of courts’ sua sponte finding of improper joinder in
multiple jurisdictions); Karunaratne, supra note 14, at 287-88 & nn.25-28 (noting that
courts find joinder to be improper in file-sharing cases when defendants only use the same
ISP and peer-to-peer networks). The limited courts not ordering de-aggregation generally
concluded only that severance was premature until the Does were identified. See Does 1–
11, 2008 WL 4823160, at *3-5.
112. Does 1–11, 2008 WL 4823160, at *6.
113. See Does 1–27, 2008 WL 222283, at *6 n.5.
114. See Bridy, supra note 23, at 721-22. See generally Karunaratne, supra note 14
(discussing the propriety of joinder in copyright infringement cases against John Does).
115. See Hard Drive Prods., Inc. v. Does 1–188, 809 F. Supp. 2d 1150, 1153 (N.D. Cal. 2011).
116. See Next Phase Distrib., Inc. v. Does 1–27, 284 F.R.D. 165, 166 (S.D.N.Y. 2012).
117. Digital Sins, Inc. v. Does 1–245, No. 11 Civ. 8170(CM), 2012 WL 1744838, at *2
(S.D.N.Y. May 15, 2012).
1028
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
Courts in the pornography cases are more divided on joinder than
in the RIAA cases, but the division is over whether the defendants
acted in concert, not whether similar claims against unrelated defendants can be aggregated.118 In earlier P2P networks, a single user
uploaded a copyrighted file, which was then downloaded in its entirety separately by other users. BitTorrent breaks the copyrighted material into pieces, which a user then collects from various other users
and must share with others once in her possession.119 Some courts
have upheld joinder (or postponed resolution), concluding that concerted action existed to the extent the defendants were part of the
same group of users (called a “swarm”) sharing the same pieces at
the same time or in the same time period.120
3. Telecommunications Cases
Controversy over aggregation of similarly situated defendants has
frequently arisen in telecommunications cases,121 most notably in DirecTV’s campaign against piracy. DirecTV raided businesses selling
devices that could unscramble its signals and sent demand letters to
more than 170,000 purchasers, offering to settle for $3500 per purchaser (presumably less than the cost of defense).122 DirecTV ultimately sued over 24,000 individuals under various wiretap, communications, and copyright laws, often by suing several unrelated and
independent violators in a single suit.123
Once again, many potential problems existed with these cases,124
but the joinder issue, raised either by defendants’ motions or sua
sponte by the court,125 was the focal point of resistance.126 “Most
courts presented with a suit of this type have concluded that the
claims against the various defendants are not transactionally relat118. See Next Phase, 284 F.R.D. at 168-69.
119. Karunaratne, supra note 14, at 288-90.
120. Digital Sin, Inc. v. Does 1–27, No. 12 Civ. 3873(JMF), 2012 WL 2036035, at *2
(S.D.N.Y. June 6, 2012).
121. See, e.g., Don King Prods., Inc. v. Colon-Rosario, 561 F. Supp. 2d 189 (D.P.R.
2008); Movie Sys., Inc. v. Abel, 99 F.R.D. 129, 129-30 (D. Minn. 1983) (18 suits against
1,795 total individuals).
122. In re DIRECTV, Inc., No. C-02-5912-JW, 2004 WL 2645971, at *2 (N.D. Cal. July
26, 2004); Hamdani & Klement, supra note 4, at 703-06.
123. See David W. Opderbeck, Peer-to-Peer Networks, Technological Evolution, and
Intellectual Property Reverse Private Attorney General Litigation, 20 BERKELEY TECH. L.J.
1685, 1725-26 (2005). For example, DirecTV sued 775 defendants in 180 lawsuits in the
Northern District of California. See DIRECTV, 2004 WL 2645971, at *2.
124. See Hamdani & Klement, supra note 4, at 703-06 (noting concerns about viability
of statutory claims, investigative techniques, and lack of necessary correlation between
device and piracy).
125. See, e.g., DIRECTV, 2004 WL 2645971, at *1 (sua sponte); DirecTV, Inc. v. Beecher, 296 F. Supp. 2d 937, 939 (S.D. Ind. 2003) (motion).
126. See DIRECTV, 2004 WL 2645971, at *5 (summarizing cases).
2014]
AGGREGATING DEFENDANTS
1029
ed” for joinder.127 They have held that “[i]ndividual purchasers, who
have no business connection with one another and who make their
purchases independently of one another are not engaged in the same
transaction,” even though they received their devices from the same
shipping facility, “perform[ed] the same act in the same [geographic]
area,” and engaged in “similar statutory violations” that “injured
DIRECTV in the same manner.”128
4. Other Examples
Technology cases are not the only examples of efforts by plaintiffs
to aggregate similarly situated defendants and corresponding resistance by defendants, courts, and policymakers. This has also occurred in products liability,129 environmental,130 consumer protection,131 indemnification,132 and other types of cases.133
E. The Puzzle of Defendant Aggregation
The narrative of plaintiff aggregation has become well defined in
its nearly half-century at the forefront of American litigation. Plaintiffs (and their attorneys) seek aggregation because they benefit from
its economies of scale, while defendants generally oppose plaintiff
aggregation because of the in terrorem effect of the aggregated liability.134 Courts and policymakers walk a middle ground, allowing ag127. MOORE ET AL., supra note 49, § 20.05[3], at 20-37; see also DIRECTV, 2004 WL
2645971, at *5 (summarizing cases and rationales); McFarland, supra note 55, at 268 (noting that out of the several courts that have considered joinder in television-pirating cases,
one has allowed joinder while a dozen others have denied it).
128. DIRECTV, 2004 WL 2645971, at *4.
129. Erichson, supra note 20, at 403 & n.80. Many multi-defendant products liability
cases, and multi-defendant cases in other contexts, involve aggregated plaintiffs each
injured in similar ways but only by one defendant. Defendants have contested the propriety of defendant aggregation in these cases, see, e.g., Turpeau v. Fidelity Fin. Servs., 936 F.
Supp. 975, 980 (N.D. Ga. 1996), but more often, defendants contest whether plaintiffs meet
the requirements for a class action, see, e.g., Arch v. Am. Tobacco Co., 175 F.R.D. 469,
489 (E.D. Pa. 1997), or joinder, see, e.g., Abdullah v. Acands, Inc., 30 F.3d 264, 268 & n.5
(1st Cir. 1994).
130. See, e.g., Coal. for a Sustainable Delta v. U.S. Fish & Wildlife Serv., No. 1:09–CV–
480 OWW GSA, 2009 WL 3857417 (E.D. Cal. Nov. 17, 2009).
131. See, e.g., Turpeau, 936 F. Supp. at 979-80.
132. See, e.g., United States v. Katz, 494 F. Supp. 2d 645, 646 (S.D. Ohio 2006).
133. See, e.g., Nassau Cnty. Ass’n of Ins. Agents v. Aetna Life & Cas. Co., 497 F.2d
1151, 1152 (2d Cir. 1974) (employment and antitrust case).
134. See Richard A. Nagareda, Aggregation and Its Discontents: Class Settlement Pressure, Class-Wide Arbitration, and CAFA, 106 COLUM. L. REV. 1872, 1873-74 (2006) [hereinafter Nagareda, Discontents]. This description is admittedly a generalization. See Elizabeth
Chamblee Burch, Litigating Groups, 61 ALA. L. REV. 1, 11-12 (2009) (discussing a defendant’s preference for aggregated settlements of mass joinder actions); Richard A. Nagareda,
Autonomy, Peace, and Put Options in the Mass Tort Class Action, 115 HARV. L. REV. 747,
751 (2002) (discussing defendants’ support for settlement classes) [hereinafter Nagareda,
Autonomy]; David Rosenberg, Adding a Second Opt-Out to Rule 23(b)(3) Class Actions: Cost
1030
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
gregation of similarly situated plaintiffs but imposing limits in some
cases and some contexts. By contrast, the narrative of defendant aggregation is utterly confused. Theoretically, defendants obtain similar economies of scale and the judicial system obtains similar efficiencies from aggregation. Defendants face no in terrorem effect, as
an individual defendant’s liability does not change with aggregation.
To the contrary, defendant aggregation increases resources on the
defense side and raises the stakes for the plaintiff, since a single loss
extinguishes all of its claims.
Yet, plaintiffs have generally sought to aggregate defendants
while defendants, courts, and policymakers have resisted, assuming
that defendant aggregation benefits plaintiffs at the expense of defendants.135 Parts III and IV seek to solve this puzzle of defendant
aggregation, with the former focusing on defendant resistance and
the latter systemic resistance.
III. DEFENDANTS’ OPPOSITION TO DEFENDANT AGGREGATION
This Part explores why defendants oppose being aggregated despite the economies of scale benefits they should obtain. In some contexts, the explanation is easy because aggregation is clearly against
defendants’ interests. In other contexts, the effects of aggregation on
defendants are unclear. And, in at least one category of cases, defendants lose nothing from being aggregated and their resistance
appears irrational.
A. Creating Additional Litigation for Defendants
Defendants’ opposition to being aggregated is easy to explain in
cases, like the BitTorrent copyright litigation, where the stakes are
small, each defendant’s liability is less than the cost of individual litigation, and the claims are only economically viable because of the
cost-savings the plaintiff realizes from defendant aggregation. Defendants are better off without being aggregated because these
claims would never be brought individually and defendants would
have no need for the economies of scale aggregation offers them.
This category of cases may not be particularly large. Even without
aggregation, a plaintiff can spread many of its costs over the full
portfolio of individual cases through, for example, form complaints;
reuse of briefs, expert reports, and discovery requests; and a single
fact investigation and document collection.136 In only a limited set of
Without Benefit, 2003 U. CHI. LEGAL F. 19, 22-23 (2003) (discussing hold-out and opt-out
plaintiffs in mass joinders and class actions).
135. See, e.g., Bridy, supra note 23, at 722-24; Taylor, supra note 7, at 671-78.
136. See Hamdani & Klement, supra note 4, at 696 n.36.
2014]
AGGREGATING DEFENDANTS
1031
cases will the additional savings the plaintiff realizes from defendant
aggregation—filing fees, administrative costs related to preparing
and submitting court documents, and attendance at depositions,
hearings, and trial137—be the difference in the viability of the claim.
In the BitTorrent cases, the plaintiff had unusually high up-front
costs because it only knew the infringers’ IP addresses and needed
expedited subpoenas to obtain the identities of those whom it could
target for settlement demands.138 The cost of separate filing fees,
moving individually for expedited subpoenas, attending separate
subpoena hearings, and executing individual subpoenas on the ISPs
would likely dwarf the few thousand dollars at stake.139 A small subset of extreme patent troll cases also may fit in this category, where
the patentee sues large numbers of end-users of common technology
(e.g., hotels using Wi-Fi related patents or small businesses using
scanners) and demands only a few thousand dollars from each.140 The
patent anti-joinder provision seems to have deterred these smallstakes patent cases.141
On the other hand, higher stakes cases where defendant aggregation is not the difference in the viability of the plaintiffs’ claims will
be brought even if defendants cannot be aggregated. This is true of
most patent cases; there was only a small decrease in the total
number of defendants sued and no significant effect on patent assertion entities’ share of defendants after enactment of the patent
anti-joinder provision.142 Similarly, the financial stakes in the individual RIAA and DirecTV cases were small, but the corporate plaintiffs derived significantly greater value from the larger deterrent
effect of the suits.143 While it is possible they would have sued fewer
defendants without aggregation, depending on how many suits were
necessary to provide sufficient deterrence, both the RIAA and
DirecTV plaintiffs often pursued individual, negative-value lawsuits
when necessary.144
137. See Bridy, supra note 23, at 722; Taylor, supra note 7, at 672.
138. See Karunaratne, supra note 14, at 291.
139. See Karunaratne, supra note 14, at 303-04.
140. See, e.g., Joe Mullin, Wi-Fi Patent Troll Hit with Racketeering Suit Emerges Unscathed, ARS TECHNICA (Feb. 13, 2013, 10:05 AM), http://arstechnica.com/tech-policy/
2013/02/wi-fi-patent-troll-hit-with-novel-anti-racketeering-charges-emerges-unscathed/.
141. See Colleen Chien, Patent Trolls by the Numbers 3 (Mar. 13, 2013) (unpublished
manuscript), available at https://ssrn.com/abstract=2233041.
142. Id.; Pistorino, supra note 89, at 4.
143. See supra Part III.D.2-3.
144. See Arista Records, LLC v. Does 1–11, No. 1:07–CV–2828, 2008 WL 4823160, at *5
(N.D. Ohio Nov. 3, 2008) (“Plaintiffs indicated to the Magistrate Judge that they intend to
sever the Doe Defendants’ cases once they have been identified.”).
1032
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
B. Cost-Differentials
Even for claims with high enough stakes to be brought individually, defendant aggregation could create, or accentuate, a pro-plaintiff
cost-differential, incentivizing plaintiffs to bring weak claims and defendants to settle meritorious defenses, at least at the margins.145
Despite the suggestions that this has occurred in multi-defendant
patent “troll” cases,146 it is unclear why the net costs of defendant aggregation would favor plaintiffs.147
Although empirical evidence is lacking, it is doubtful that aggregating defendants would eliminate more costs from individual litigation for plaintiffs than defendants. Plaintiffs can spread many of
their costs over individual cases, and any efforts by defendants in
individual suits to spread costs through a joint defense group or other
informal aggregation are likely to be less effective, efficient, and
substantial.148 With the exception of the $400 filing fee, an insubstantial amount for any individually viable claim, defendants can match
or exceed whatever additional cost-savings plaintiffs realize from
defendant aggregation by, for example, dividing responsibility for
depositions, hearings, and trial; splitting up document review and
fact investigation; and preparing joint briefing or expert reports on
common issues.149
On the other hand, defendant aggregation may impose new costs
on defendants not present in individual litigation and not matched by
the individual plaintiff. Aggregated parties often are required, either
by court order or strategic considerations, to agree on a common
strategy, achieve consensus on the myriad of issues that arise in litigation, divide tasks, file a single brief on common issues, or even present a single argument at a hearing or single case at trial,150 all of
which could impose substantial coordination costs in attorney time
and client money.151 Aggregated plaintiffs minimize these coordination costs because plaintiffs’ lawyers’ contingency fee arrangements
provide an incentive to reduce costs by dividing, rather than duplicat145. See John C. Coffee, Jr., The Regulation of Entrepreneurial Litigation: Balancing
Fairness and Efficiency in the Large Class Action, 54 U. CHI. L. REV. 877, 891 (1987).
146. See Taylor, supra note 7, at 673-75.
147. See Coffee, supra note 145, at 891-94 (describing difficulty in identifying existence
and extent of cost-differential in litigation).
148. See Taylor, supra note 7, at 673-75.
149. See Erichson, supra note 20, at 403-05.
150. Taylor, supra note 7, at 673 & n.103, 674.
151. See Mark Baghdassarian & Aaron Frankel, Litigation: Managing Joint Defense
Groups in Asymmetrical Lawsuits, INSIDECOUNSEL (Aug. 23, 2012), http://www.inside counsel.com/2012/08/23/litigation-managing-joint-defense-groups-in-asymme?t=litigation&page=2;
see also Judith Resnik et al., Individuals Within the Aggregate: Relationships, Representation, and Fees, 71 N.Y.U. L. REV. 296, 304 (1996) (noting increase in plaintiff aggregation of
“multiple sets of lawyers within a single aggregated litigation”).
2014]
AGGREGATING DEFENDANTS
1033
ing, work and compromising, rather than disputing, strategy or responsibilities.152 But for aggregated defendants, defense attorneys’
hourly fee arrangements are likely to increase coordination costs, as
duplicating work (e.g., attending every deposition or carefully editing
every joint submission) and disputing strategy and responsibilities
maximizes fees.153 Yet, if the resulting high coordination costs eliminated the cost-savings defendants otherwise realize from being aggregated, a rational defendant would respond by insisting on fee arrangements that minimized coordination costs, not by rejecting aggregation. Before Congress passed the patent anti-joinder provision,
companies sued in multi-defendant patent cases in the Eastern District of Texas had begun to share a common attorney with other defendants, insist on a fixed fee, or refuse to pay for duplicative work.154
Aggregated defendants also face potential free-rider problems,
that is, some defendants do the minimal work necessary for their individual cases and rely on other defendants to develop common defenses that apply to all defendants regardless of who pays for them.155
However, common defenses are public goods and a defendant that
prevails on a common defense in individual litigation will bear the
full cost but realize only some of the benefits, which are shared with
any other similarly situated defendant.156 Thus, aggregation poses no
greater free-rider costs than individual litigation.157
C. Substantive Effects
Defendant aggregation does not just affect the costs of litigation; it
also affects its substance. Perhaps defendants oppose being aggregated
despite its potential economies of scale, because of substantive concerns.
1. Asymmetric Preclusion
The doctrine of non-mutual issue preclusion precludes a party
from re-litigating an issue it lost in a prior suit but does not bind a
non-party to the prior suit.158 As a result, if a plaintiff loses on a key
issue in an individual case then it will be bound by that loss in all
other cases against similarly situated defendants, but if it prevails, it
152. See Coffee, supra note 145, at 889-94; see also Resnik et al., supra note 151, at
309-14 (describing workings of aggregated plaintiffs’ lawyers).
153. See Coffee, supra note 145, at 892.
154. See Baghdassarian & Frankel, supra note 151 (suggesting shared counsel).
155. See id.
156. See Roger Allan Ford, Patent Invalidity Versus Noninfringement, 99 CORNELL L.
REV. 71, 109-12 (2013).
157. See Defendant Class Actions, supra note 20, at 648. The public good nature of
common defenses is normally an argument for, not against, aggregation of similarly situated defendants. See Parchomovsky & Stein, supra note 23, at 1534-35.
158. See Blonder-Tongue Labs., Inc. v. Univ. of Ill. Found., 402 U.S. 313, 329 (1971).
1034
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
still has to re-litigate the issue with each subsequent defendant, with
the risk that a subsequent loss will be binding in all remaining cases.
Aggregation deprives defendants of the multiple bites at the apple
offered in serial litigation by non-mutual preclusion.
However, this is only a loss to defendants if their defenses are individually positive-value, that is, the potential liability exceeds the
cost of individual litigation. If the cost of individual litigation exceeded the potential liability, defendants would settle, and there would be
no final judgment to which asymmetric preclusion could apply.159 For
example, some patent troll cases, the RIAA file-sharing litigation,
and the DirecTV suits likely involved negative-value defenses where
asymmetric preclusion would be largely irrelevant.160
Even in higher stakes cases with individually positive-value defenses, asymmetric preclusion only is significant in limited circumstances. If the primary issues are legal, subsequent defendants often
will be effectively bound by resolution of an issue in earlier litigation
through stare decisis, persuasive power, or simply disinclination to
revisit something already decided.161 A rational defendant probably
would prefer aggregation and the opportunity to influence the initial
decision than the gamble that it can convince a subsequent court to
revisit the issue after it was botched by the first defendant,162 especially since plaintiffs in serial litigation target weak, underfunded
defendants first to cheaply and quickly obtain favorable precedent.163
Asymmetric preclusion also is of little help in cases where the primary issues relate to fact questions unique to each defendant—say the
purpose to which the defendant put the DirecTV unscrambling device—since favorable findings for earlier defendants would be inapplicable to subsequent defendants.
Thus, asymmetric preclusion only is advantageous to defendants
in cases where the primary issues are factual but common, such as
questions related to the plaintiff’s conduct. Patent litigation is one
such example: two key patent defenses—invalidity and misconduct in
the Patent Office (i.e., “inequitable conduct”)—have significant factual components common to each defendant.164 It is possible that patent
159. See Hamdani & Klement, supra note 4, at 737.
160. See id. at 701-04 (describing negative-value nature of defenses in RIAA and DirecTV cases); Jeremy P. Oczek, Rethinking Defense in “Patent Troll” Cases, CORP. COUNS.,
Mar. 27, 2013, available at http://www.bsk.com/media-center/2730-rethinking-defensepatent-troll-cases (noting that the average cost to defend patent cases is $1 million when
less than $1 million is at stake and $3 million when $1 to 25 million is at stake).
161. See Markman v. Westview Instruments, Inc., 517 U.S. 370, 391 (1996) (noting that
stare decisis would apply to patent claim interpretation even without issue preclusion).
162. See Jeffrey T. Haley, Strategies and Antitrust Limitations for Multiple Potential
Patent Infringers, 21 AIPLA Q.J. 327, 334-35 (1993).
163. See Parchomovsky & Stein, supra note 23, at 1488.
164. See id. at 1501-03.
2014]
AGGREGATING DEFENDANTS
1035
defendants realize more benefits from asymmetric preclusion than
they do from aggregation’s economies of scale, though, even here,
commentators have suggested that the rejection of an invalidity defense in an earlier case makes a later finding of invalidity less likely.165
2. Jury and Judicial Confusion
Pursuing separate claims simultaneously against multiple defendants may allow the plaintiff to exploit judicial or jury confusion
over which evidence and arguments apply to which defendants.166
Confusion is not inherently detrimental to aggregated defendants.
For example, a similar risk of confusion of evidence and arguments is
seen as benefitting aggregated plaintiffs by allowing them to focus on
the strongest claims or combine strong parts of various claims to create a collective claim stronger than any individual one.167
However, plaintiffs may strategically use their control over aggregation to combine weaker claims with stronger claims, hoping the
evidence and arguments for the latter will bolster the former. For
example, a DirecTV defendant that used the unscrambling device for
some legitimate purpose will have a difficult time highlighting this
when aggregated with lots of defendants who used the device to steal
DirecTV’s signals. Similarly, aggregating claims that a few defendants willfully and knowingly infringed a patent may bolster claims of
ordinary (strict liability) infringement against other defendants.
When applicable, the risk of judicial or jury confusion is a strong
reason for defendants to resist being aggregated. On the other hand,
potential confusion from aggregation does not make defendants
whose individual defenses are negative-value worse off, since they
would default or settle individual litigation without the opportunity
to benefit from an unconfused decision-maker. Even for positivevalue defenses, confusion only poses a problem if there is significant
heterogeneity among the defendants.168 If the primary defenses are
common (e.g., patent invalidity when the claims are broad and clearly cover the accused products) or very similar (e.g., patent noninfringement when the patentee alleges the patent covers all prod-
165. Glynn S. Lunney, Jr., FTC v. Actavis: The Patent-Antitrust Intersection Revisited
44 (Tulane Sch. of Law Pub. Law & Legal Theory Working Paper Series, Paper No. 13-19,
2013), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2348075.
166. See, e.g., Bridy, supra note 23, at 722-23.
167. See Erbsen, supra note 28, at 1009-14.
168. See id. at 1014; see also, e.g., Philips Elecs. N. Am. Corp. v. Contec Corp., 220
F.R.D. 415, 418 (D. Del. 2004) (citing potential prejudice in ordering de-aggregation because co-defendant was not presenting defense to infringement at trial).
1036
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
ucts with certain functionality regardless of how implemented),169
there is little risk of prejudice to defendants from jury or judicial confusion. For example, potential confusion does not explain resistance
to aggregation in many patent troll cases, which often involve broadly asserted claims and, consequently, defenses of either invalidity or
non-infringement that apply equally to all defendants.
3. Autonomy
Autonomy includes “the power of individuals to make fundamental choices concerning their legal rights of action – for example, the
power to choose when and how to sue, whether to settle, and if so,
under what terms.”170 Plaintiff aggregation increasingly faces
autonomy concerns,171 and these concerns may also underlie resistance to defendant aggregation.172 However, aggregation can only
raise autonomy concerns for positive-value defenses; for negativevalue defenses, aggregation promotes defendants’ autonomy by allowing them to defend rather than default and by deterring plaintiffs
from bringing strike suits.173 Individual autonomy also is of questionable relevance to corporate defendants, who lack the strong autonomy interests of individuals.174
Moreover, in practice, defendant aggregation almost never is attempted through representative procedures, like the class action,
that pose the greatest threat to individual autonomy.175 Rather, defendant aggregation has generally occurred via procedures (e.g.,
permissive joinder, consolidation, or multi-district litigation) that
leave the substantive rights of the defendants separate and allow
each defendant to retain control over its own case, obtain a separate
judgment as to its liability, choose its own attorney, settle whenever
and on whatever terms it chooses, and present its individual de-
169. See Mark A. Lemley, Software Patents and the Return of Functional Claiming,
2013 WIS. L. REV. 905, 907-08 (describing common problem of broad functional claims that
“purport to cover any possible way of achieving a goal”).
170. Nagareda, Autonomy, supra note 134, at 750; see also Epstein, supra note 69, at 5
(stating that autonomy over one’s lawsuit is a critical element of fairness).
171. Sergio J. Campos, Mass Torts and Due Process, 65 VAND. L. REV. 1059, 1060-61 (2012).
172. See, e.g., Wiav Networks, LLC v. 3COM Corp., No. C 10-03448 WHA, 2010 U.S.
Dist. LEXIS 110957, at *16 (N.D. Cal. Oct. 1, 2010) (“[T]he accused defendants — who will
surely have competing interests and strategies — are also entitled to present individualized assaults on questions of non-infringement, invalidity, and claim construction.”).
173. Cf. Epstein, supra note 69, at 6 (concluding that plaintiff aggregation raises no
autonomy concerns when the plaintiffs’ claims are too small to pursue individually because
individual control is too expensive for anyone to rationally choose it).
174. See Nathan Oman, Corporations and Autonomy Theories of Contract: A Critique of
the New Lex Mercatoria, 83 DENV. U. L. REV. 101, 114 (2005).
175. REDISH, supra note 48, at 230.
2014]
AGGREGATING DEFENDANTS
1037
fense.176 Thus, aggregated defendants are not “forced to adjudicate
their [defenses] passively” but rather are “able to represent [their]
own interests fully as a litigant before the court.”177
Thus, the only autonomy interest inherently threatened by defendant aggregation is the ability to present one’s defense without
the judge or jury hearing any other defense or to pursue one’s preferred strategy without any compromise or potential dilution from
the judge or jury being presented with competing strategies from codefendants. A defendant would have to value autonomy unusually
highly to conclude that this limited threat outweighed the benefits
from aggregation’s economies of scale. Admittedly, judicial practices
aimed at maximizing efficiency or minimizing workloads do accentuate autonomy concerns, as judges sometimes mandate that all aggregated defendants file a single brief, speak with a single voice, or
share the same amount of trial time as allotted to a single plaintiff.178
If these practices were defendants’ real concerns, the proper recourse
would be to challenge judicial case management procedures or judicial discretion, not to seek full-scale de-aggregation.
4. Tool to Achieve Other Substantive Objectives
Another possibility is that defendants oppose being aggregated not
because they dislike defendant aggregation as such but because seeking de-aggregation is doctrinally possible and serves other strategic
objectives, such as creating delay and cost for the plaintiff from motion practice and re-filing179 or eliminating a co-defendant with ties to
a plaintiff-friendly jurisdiction that serves as an anchor preventing
transfer.180 Or perhaps defendants take the doctrinal opportunity for
de-aggregation simply as a reflex from always opposing plaintiff aggregation or because they fear that greater use of aggregative procedures will have spillover effects that will promote plaintiff aggregation. Of course, the benefits from these other strategic objectives
176. See, e.g., FED. R. CIV. P. 20(a)(3); Malcolm v. Nat’l Gypsum Co., 995 F.2d 346, 352
(2d Cir. 1993); Dupont v. S. Pac. Co., 366 F.2d 193, 195-96 (5th Cir. 1966); MacAlister v.
Guterma, 263 F.2d 65, 68 (2d Cir. 1958); MANUAL FOR COMPLEX LITIGATION
(FOURTH) §§ 10.222, 12.21–.22, 13.21 (2004); MOORE ET AL., supra note 49, § 20.02[4].
177. REDISH, supra note 48, at 230.
178. See, e.g., Trial Scheduling Order, Fractus S.A. v. Samsung Elecs. Corp., No. 6:09–
CV–203 (E.D. Tex. May 6, 2011) (allotting thirty minutes per side for jury selection, forty
minutes per side for opening arguments, fifteen hours per side for direct/cross examination,
and one hour per side for closing arguments in patent case involving single plaintiff and
multiple unrelated defendants).
179. See Rosenberg, supra note 4, at 571.
180. See Taylor, supra note 7, at 676-79. De-aggregation has done little to help patent
defendants get out of the Eastern District of Texas, as courts have relied on the pending
litigation in the same district against the previously aggregated defendant as a basis for
denying transfer. See In re EMC Corp., 501 F. App'x 973, 976 (Fed. Cir. 2013).
1038
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
would have to be greater than the benefits defendants realize from
aggregation’s economies of scale, which is unlikely to be true at least
when the defenses are individually negative-value.
De-aggregation is doctrinally viable for defendants. Under Rule
20, the primary tool for defendant aggregation, joinder is permitted
when “any right to relief is asserted against [the defendants] jointly,
severally, or in the alternative with respect to or arising out of the
same transaction, occurrence, or series of transactions and occurrences.”181 Although courts only require a “logical relationship” for events
to be part of the same transaction or occurrence,182 they have narrowly defined from what events the “right to relief” arises, requiring that
there be a “logical relationship” in the defendants’ activities that allegedly violate the legal duty.183 Even though the right to relief arguably arises out of each of the four elements of a prototypical legal
claim—(1) a violation of (2) a legal duty that (3) causes (4) harm to
the plaintiff—courts generally have held that a logical relationship in
the events giving rise to the legal duty (e.g., the issuance of the patent)184 or the harm caused to the plaintiff (e.g., the stealing of the
same broadcast signal without paying for the rights)185 is insufficient
for joinder.186 This interpretation of Rule 20 favors joinder of multiple
unrelated plaintiffs suing a single defendant, where it is more likely
that the defendant’s alleged violation of the legal duty is the same or
factually related for all plaintiffs (e.g., the same allegedly illegal
practice or policy). By contrast, for a single plaintiff suing multiple
unrelated defendants, the defendants’ alleged violations of the legal
181. FED. R. CIV. P. 20(a)(2)(A) (emphasis added).
182. Mosley v. Gen. Motors Corp., 497 F.2d 1330, 1333 (8th Cir. 1974); WRIGHT,
MILLER & KANE, supra note 61, § 1653, at 409.
183. See, e.g., In re EMC Corp., 677 F.3d 1351, 1358 (Fed. Cir. 2012) (requiring that
“the defendants’ allegedly infringing acts” share operative facts to satisfy logical relationship test); Nassau Cnty. Ass’n of Ins. Agents, v. Aetna Life & Cas. Co., 497 F.2d 1151, 1154
(2d Cir. 1974) (finding no logical relationship because defendants’ allegedly wrongful actions were “separate and unrelated”).
184. See, e.g., Androphy v. Smith & Nephew, Inc., 31 F. Supp. 2d 620, 623 (N.D. Ill.
1998) (allegations of infringement of same patents “does not mean the claims against the
two companies arise from a common transaction or occurrence”); see also Nassau Cnty., 497
F.2d at 1154 (“violations of the same statutory duty” are not enough to permit joinder
(quoting Kenvin v. Newburger, Loeb & Co., 37 F.R.D. 473, 475 (S.D.N.Y. 1965))).
185. McFarland, supra note 55, at 268-69; see also Colt Def. LLC v. Heckler & Koch
Def., Inc., No. 2:04CV258, 2004 U.S. Dist. LEXIS 28690, at *12 (E.D. Va. Oct. 22, 2004)
(rejecting argument that transaction or occurrence requirement was satisfied because both
defendants were “attempting to commit ‘genericide’ on Colt’s M4 trademark”).
186. Consolidation under Rule 42(a) and the multi-district litigation statute only require a common question, but courts tend to require a degree of similarity approaching
what is required of permissive joinder. See supra Part III.A. In fact, when Congress prohibited joinder of similarly situated patent defendants, it also prohibited consolidation for trial
even though such claims certainly possess the single common question required by the
plain language of Rule 42(a). See 35 U.S.C. § 299(a) (Supp. V 2011).
2014]
AGGREGATING DEFENDANTS
1039
duty are unlikely to be factually related, even if similar (e.g., developing and selling different products alleged to infringe the same patent).
D. Cognitive Limitations
Cost and substantive explanations only partially explain defendants’ widespread resistance to being aggregated. Since defendants
should realize significant economies of scale from being aggregated
and, in some cases, lack economic or strategic reasons to oppose collective resolution, perhaps defendants’ opposition to being aggregated
is not fully motivated by rational concerns. Information problems
may exist, with defendants and their counsel unaware of the benefits
of being aggregated (given that defendant aggregation is rare and
novel compared to plaintiff aggregation) or failing to adequately distinguish between plaintiff aggregation (which is generally against
defendants’ interests) and defendant aggregation (which often aligns
with defendants’ interests).
Alternatively, some of the resistance to defendant aggregation
may result from cognitive biases—biases and aversions that can lead
to inaccurate perceptions of what is and is not in defendants’ interest.187 There is good reason to think that cognitive biases are part of
the explanation, as the resistance to defendant aggregation has been
vehement even though the economic and strategic arguments are, at
best, ambiguous. Anecdotal evidence points to three relevant cognitive limitations: bias against forced groups, self-serving biases, and
loss aversion.
1. Involuntary Groups
Research in a variety of fields indicates that, compared to individuals, groups perform better, make better decisions, and are better
problem-solvers.188 Admittedly, aggregated plaintiffs generally choose
their litigation group,189 whereas similarly situated defendants
187. Christine Jolls, Cass R. Sunstein & Richard H. Thaler, A Behavioral Approach to
Law and Economics, in BEHAVIORAL LAW AND ECONOMICS 13, 14-15 (Cass R. Sunstein ed.,
2000); Cass R. Sunstein, Introduction to BEHAVIORAL LAW AND ECONOMICS 1, 3-5 (Cass R.
Sunstein ed., 2000) [hereinafter Sunstein, Introduction].
188. See Martin Kocher et al., Individual or Team Decision-Making—Causes and Consequences of Self-Selection, 56 GAMES AND ECON. BEHAV. 259, 259-60, 268 (2006). See generally CASS R. SUNSTEIN, INFOTOPIA: HOW MANY MINDS PRODUCE KNOWLEDGE (2006) (discussing how aggregating knowledge and perspectives can produce more accurate information and decisions).
189. See Freer, supra note 66, at 823-24. Joinder by definition requires a voluntary
choice by plaintiffs. Id. Common-question class actions include an opt-out mechanism,
while other types of class actions normally track groups that pre-date litigation. See Elizabeth Chamblee Burch, Aggregation, Community, and the Line Between, 58 U. KAN. L. REV.
889, 890-92 (2010) (describing, though questioning, reliance on pre-existing groups). Consolidation via Rule 42(a) or multi-district litigation can be initiated by the court sua sponte
or on a defendant’s motion, but plaintiffs are more likely to initiate these procedures. See
1040
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
normally are forced to litigate together by the plaintiff with no ability
to avoid group litigation if the requisite commonality in the claims
is present.190 Yet, research demonstrates that self-selection of groups
is not a prerequisite for successful group performance and involuntarily selected groups are as capable of successful and cohesive performance as voluntarily selected groups.191 Thus, a rational defendant
would seem to prefer litigating as part of a defendant group, rather
than individually.
Despite these benefits of even involuntary groups, people are often
biased against working in forced groups, resisting participation, expressing skepticism of the group’s objectives, showing distrust of fellow group members, being less cooperative, and being worse at resolving conflicts.192 These biases against forced groupings are exacerbated when, as with aggregated defendants, the group is chosen by
an adversary, not a neutral. “Reactive devaluation” suggests that
“proposals made by adversaries will be valued lower than identical
proposals made by a neutral party or a member of one’s own group”193
as a result of zero-sum or fixed-pie bias (i.e., the assumption that a
gain for the opponent equals a loss for self).194
To overcome initial biases and perform effectively, an involuntary
group often requires a collective goal that generates a task-based cohesion.195 Aggregated defendants lack an inherently collective goal,
as each defendant’s goal is to defeat the plaintiff’s claim with whatever combination of individual and common defenses is best for that
Olson, supra note 66, at 360-63; Notice of Hearing Session, (J.M.P.L. July 25, 2013) (MDL No.
875), available at http://www.jpml.uscourts.gov/sites/jpml/files/Hearing_Order-7-25-13.pdf (listing plaintiff as moving party for multi-district proceedings in twelve of sixteen matters).
190. See Freer, supra note 66, at 823-26; Parchomovsky & Stein, supra note 23, at 1522.
191. See Dwight R. Norris & Robert E. Niebuhr, Group Variables and Gaming Success,
11 SIMULATION & GAMES 301, 306-09 (1980).
192. See, e.g., Ronald Rooney & Michael Chovanec, Involuntary Groups, in HANDBOOK
OF SOCIAL WORK WITH GROUPS 212, 215-16 (Charles D. Garvin et al. eds., 2004) (involuntary psychotherapy groups); Work with Involuntary Groups, in STRATEGIES FOR WORK
WITH INVOLUNTARY CLIENTS 244, 246, 256 (Ronald H. Rooney ed., 2d ed. 2009) (involuntary
psychotherapy groups); Donald R. Bacon et al., Lessons from the Best and Worst Student
Team Experiences: How a Teacher Can Make the Difference, 23 J. MGMT. EDUC. 467, 468,
482 (1999) (management/business education); Kenneth J. Chapman et al., Can’t We Pick
Our Own Groups? The Influence of Group Selection Method on Group Dynamics and Outcomes, 30 J. MGMT. EDUC. 557, 565 (2006) (management/business education); Paul Miesing
& John F. Preble, Group Processes and Performance in a Complex Business Simulation, 16
SMALL GROUP BEHAV. 325, 334-36 (1985) (management/business education).
193. Maurits Barendrecht & Berend R. de Vries, Fitting the Forum to the Fuss with
Sticky Defaults: Failure in the Market for Dispute Resolution Services?, 7 CARDOZO J.
CONFLICT RESOL. 83, 96-97 (2005).
194. See Robert S. Adler, Flawed Thinking: Addressing Decision Biases in Negotiation,
20 OHIO ST. J. ON DISP. RESOL. 683, 740-42 (2005).
195. Norris & Niebuhr, supra note 191, at 306-09.
2014]
AGGREGATING DEFENDANTS
1041
defendant, regardless of the outcome for the rest of the group.196 In
fact, aggregated defendants that are competitors have an incentive to
favor individual defenses over class or general defenses exactly because they can prevail without benefitting their competitors.197
Anecdotal evidence supports the intuition that biases against involuntary groups create resistance to defendant aggregation. Defendants seeking de-aggregation have argued “that it is inherently
unfair to join separate defendants in the same proceeding when they
are competitors,”198 that is, to force a group where none would otherwise exist. Similarly, even though Rule 20 focuses on the relationship
of the claims, arguments often focus on the lack of relationship or existence of competition among the defendants, that is, the lack of a preexisting group, even when the claims have common roots.199
2. Self-Serving Bias
Aside from, or perhaps in conjunction with, biases against forced
groupings, “[p]eople tend to make judgments about themselves, their
abilities, and their beliefs that are ‘egocentric’ or ‘self-serving.’ ”200
This bias can manifest itself in several ways, including “a tendency
for people to see themselves as having a greater than average share
of some desirable quality” or “skewed predictions; that which is desired is thought more likely to occur than that which is undesired.”201
Applied to the litigation context, self-serving bias suggests that “litigants, their lawyers, and other stakeholders might overestimate
their own abilities, the quality of their advocacy, and the relative
merits of the positions they are advocating.”202
In the litigation context, egocentrism is most commonly applied to
explain settlement failures,203 but it also suggests that each defendant will over-value the strength of its own defense and each defense
196. See Ford, supra note 156, at 93-112; see also Parchomovsky & Stein, supra note
23, at 1483 (categorizing defenses into general, class, and individual).
197. See Ford, supra note 156, at 109-12.
198. Sprint Commc’ns Co. v. Theglobe.com, Inc., 233 F.R.D. 615, 617 (D. Kan. 2006)
(emphasis added).
199. See, e.g., Spread Spectrum Screening, LLC v. Eastman Kodak Co., No. 10 C 1101,
2010 U.S. Dist. LEXIS 90549, at *6 (N.D. Ill. Sept. 1, 2010) (noting alleged common roots
between each defendant’s accused product, but emphasizing that “Kodak and Heidelberg
are business competitors of one another” in rejecting joinder); Pergo, Inc. v. Alloc, Inc., 262
F. Supp. 2d 122, 128 (S.D.N.Y. 2003).
200. Chris Guthrie & Jeffrey J. Rachlinski, Insurers, Illusions of Judgment & Litigation, 59 VAND. L. REV. 2017, 2042 (2006).
201. Ward Farnsworth, The Legal Regulation of Self-Serving Bias, 37 U.C. DAVIS L.
REV. 567, 569-70 (2003).
202. Guthrie & Rachlinski, supra note 200, at 2044.
203. Linda Babcock et al., Creating Convergence: Debiasing Biased Litigants, 22 LAW &
SOC. INQUIRY 913, 922-23 (1997).
1042
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
attorney will over-value her own abilities and strategies. This has
two important consequences for defendant aggregation. First, even
absent a rational reason to assume a stronger defense than codefendants, an egocentric defendant will perceive aggregation as detrimentally mixing its “stronger” position with the “weaker” positions
of the co-defendants. Second, if each defendant (and defense attorney) is overly confident in her own strategy, compromise will be more
difficult when coordinated action is required, leading to conflict and
increased coordination costs. Similarly, an egocentric defense attorney will be more likely to duplicate efforts (e.g., ask her own questions at depositions, revise every joint brief, etc.) rather than divide
labor and trust the work of other defense lawyers perceived as of
lower quality. This again increases coordination costs. Although
plaintiffs and plaintiffs’ attorneys should be subject to similar selfserving biases, contingent fee arrangements necessitate the division
of labor and reduction of coordination costs to maximize fees. Plaintiffs’ attorneys’ economic incentives thus mitigate egocentrism.
Anecdotal evidence again supports the role of egocentrism in multi-defendant cases. Practitioners report that managing aggregated
defendants in patent litigation “can sometimes be like herding cats,
with each defendant wanting to go in its own direction. Clients and
their counsel may have differing strategies, and it can be necessary
to remind co-defendants that healthy compromise may be necessary
to get everyone rowing in the same direction.”204
3. Loss Aversion
A common reaction to aggregation of similarly situated defendants
is that the problems created by collective resolution are worse for aggregated defendants, who face a potential judgment against them,
than for aggregated plaintiffs, who only seek to obtain new benefits.205 This reaction reflects loss aversion; that is, despite being economically equivalent, people “are more displeased with losses than they are
pleased with equivalent gains – roughly [sic] speaking, twice as displeased.”206 As a result, defendants will care more about paying $100 in
damages than plaintiffs will care about recovering $100 in damages.
Of course, loss aversion does not directly explain resistance to defendant aggregation; if aggregation helps defendants, a loss-averse
defendant should strongly favor it. But loss aversion does seem rele204. Baghdassarian & Frankel, supra note 151.
205. Cf. Phillips Petroleum Co. v. Shutts, 472 U.S. 797, 810-11 (1985) (suggesting that
due process protections are greater for defendants than plaintiffs because “an adverse
judgment typically [will not] bind an absent plaintiff for any damages” even though it will
“extinguish any of the plaintiff's claims which were litigated”).
206. Sunstein, Introduction, supra note 187, at 5.
2014]
AGGREGATING DEFENDANTS
1043
vant in two ways. First, defendants are more likely to overvalue the
potential problems with aggregation than are plaintiffs because they
are more concerned about their potential loss than plaintiffs are
about their potential gain. Second, the status quo is normally the reference point for determining whether something is a gain or loss.207
The individual lawsuit is the status quo for all procedural rules,208
particularly for the relatively uncommon and novel aggregation of
similarly situated defendants. Thus, defendants are likely to overvalue the “losses” they realize from being aggregated (as compared to
the individual lawsuit) and undervalue the gains.
E. Summary and the Role of Selection Effects
The justifications for defendants’ opposition to aggregation of similarly situated defendants fall on a spectrum. At one end, collective
resolution is clearly against defendants’ interests when the plaintiff’s
claim is so small that it could not be profitably brought individually,
such as in the BitTorrent copyright cases. Without aggregation, defendants would never be sued.
At the other end, defendants have no economic or strategic reasons to oppose being aggregated when there are asymmetric stakes,
such that the plaintiff’s claim against each defendant would be individually viable but each defendant’s cost of individual defense would
be greater than its expected liability. These asymmetric stakes can
arise, first, because the plaintiff has a cost advantage due to the lesser amount of discoverable information in its possession or its ability
to spread costs across multiple individual cases. This is often the case
with patent trolls.209 Alternatively, asymmetric stakes exist when the
plaintiff receives some additional benefit from the litigation beyond
the damages paid by the defendant, such as the deterrence sought by
the RIAA and DirecTV plaintiffs. A rational defendant in these circumstances would default or settle and is no worse off from aggregation regardless of any strategic consequences. To the contrary, these
defendants are better off because the economies of scale that aggregation offers increase the chances of a positive-value defense. Thus,
cognitive biases are the only possible explanation for resistance to
defendant aggregation when there are asymmetric stakes.
Between these two extremes lie cases where there is enough at
stake to justify the plaintiff in bringing litigation individually and
207. Id. at 6.
208. See David Marcus, From “Cases” to “Litigation” to “Contract”: A Comment on Stability in Civil Procedure, 56 ST. LOUIS U. L.J. 1231, 1232 (2012).
209. See Coffee, supra note 145, at 899; see also In re Genentech, Inc., 566 F.3d 1338,
1345 (Fed. Cir. 2009) (noting that patent defendants generally have more discoverable
information than patentees).
1044
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
each defendant to defend individually, as is true for most patent cases, even those brought by so-called trolls. In these cases, objecting
to being aggregated with similarly situated defendants may or may
not be rational for a defendant, depending on the fact-specific question of whether the strategic benefits of, for example, asymmetric
preclusion or avoiding jury confusion outweigh the economies of scale
of a collective defense. Part III.C suggests that the strategic considerations often will not outweigh the economies of scale aggregation
and therefore defendants opposing aggregation in these cases often
will be acting irrationally.
Perhaps selection effects provide a simple explanation for why defendants oppose being aggregated. The plaintiff generally controls
whether or not to aggregate defendants,210 and it is reasonable to
think that aggregation only occurs when it is in the plaintiff’s selfinterest and, presumably, against the defendants’ self-interest. Defendants’ consistent resistance to being aggregated thus may result
from selection effects—they are only aggregated in those cases where
it is detrimental to their interests. The limited cases in which the
plaintiff aggregates defendants even when it is beneficial to defendants may simply go unnoticed because no one objects to collective
resolution. Although perfectly plausible, there are three reasons to
doubt the simplicity of this explanation.
First, defendants appear to oppose aggregation even when it is not
a self-interested choice by the plaintiff. Defendants have some mechanisms to seek aggregation even if the plaintiff chooses individual
litigation, yet rarely use them.211 For example, in the RIAA and DirecTV cases where the plaintiff proceeded individually, the defendants did not request consolidation or multi-district litigation.
Second, aggregated defendants have objected to collective resolution even in the category of cases in which they should favor it, that
is, where a plaintiff with a positive-value claim sues a defendant with
a negative-value defense. For example, patent defendants pushed for
and obtained a blanket prohibition on joinder of similarly situated defendants, even in the many patent troll cases where a cost-differential
makes the patentee’s claim greater than its litigation costs but less
than the defendant’s cost of defense. Similarly, the RIAA and DirecTV
210. See Freer, supra note 66, at 823-26; Parchomovsky & Stein, supra note 23, at 1522.
211. Defendants appear to be able to request defendant class treatment, yet virtually
never do so. See Defendant Class Actions, supra note 20, at 630 n.3. Defendants also can
seek permissive intervention in similarly situated defendants’ cases, but never do so. See
Haley, supra note 162, at 334. Nor have similarly situated defendants made significant use
of opportunities to aggregate themselves via Rule 42(a) or multi-district consolidation. See
Olson, supra note 66, at 360-63; see also Notice of Hearing Session, (J.M.P.L. July 25, 2013)
(MDL No. 875), available at http://www.jpml.uscourts.gov/sites/jpml/files/Hearing_Order-725-13.pdf (listing plaintiff as moving party for multi-district proceedings in twelve of sixteen matters).
2014]
AGGREGATING DEFENDANTS
1045
defendants repeatedly opposed being aggregated despite their negative-value defenses and the likely positive-value claims.
Third, just because plaintiffs choose aggregation does not mean
that it is against the defendants’ interest. Multiple lawsuits are more
time-consuming and bothersome to the plaintiff than a single suit,
regardless of the strategic benefits.212 Moreover, the plaintiff may
have capital constraints that necessitate the immediate savings from
a collective lawsuit even if the long-term benefits favor defendants, or
the plaintiff may have agency problems if its lawyer realizes benefits
from aggregating defendants that are not passed on to the plaintiff
(e.g., less work in a contingency fee arrangement). Finally, aggregation is not necessarily zero-sum but instead may reduce the transaction costs for both parties or offer both the plaintiff and defendants
benefits that are not mutually exclusive. For example, a plaintiff suing multiple defendants internalizes the entire risk from asymmetric
preclusion, whereas each defendant only internalizes a part of the
benefit, making elimination of asymmetric preclusion a greater gain
for the plaintiff than loss for each defendant. Thus, both the plaintiff
(from eliminating asymmetric preclusion) and the defendants (from
economies of scale) could be better off from defendant aggregation.
IV. DEFENDANT AGGREGATION FROM THE SYSTEMIC PERSPECTIVE
Part III identified several potential problems with defendant aggregation, finding them more limited or weaker than may be initially
thought. But Part III focused only on whether defendant aggregation
was problematic for defendants. That defendant aggregation may be
bad for defendants in some circumstances tells us little about whether it is socially optimal. To the contrary, it may be socially optimal
exactly because it is against defendants’ interest, for example, by increasing the chances that wrongful conduct will be remedied. From a
societal or systemic perspective, there are two key questions. The
first is the economic question of whether aggregating similarly situated defendants makes litigation more efficient or more costly.213 The
second is the substantive question of whether defendant aggregation
promotes the substantive remedial scheme by increasing the chances
that wrongdoers are found liable and innocent parties escape liability
or whether defendant aggregation distorts the substantive remedial
scheme by permitting wrongdoers to escape liability or by imposing
liability on innocent actors.214
212. See Freer, supra note 66, at 824 (arguing that plaintiffs are unlikely to omit defendants).
213. See Rosenberg, supra note 4, at 562-65.
214. See Nagareda, Discontents, supra note 134, at 1874-78. Common-question plaintiff
class actions have been accused of distorting substantive law by increasing the costs of
1046
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
As discussed earlier, defendant aggregation is socially optimal
from a theoretical perspective, as it eliminates duplicative litigation,
promotes deterrence by permitting negative-value claims and defenses to be brought, and encourages resolution based on the merits, not
costs.215 Yet, in practice, courts, policymakers, and commentators
have been highly critical of efforts to aggregate similarly situated defendants. This Part considers possible justifications for this systemic
resistance, finding them insufficient to reject defendant aggregation
wholesale but suggestive of the proper procedures for aggregating
similarly situated defendants. The first three Sections consider
whether systemic opposition to defendant aggregation is warranted
based on its potential negative consequences for defendants, with
Section A addressing the economic consequences for defendants, Section B the substantive consequences for defendants, and Section C
the possibility of capture by defendant interests. Section D considers,
and rejects, the possibility that systemic resistance to defendant aggregation is warranted based on negative consequences for judges or
the judicial system.
A. Cost-Differentials, Weak Claims, and Unsympathetic Plaintiffs
Defendants sometimes have sound economic reasons to resist being aggregated, despite its theoretical economies of scale, because
aggregation offers cost-savings to plaintiffs that allow litigation that
otherwise would not be brought or accentuates pro-plaintiff costdifferentials by imposing coordination costs on defendants.216 However, the fact that aggregating defendants makes some otherwise negative-value claims viable is a reason for the judicial system to embrace
defendant aggregation, not resist it, as this promotes optimal deterrence and secures faithful implementation of the substantive remedial regime.217 Likewise, aggregated defendants’ potentially high coordination costs are largely a result of defense attorneys’ fee structure,218 and systemic resistance is not warranted simply to preserve
the fees of defense attorneys, countenance inefficient lawyer-client
arrangements, or stifle innovation in fee arrangements.
However, courts and policymakers may be concerned that the additional litigation and cost-differentials created by defendant aggregation encourage strike suits, that is, non-meritorious claims that are
defense or increasing the risks of litigation through the massive potential liability. See
Resnik, supra note 24, at 16.
215. See supra Part III.C.
216. See supra Parts IV.A-B.
217. See supra Part III.C.
218. See supra Part IV.B.
2014]
AGGREGATING DEFENDANTS
1047
settled for less than the cost of defense.219 For example, one district
court cited a BitTorrent plaintiff’s prior “abusive litigation practices,”
“coercive settlements,” and “thus-far-unsubstantiated and perhaps
erroneous allegation”220 in rejecting defendant joinder, while Congress relied on “abusive” litigation practices by patent assertion entities to justify the patent anti-joinder provision.221 De-aggregation and
re-filing may be seen as a way to eliminate individually negativevalue strike suits and make positive-value strike suits less profitable.
The Supreme Court recently cautioned against allowing merits
questions to drive the aggregation decision.222 Addressing strike suits
through de-aggregation has spillover effects, as it creates doctrine—
and in the case of patent law, statutory provisions—hostile to defendant aggregation generally, foreclosing meritorious negative-value
suits that promote the substantive remedial scheme and optimal deterrence. Courts have more direct means that are better tailored to
addressing weak claims and strike suits, such as dismissal for failure
to state a claim upon which relief can be granted,223 sanctions for
harassing or frivolous litigation,224 and review of the merits before
entering a default judgment.225
More troubling, courts’ concern with the multi-defendant suits
may be that they are undesirable as a matter of policy, not that they
lack legal merit.226 Multi-defendant cases have often involved unsym219. Next Phase Distrib., Inc. v. Does 1–27, 284 F.R.D. 165, 170-71 (S.D.N.Y. 2012)
(“[T]he Court recognizes that joining 27 defendants, a substantial number of whom may
have no liability in this case, in a copyright infringement case when the copyright itself
might be deemed invalid, could prove to be a costly and futile exercise for Next Phase and
the Court, and a damaging an [sic] unnecessary ordeal for the John Does.”); Taylor, supra
note 7, at 674-75 (noting concern with strike suits in multi-defendant cases).
220. Digital Sins, Inc. v. Does 1–245, No. 11 Civ. 8170(CM), 2012 WL 1744838, at *3-4
(S.D.N.Y. May 15, 2012).
221. Taylor, supra note 7, at 702.
222. See Amgen Inc. v. Conn. Ret. Plans & Trust Funds, 133 S. Ct. 1184, 1194-95
(2013) (“Rule 23 grants courts no license to engage in free-ranging merits inquiries at the
certification stage. Merits questions may be considered to the extent—but only to the extent—that they are relevant to determining whether the Rule 23 prerequisites for class
certification are satisfied.”).
223. FED. R. CIV. P. 12(b)(6); see also Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009) (describing what is needed for a complaint to survive a motion to dismiss for failure to state a claim).
224. FED. R. CIV. P. 11(b)(2).
225. In re DIRECTV, Inc., No. C–02–5912–JW, 2004 WL 2645971, at *7-12 (N.D. Cal.
July 26, 2004).
226. See Ingenuity 13 LLC v. Doe, No. 2:12–cv–8333–ODW (JCx), 2013 WL 1898633, at
*1 (C.D. Cal. May 6, 2013) (noting that BitTorrent plaintiffs “outmaneuvered the legal
system” by “discover[ing] the nexus of antiquated copyright laws, paralyzing social stigma,
and unaffordable defense costs” (emphasis added)); DIRECTV, 2004 WL 2645971, at *7-12
(finding that most of DirecTV’s claims had sufficient legal merit to warrant a default judgment); Karunaratne, supra note 14, at 303 (noting that file-sharing plaintiffs “generally
have legitimate substantive grounds for their allegations of copyright infringement” but
still suggesting need to make litigation less profitable for them).
1048
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
pathetic plaintiffs: patent trolls;227 the “porno-trolling” BitTorrent
plaintiffs accused of fraud, extortion, and exploiting antiquated copyright laws and social stigma;228 and large corporate interests like DirecTV and the RIAA suing individuals or small businesses.229 Many
today view: patents as too prevalent or too broad;230 copyright as a
poor fit for the internet;231 and large corporate interests suing financially limited individuals as not a proper use of judicial resources.232
Using de-aggregation as a de facto reform of substantive law is
inconsistent with the ideal that “the format for the resolution of civil
disputes . . . should not alter substantive law.”233 Of course, it would
not be the first time that “[t]he affording or withholding of aggregate
treatment . . . operates as a backdoor vehicle to restructure the remedial scheme in applicable substantive law.”234 However, judicial
nullification of substantive law via de-aggregation hinders the development of substantive law. Legislators who know judges will use procedural tools to avoid implementation of substantive law when it
seems the most unfair or antiquated will have little incentive to undertake efforts to repeal or amend the law. For example, to the extent
that the patent anti-joinder provision decreases litigation by patent
assertion entities, it could relieve some of the pressure on Congress
or the Federal Circuit to adjust substantive patent law doctrines that
create broad patents, even though the problem of broad patents is not
limited to multi-defendant cases or patent assertion entities.
B. Substantive Effects, Fairness, Autonomy, and Due Process
Aside from its effects on litigation costs, defendants’ resistance to
being aggregated may be motivated by non-economic concerns. Three
of defendants’ possible concerns are clearly inapplicable from the systemic perspective. Asymmetric preclusion, though potentially beneficial to defendants, is generally seen as undesirable from a systemic
perspective because it encourages gamesmanship and duplicative
227. See Bryant, supra note 23, at 690-94.
228. See Ingenuity, 2013 WL 1898633, at *1.
229. See supra Part III.D.2-3.
230. See, e.g., Charles Duhigg & Steve Lohr, The Patent, Used as a Sword, N.Y. TIMES
(Oct. 7, 2012), http://www.nytimes.com/2012/10/08/technology/patent-wars-among-tech-gia
nts-can-stifle-competition.html?pagewanted=all&_r=0.
231. See, e.g., Jessica Litman, Real Copyright Reform, 96 IOWA L. REV. 1, 1 (2010).
232. See, e.g., Opderbeck, supra note 123, at 1727.
233. Nagareda, Discontents, supra note 134, at 1874-78.
234. Id. at 1877-78. Professor Nagareda argues “that aggregate procedure is under
constant pressure—sometimes from plaintiffs and sometimes from defendants—to serve as
the vehicle for reform of underlying substantive law through means other than actual reform legislation.” Id. at 1877.
2014]
AGGREGATING DEFENDANTS
1049
litigation. Aggregation procedures are often justified exactly because
they eliminate asymmetries in the preclusive effect of judgments.235
Similarly, doctrinal opportunities to de-aggregate may offer defendants some benefits and may explain some recent court decisions
taking a narrow doctrinal focus,236 but they do not explain why
the doctrine developed or has persisted in this manner, or why Congress has endorsed the doctrine. Permitting joinder based on overlap
in the underlying legal duty or the harm caused to the plaintiff, not
just the defendants’ allegedly unlawful conduct, is equally consistent
with Rule 20’s plain language and policy goal, that is, promoting efficiency by avoiding separate litigation where there is a substantial
evidentiary overlap.237
Finally, judges and policymakers are not immune from cognitive
biases,238 but biases against participation in involuntary groups, selfserving biases, and loss aversion more directly explain why defendants would conclude aggregation is against their interests than why
judges or policymakers would.239
On the other hand, the risk of judicial or jury confusion is a significant systemic concern and is often cited by courts in ordering deaggregation.240 Allowing a defendant to be found liable, or escape liability, simply because the jury or judge misattributed evidence or arguments against another defendant could distort the substantive remedial regime. Of course what this means for defendant aggregation
depends on how strong and extensive this risk is and whether the
other ways in which aggregation promotes the substantive remedial
scheme are greater than the distortions created by confusion. In any
event, courts’ severance orders, as well as Congress’s patent
anti-joinder provision, are written broadly to bar aggregation of similarly situated defendants generally, even in the many situations in
235. See 18A CHARLES ALAN WRIGHT, ARTHUR R. MILLER & EDWARD H. COOPER,
FEDERAL PRACTICE AND PROCEDURE § 4464, at 692-93 (2d ed. 2002); Developments in the
Law, supra note 64, at 1394.
236. See Taylor, supra note 7, at 701.
237. See McFarland, supra note 55, at 268-70; see also Kane, supra note 66, at 1729-30,
1746 (explaining that the policy goal of joinder rules is to encourage efficient resolution of
claims, especially when there is factual overlap).
238. See Chris Guthrie et al., Inside the Judicial Mind, 86 CORNELL L. REV. 777, 778 (2001).
239. See supra Part III.D.1–3. On the other hand, courts in other contexts have been
skeptical of forced associations. See, e.g., Boy Scouts of Am. v. Dale, 530 U.S. 640,
648 (2000) (right of expressive organizations to exclude members); Jay M. Feinman,
The Development of the Employment at Will Rule, 20 AM. J. LEGAL HIST. 118 (1976)
(employment relationships).
240. See, e.g., Colt Def. LLC v. Heckler & Koch Def., Inc., No. 2:04CV258, 2004 U.S.
Dist. LEXIS 28690, at *15 (E.D. Va. Oct. 22, 2004).
1050
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
which there is little risk of problematic confusion.241 Moreover, courts
and policymakers have more tailored ways to address potential
confusion short of de-aggregation, such as requiring that common
questions predominate or satisfy some other threshold of similarity,
asking whether individual issues are so prevalent as to make aggregate resolution unduly burdensome,242 or holding separate trials for
individual issues.243
There also seems to be some sense that aggregating similarly situated defendants is unfair or contrary to due process. This largely
reflects concerns about forced groupings, autonomy, or potential confusion that have already been addressed. There may be some additional belief that defendants facing liability are entitled to pursue
whatever strategy they choose without having to compromise or be
affected by defendants pursuing other strategies. This contention is
doubtful. Even without aggregation, defendants will often be effectively bound by the strategic decisions of earlier defendants via principles like stare decisis, and it is equally unfair to require earlier defendants to bear the burden of defense alone.244 Moreover, extreme
defendant autonomy creates externalities unfair to third parties by
consuming judicial resources that otherwise could be spent on other
cases or other socially beneficial activities. Admittedly, joint resolution of individual issues and judicial case management procedures
that limit the ability of defendants to present defenses on individual
issues may raise greater fairness or due process concerns, but this
problem is with the aggregation procedures, not aggregation itself.
C. Capture
Even if the potential problems aggregation creates for defendants
are not troubling from a systemic perspective, these concerns could
still explain systemic resistance based on a “capture” theory. Public
choice theory predicts that policy outcomes will favor concentrated
groups with high individual stakes, such as organized corporate interests, because more diffuse interests, such as taxpayers, consumers, or citizens, generally suffer greater free-rider problems and are
more difficult to mobilize.245 Because defendants are more likely to be
241. See Taylor, supra note 7, at 708-15 (describing situations where joinder is likely
not permitted despite little risk of harmful confusion regarding patent infringement); supra Part IV.C.2.
242. See Erbsen, supra note 28, at 1002.
243. See, e.g., Jaime Dodge, Disaggregative Mechanisms: The New Frontier of MassClaims Resolution Without Class Actions, 63 EMORY L.J. (forthcoming 2014), available at
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2243032.
244. See supra Part IV.C.1.
245. See Daniel A. Farber, Introduction to PUBLIC CHOICE AND PUBLIC LAW ix, x (Daniel A. Farber ed., 2007).
2014]
AGGREGATING DEFENDANTS
1051
large corporate interests and plaintiffs to be individuals, public
choice theory would predict that procedural rules are more likely to
reflect the interests of defendants than plaintiffs.
For example, Congress is widely seen as enacting the anti-joinder
provision for patent cases at the behest of corporate interests, largely
in the high technology area, which are frequently targeted by small
patent assertion entities.246 However, in other prominent examples,
concentrated corporate plaintiffs (e.g., DirecTV and RIAA) sought to
aggregate diffuse individual defendants.247 Therefore, resistance to
defendant aggregation is the opposite of public choice predictions.
Moreover, the public choice literature debates to what extent the independent judiciary is subject to interest group pressures.248
Whatever the descriptive power of public choice theory for systemic opposition to defendant aggregation, it does not provide a normatively sound justification on which to deny aggregation of similarly
situated defendants. However, it does provide an important insight.
Overcoming systemic resistance to defendant aggregation and obtaining the efficiencies and other benefits it offers likely requires blunting defendants’ opposition to being aggregated. Thus, a procedural
mechanism that mitigates the problems for defendants created by
aggregation is more likely to be adopted and be effective in practice.
D. Judicial Self-Interest
A judge may resist aggregating defendants, not because it is
against defendants’ interests, but because it is against the judge’s
own self-interest.249 In granting de-aggregation, judges sometimes
cite self-interested justifications that are weak from a disinterested
perspective, such as the plaintiff’s circumvention of filing fees or the
single credit the judge gets for purposes of caseload distribution.250
Judges also often cite concerns that multi-defendant cases will
require too much effort to manage, contending, for example, that aggregation lacks “litigation economies” because each defendant “is
246. See Bryant, supra note 23, at 701-02 & n.102.
247. See supra Part II.D.2-3.
248. See William M. Landes & Richard A. Posner, The Independent Judiciary in an
Interest-Group Perspective, 18 J.L. & ECON. 875 (1975), reprinted in PUBLIC CHOICE AND
PUBLIC LAW 22, 22-24 (Daniel A. Farber ed., 2007).
249. See Richard A. Posner, What Do Judges and Justices Maximize? (The Same Thing
Everybody Else Does), 3 SUP. CT. ECON. REV. 1 (1993). On the other hand, aggregation
(whether of plaintiffs or defendants) may be in judges’ self-interest, since handling “big”
cases is a way for a judge to enhance prestige and reputation. Id. at 13-15.
250. Order Severing Parties Due to Misjoinder and Dismissing All but the First Named
Defendant at 2, Finisar Corp. v. Source Photonics, Inc., No. C 10–00032 WHA (N.D. Cal.
dismissed May 5, 2010).
1052
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
likely to have some individual defense to assert.”251 Of course, the fact
that there may be “some individual defense” says little about litigation
economies without also considering the common questions that exist,
the relative importance of common and individual questions in the
particular case, and the ability to efficiently resolve individual questions through means short of de-aggregation, such as separate trials.
Moreover, the concern that multi-defendant litigation is too complex to handle collectively reflects a narrow focus on the self-interest
of the individual judge in the aggregated proceedings, rather than
the interests of the judiciary as a whole. Individual litigation requires
resolution of both individual and common issues for each defendant.
By contrast, aggregate litigation allows resolution of common issues
collectively for all defendants, even if it still necessitates some form
of separate proceedings to resolve individual issues. Unless common
issues are minor, greater systemic efficiency should result from aggregate proceedings (even if individual issues must be addressed
separately) than individual proceedings (where both individual and
common issues must be addressed separately). Managing individual
issues in aggregate litigation may be more work for the individual
judge, but it normally will be less work for judges as a whole than
numerous individual cases. Thus, concerns about the complexity of
aggregated litigation are unpersuasive from a systemic perspective.
V. OVERCOMING RESISTANCE TO DEFENDANT AGGREGATION
To this point, this Article has focused on the why (and the what) of
defendant aggregation, explaining that aggregating defendants is
socially desirable because it can improve efficiency, deterrence, and
compensation. This Part turns to the how of defendant aggregation,
identifying how defendant aggregation can be implemented while
minimizing the problems it may create. Minimizing the problems
from aggregation both enhances its social desirability and, more
pragmatically, increases the chances that courts and policymakers
will actually permit collective resolution of claims against similarly
situated defendants.
This Part does not purport to definitively resolve the proper procedural mechanism for aggregating defendants, nor does it purport to
work out all of the logistical considerations. That will have to wait for
further work, debate, and practical experience. Rather, this Part has
three goals: to question the common assumption that defendant aggregation is synonymous with defendant class actions, to identify key
251. Digital Sins, Inc. v. Does 1–245, No. 11 Civ. 8170(CM), 2012 WL 1744838, at *3
(S.D.N.Y. May 15, 2002) (emphasis added); see also Next Phase Distribution, Inc. v. Does
1–27, 284 F.R.D. 165, 169-70 (S.D.N.Y. 2012); Colt Def. LLC v. Heckler & Koch Def., Inc.,
No. 2:04CV258, 2004 U.S. Dist. LEXIS 28690, at *15-16 (E.D. Va. Oct. 22, 2004).
2014]
AGGREGATING DEFENDANTS
1053
principles to guide design of defendant aggregation procedures, and
to sketch the initial outlines of a procedural mechanism that implements these principles: inter-district related case coordination.
A. Principles of a Defendant Aggregation Procedure
Regardless of the precise details, a procedural mechanism for aggregating similarly situated defendants should be nonrepresentative, subject to centralized (neutral) control and limited to
common issues. These features will both optimize defendant aggregation and reduce the resistance from defendants, courts, and policymakers that stands as a practical obstacle to achieving the benefits of
defendant aggregation.
1. Non-Representative Aggregation
Commentators often assume that if defendants are to be aggregated, the procedure must track the most common form of plaintiff
aggregation: the class action.252 This is probably unsurprising given
the scholarly, political, and popular obsession with the class action.
Yet, the possibilities for aggregation are far richer than just the representative class action and include procedures that permit collective
resolution while offering much greater protection for individual interests than the class action, like joinder and multi-district litigation.253
Admittedly, the class defense is probably the mechanism for aggregating defendants with the lowest litigation costs, since treating
the defendants as a unitary body with centralized representation
maximizes cost-savings and cost-spreading, while minimizing coordination costs.254 But it also accentuates the substantive concerns with
aggregating similarly situated defendants. Because it minimizes the
individualized nature of the claims and defenses, it increases the
chances of jury confusion and the possibility that evidence will be
misattributed in a way that improperly imposes or excuses liability.255
Moreover, a class defense is a representative procedure that would
impose liability on absent defendants based on the decisions and
strategies pursued by other defendants and their chosen counsel.
This raises autonomy concerns (to those who emphasize individual
252. See, e.g., Hamdani & Klement, supra note 4; Netto, supra note 21; Shen, supra
note 22; Simpson & Perra, supra note 20; Matthew K.K. Sumida, Note, Defendant Class
Actions and Patent Infringement Litigation, 58 UCLA L. REV. 843, 853-57 (2011). But see
Parchomovsky & Stein, supra note 23, at 1513-15 (semi-voluntary joinder); Edward Hsieh,
Note, Mandatory Joinder: An Indirect Method for Improving Patent Quality, 77 S. CAL. L.
REV. 683 (2004).
253. See supra Part III.A.
254. See Hamdani & Klement, supra note 4, at 711-13.
255. See Erbsen, supra note 28, at 1009-14 (recognizing the problem of jury confusion
in context of plaintiff class actions).
1054
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
autonomy)256 and is likely to generate practical opposition from
defendants, courts, and policymakers. Of course, defendants could be
permitted to opt out,257 minimizing autonomy concerns, but this
would likely undermine the practical benefits of defendant aggregation. Many defendants likely would opt out because: the plaintiff’s
claims would not be individually viable; other self-interested reasons
exist; or self-serving biases make them unwilling to hand over
control of their defense to the class representative.258 If opting out
were not permitted, a mandatory class defense would likely run afoul
of due process protections, at least if damages were at stake,259 as
well as raise serious concerns about jury confusion, autonomy, and
involuntary groupings.
Thus, defendant aggregation is unlikely to be adopted by courts
and policymakers in practice unless it uses a non-representative procedure—more akin to joinder, consolidation, or multi-district litigation—where the claim against each defendant remains separate and
each defendant can retain its own lawyer and make its own decision
regarding settlement.
2. Centralized Control Over Aggregation
Professor Richard Freer has argued that aggregation generally
should be controlled by a neutral judge, rather than the plaintiff, to
minimize duplicative litigation, maximize efficiency, and avoid the
whims of the plaintiff’s strategic interests.260 Centralized neutral control is a particularly sound principle for defendant aggregation.
Plaintiff aggregation is generally in the interest of plaintiffs, and
therefore the plaintiffs’ incentives will normally line up with the sys256. See, e.g., REDISH, supra note 48, at 169-73 (raising autonomy concerns in context
of plaintiff class actions).
257. See FED. R. CIV. P. 23(c)(2)(B)(v).
258. Hamdani and Klement struggle with opt-out. They suggest that plaintiffs opt out
only in low numbers and conclude that this will likely be true of defendants, overlooking
the effects of loss aversion on defendants. See Hamdani & Klement, supra note 4, at 725.
Although they suggest that defendants are unlikely to opt out because they will not want
to identify themselves to the plaintiff, in the primary examples of defendant aggregation,
the defendants were either already known to the plaintiff (patent and DirecTV litigation)
or easily identifiable through other means (subpoenas to ISPs in the file-sharing litigation).
See id. at 722. Moreover, because claims against similarly situated defendants normally
involve money damages, the potential defendants would already need to be identified for
purposes of notice if they are to be bound by the class judgment. See FED. R. CIV. P.
23(c)(2)(B); Phillips Petroleum Co. v. Shutts, 472 U.S. 797, 811-13 (1985). Ultimately,
Hamdani and Klement acknowledge that “[t]he barriers to opting out pose a major challenge to the legitimacy of the class defense as the opt-out option is important in preserving
due process rights.” Hamdani & Klement, supra note 4, at 722.
259. See Phillips Petroleum, 472 U.S. at 811-13; see also Netto, supra note 21, at 68
(proposing mandatory defendant class actions without considering opt -out and due
process requirements).
260. See Freer, supra note 66, at 812-13.
2014]
AGGREGATING DEFENDANTS
1055
temic interest in collective resolution.261 By contrast, who benefits
from defendant aggregation is more complex and dependent on the
facts of particular cases; therefore, the plaintiffs’ incentives will not
always correspond to the systemic interest in collective resolution.262
Defendant aggregation at the plaintiff’s behest, such as permissive
joinder, has provoked resistance from defendants, courts, and policymakers. The different stakeholders likely resist aggregation because the plaintiff is perceived as aggregating only when it benefits
and when the defendants suffer from aggregation. Alternatively, resistance could stem from biases against forced groupings, reactive
devaluation, and zero-sum biases.263 On the other hand, defendant
aggregation at the defendants’ behest, as some have suggested,
wound tend to be underutilized, as is true of the existing procedures
defendants have to aggregate themselves.264 Self-interested concerns
and self-serving biases would lead defendants to think they were better off litigating alone.265 Finally, if the aggregation decision were left
to the individual presiding judge, the self-interest of that judge may
cause him or her to reject aggregation even when beneficial to the
judicial system as a whole.
Thus, to best overcome obstacles to defendant aggregation, a centralized body representing the judicial system’s interests should be
provided with information about related cases and then allowed to
determine sua sponte whether to aggregate them.266
3. Issue-Only Aggregation
Most aggregative devices presumptively apply to entire cases, not
merely common issues. Permissive joinder applies to entire cases,
though a court is permitted to order separate trials or take other precautions “to protect a party [from] embarrassment, delay, expense, or
other prejudice . . . .”267 Similarly, the multi-district litigation statute
generally provides for transfer of the entire “civil action[],” though
the statute allows the panel to separate and remand “any claim,
cross-claim, counter-claim, or third-party claim” to the original
judge.268 At the same time, the Judicial Panel on Multi-District Litigation has concluded that it “does not have power to separate issues
261. See supra Part III.E.
262. See supra Part III.E.
263. See supra Part III.
264. See Freer, supra note 66, at 823-26; Parchomovsky & Stein, supra note 23, at 1522.
265. See supra Part III.
266. Cf. Freer, supra note 66, at 841-51 (proposing duty to notify court of potential
duplicative litigation and court determination of whether to “package” litigation).
267. FED. R. CIV. P. 20(b).
268. 28 U.S.C. § 1407(a) (2012).
1056
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
in civil actions, assigning one or more to the transferee court and one
or more to transferor courts.”269 Although Rule 23 appears to provide
for class actions limited to specific common issues, with resolution of
other issues left for subsequent individual cases, issue class actions remain an exception.270 Consolidation under Rule 42(a) is the only aggregative device that does not presumptively apply to the entire case.271
When entire cases are aggregated, a mix of common and individual issues will exist, creating concerns about fact-finder confusion,
fairness and autonomy, and coordination costs. Moreover, the different portfolio of individual and common defenses possessed by each
defendant can hinder the development of a shared-task focus to overcome bias against forced groupings and can create room for egocentric perceptions of the relative strengths of the different defendants’
positions. Limiting aggregation to common issues, with individual
issues resolved separately outside of the group litigation, eliminates
or minimizes these problems.
B. Inter-District Related Case Coordination
1. Overview of Inter-District Coordination
Multi-district litigation is a good starting point for a defendant
aggregation procedure because it offers a non-representative structure, uses a centralized, neutral body to make the aggregation decision, and allows the exercise of nationwide jurisdiction and venue.272
However, multi-district litigation suffers from three shortcomings
that undermine its effectiveness for aggregating similarly situated
defendants. First, in practice, multi-district litigation only happens
at the request of one of the parties, with only four percent of proceedings initiated via the panel’s authority to order aggregation sua sponte.273 Second, multi-district litigation normally applies to entire cases,
269. In re A.H. Robins Co., 610 F. Supp. 1099, 1101 (J.P.M.L. 1985); see also 15
CHARLES ALAN WRIGHT, ARTHUR R. MILLER, EDWARD H. COOPER & RICHARD D. FREER,
FEDERAL PRACTICE AND PROCEDURE § 3862, at 451 (3d ed. 2013) (“The ability to subdivide
a case is limited . . . . The Panel will not separate and transfer discrete issues for [multidistrict litigation] . . . .”).
270. See Jon Romberg, Half a Loaf is Predominant and Superior to None: Class Certification of Particular Issues Under Rule 23(c)(4)(A), 2002 UTAH L. REV. 249, 251-54, 266-67.
But see Laura J. Hines, Challenging the Issue Class Action End-Run, 52 EMORY L.J.
709, 713-14, 763-64 (2003); see also Laura J. Hines, The Dangerous Allure of the Issue
Class Action, 79 IND. L.J. 567, 568 (2004) (questioning permissibility and desirability of
issue classes).
271. FED. R. CIV. P. 42(a)(1) (allowing the court to “join for hearing or trial any or all
matters at issue in the actions”); 9A CHARLES ALAN WRIGHT & ARTHUR R. MILLER,
FEDERAL PRACTICE AND PROCEDURE § 2382, at 9-10 (3d ed. 2008).
272. 28 U.S.C. § 1407(a), (c) (2012); see also In re FMC Corp. Patent Litig., 422 F. Supp.
1163, 1165 (J.P.M.L. 1976).
273. See § 1407(c)(i); Lee et al., supra note 16, at 4.
2014]
AGGREGATING DEFENDANTS
1057
or at least claims, and therefore aggregates individual issues as well
as common issues.274 Third, aggregation of cases or claims via multidistrict litigation is only for the purpose of pre-trial proceedings and
does not allow ultimate resolution of common issues via trial.275
To overcome these shortcomings, the proposed inter-district coordination procedures are more akin to the related case procedures
most district courts use for intra-district coordination, except applied
across district lines. Most district courts require a notice of related
cases to be filed with the complaint that identifies any other action
previously filed or currently pending in the same district that, inter
alia, involves the “determination of the same or substantially related
or similar question[s] of law and fact” or “would entail substantial
duplication of labor if heard by different judges.”276 In many districts,
assignment or transfer of related cases to the same judge is automatic. For example, in the Southern District of California, the clerk is
tasked with identifying cases that meet the definition of related cases, with identified cases automatically assigned (or reassigned) to the
judge with the first-filed of the related actions.277
The procedures for resolution of these related cases vary depending on the district, the nature of the cases, and the degree of overlap.
Sometimes cases are consolidated for all purposes, while other times
they remain entirely separate with coordination simply allowing a
single judge to master the issues involved and avoid inconsistent results. But, importantly for present purposes, it is common in intradistrict related cases for the common issues to be resolved collectively
in a group proceeding and individual issues in separate individual
proceedings, with consolidation or reassignment sometimes happening only for the limited purpose of resolving common issues. For example, multiple cases alleging infringement of the same patent pending in the same district have been reassigned to a single judge only to
interpret the patent claims.278 Similarly, the Northern District of California assigned 200 DirecTV cases to a single judge only to resolve
the propriety of joinder, determine whether certain statutory provisions allowed private rights of action, and establish the necessary
showing under various statutes for default judgments.279
274. See supra Part V.A.3.
275. See § 1407(a); Lexecon Inc. v. Milberg Weiss Bershad Hynes & Lerach, 523 U.S.
26, 28 (1998). Common issues can be resolved via summary judgment in multi-district proceedings or tested via sample or “bellwether” trials.
276. C.D. CAL. L.R. 83-1.3.1.1; see also, e.g., S.D. CAL. L.R. 40.1(e)-(j).
277. S.D. CAL. L.R. 40.1(e)-(j).
278. See, e.g., Kohus v. Toys “R” Us, Inc., Nos. C-1-05-517, C-1-05-671, 2006 WL
1476209, at *1-2 (S.D. Ohio May 25, 2006).
279. In re DIRECTV, Inc., No. C-02-5912-JW, 2004 WL 2645971, at *1, *12 (N.D. Cal.
July 26, 2004).
1058
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
In essence, the proposed inter-district coordination applies the tools
developed for intra-district related case procedures—especially the use
of automatic reassignment and reassignment for the limited purpose of
(final) resolution of only common issues—to cases involving similarly
situated defendants that reach across district boundaries. It also formalizes the standards and procedures for related case coordination,
which in the intra-district context, often are ad hoc and subject to substantial judicial discretion, and therefore judicial self-interest.
Although developed for the needs of defendant aggregation, interdistrict related case coordination may be useful for certain types of
multi-plaintiff cases that share similarities with multi-defendant
cases, including the presence of both significant individual and significant common issues and the inappropriateness of representative
procedures. For example, mass tort cases, which fit these criteria and
have faced obstacles to aggregation under existing procedures, may
benefit from inter-district coordination. Indeed, aspects of interdistrict coordination echo the proposals from the American Law Institute’s Complex Litigation Project in the 1990s, which was largely
focused on plaintiff aggregation.280
2. Mechanics of Inter-District Coordination
The proposed inter-district coordination procedures would be initiated when the plaintiff filed a complaint (probably a form complaint) against each defendant in a chosen district court with jurisdiction and venue over that defendant. The plaintiff also would file a
notice listing any “related cases” already pending or concurrently
filed.281 Related cases could be defined generically to include, inter
alia, cases with a common, or perhaps “significant,” question of fact
or law.282 Substance-specific aggregation protocols, discussed in more
detail below, also would be useful to provide more specific definitions
or guidance as to what constitutes a related case in various substantive areas.283 To prevent serial litigation when it suited the plaintiff’s
interest, incentives could be provided to name all defendants at once,
such as a substantially discounted bulk rate filing fee, a penalty or
moratorium for subsequently filing related cases, or even preclusion
of related claims that were identifiable and could have been brought
at the time of original filing.284
280. AM. LAW INST., COMPLEX LITIGATION PROJECT 1-4 (Proposed Final Draft 1993).
281. See, e.g., Notice of Designation of Related Civil Cases Pending (D.D.C.), available
at https://www.dcd.uscourts.gov/dcd/sites/dcd/files/CO932-online.pdf.
282. See S.D. CAL. L.R. 40.1(f)-(g).
283. See D.D.C. L.R. 40.5(a) (providing different definitions of “related case[s]” for criminal, civil forfeiture, and civil cases).
284. Cf. Parchomovsky & Stein, supra note 23 (suggesting that procedural tools, like
preclusion and restitution, can lead to more efficient defense-side outcomes).
2014]
AGGREGATING DEFENDANTS
1059
The notice of related cases would be filed simultaneously with the
chosen court and a centralized body—perhaps the Judicial Panel on
Multidistrict Litigation (JPML). The centralized body would conduct
a preliminary review to insure the cases actually qualify as related
cases. If the definition of related cases is clearly specified, this task
could be assigned to the equivalent of a clerk to reduce costs.285 The
centralized body would then reassign the related cases to a single
judge for the limited purpose of resolving common issues. Under existing multi-district litigation procedures, the choice of judge is at the
discretion of the JPML and often is subject to briefing and a hearing.286 Because of its automatic application, inter-district coordination
will involve many more cases than multi-district litigation, and it will
be infeasible to litigate the identity of the coordinating judge in every
case. To reduce transaction costs, the identity of the coordinating
judge should be determined automatically287—perhaps by assigning
the related cases to a judge in the district with the most related cases, a judge in the district that is geographically closest to the most
defendants, a judge that has volunteered for coordinated cases either
generally or in specific substantive areas, a judge that has passed
some minimal screening for ability to handle complex litigation, or a
judge with some combination of these factors.
Once reassigned for coordination, the coordinating judge would
develop a coordination plan specifying the common issues, the plan
for discovery and resolution of the common issues, and the status of
the individual cases. This coordination plan would involve input from
the parties and probably a hearing, and the parties would also have
the chance to seek de-aggregation on limited, specific grounds (e.g.,
although the same patent is involved, totally different claims are asserted against each defendant). The coordinating judge could stay
proceedings in the individual cases pending outcome of the common
issues, especially where the common issues could be dispositive of all
cases (e.g., patent invalidity). In other circumstances, individual issues may be litigated simultaneously with the coordinated proceedings, such as when one patent has been asserted against many defendants but a second patent is asserted individually against just one
defendant. Once the common issues have been resolved, the related
cases would be returned to the individually chosen districts for reso285. See S.D. CAL. L.R. 40.1(g)-(h) (requiring clerk to identify related cases and prepare
an order for judge’s signature).
286. See Margaret S. Williams & Tracey E. George, Who Will Manage Complex Civil
Litigation? The Decision to Transfer and Consolidate Multidistrict Litigation, 10-11, 23-24
(July 1, 2010) (unpublished manuscript), available at http://ssrn.com/abstract=1633703.
287. Cf. S.D. CAL. L.R. 40.1(h)-(i) (assigning related cases to the judge with the “lownumbered” case). The low-number rule would not work for inter-district coordination because plaintiffs could consistently choose the most plaintiff-friendly districts simply by
filing the first of the related cases in that district.
1060
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
lution of individual issues.288 Any common issues resolved in the coordinated proceedings would have law of the case or collateral estoppel effect in the individual proceedings.289
To reduce litigation costs, minimize disputes over the coordination
plan, and constrain judicial variability and self-interest, presumptively binding guidelines could be used to specify what constitutes a
common issue appropriate for coordinated resolution, what constitutes an individual issue not appropriate for coordinated resolution,
in what circumstances to stay individual proceedings, how to structure discovery, and how to handle trials. Although trans-substantive
guidelines would help, substance-specific protocols would allow
greater specificity and detail and be preferable, at least for substantive areas that repeatedly produce multi-defendant cases (e.g., patent
litigation or internet file-sharing).290
For example, under a patent-specific protocol, a related case could
be defined as one alleging infringement of the same patent or perhaps patents that issued from a common application. The common
issues would include interpretation of the patent (i.e., “claim construction”), whether the patent is invalid for failing to meet the statutory requirements,291 defenses related to the patentee’s conduct in
the Patent Office (e.g., “inequitable conduct”), and maybe infringement issues in those cases where they are likely to be identical for
each defendant.292 As for structure, the patent-specific protocol could
provide for an early claim construction hearing to resolve the scope of
the patent,293 then discovery limited to invalidity and inequitable
conduct issues, and finally trial of invalidity and inequitable conduct.
Because the common defenses are potentially case-dispositive and
therefore can save the costs from discovery and trial on infringement,
288. Cf. In re DIRECTV, Inc., No. C-02-5912-JW, 2004 WL 2645971, at *12 (N.D. Cal.
July 26, 2004) (referring cases back to originating judge after resolving certain common
issues on a district-wide coordinated basis).
289. See Romberg, supra note 270, at 254 (similar suggestion for issue class actions).
290. Despite the trans-substantive nature of most procedural rules, see Paul D. Carrington, Making Rules to Dispose of Manifestly Unfounded Assertions: An Exorcism of the
Bogy of Non-Trans-Substantive Rules of Civil Procedure, 137 U. PA. L. REV. 2067, 2067
(1989), substance-specific procedures are increasingly common in complex areas, including
Patent Local Rules, see, e.g., N.D. CAL. PATENT L.R., available at http://www.cand.
uscourts.gov/localrules/patent, guidelines from court advisory councils on discovery and
case management, see Jason Rantanen, Model Order Addressing Numerosity of Claims and
Prior Art [UPDATED], PATENTLYO (July 26, 2013), http://www.patentlyo.com/
patent/2013/07/model-order-addressing-numerosity-of-claims-and-prior-art.html, and the
ad hoc guidelines on the showing required for a DirecTV default judgment created by the
Northern District of California, see In re DIRECTV, Inc., 2004 WL 2645971, at *6-11.
291. Parchomovsky & Stein, supra note 23, at 1484-86 (classifying patent law defenses
as individual or common).
292. See Taylor, supra note 7, at 717-19.
293. See, e.g., N.D. CAL. P ATENT L.R. 4 (setting procedures for early claim construction hearing).
2014]
AGGREGATING DEFENDANTS
1061
damages, and other individual issues—individual proceedings generally should be stayed until after the common issues are resolved,294 or
at least until claim construction is resolved.295
C. Potential Obstacles to Inter-District Coordination
A benefit of inter-district coordination is that it is a hybrid of two
well-established procedures (multi-district litigation and intra-district
related case coordination) and therefore poses minimal implementation problems. This Section considers a few implementation obstacles.
1. Practical Obstacles: Free Riding
Although common defenses are public goods and raise the possibility of free riding with or without aggregation,296 aggregation may accentuate the problem. In individual litigation, each defendant must
expend money and effort to develop its defenses, including common
defenses; if it does not, it loses.297 But aggregated defendants prevail
when the group succeeds on a common defense, regardless of their
contribution. Thus, in theory, each defendant’s incentive is to underinvest in the common defense, relying on other defendants to develop
defenses that also allow the free-riding defendant to escape liability.298
Free riding poses fewer obstacles to aggregation of similarly situated plaintiffs because successful aggregated claims create a common
pool of money from which to pay the lawyers, incentivizing lawyers to
vigorously pursue the common cause.299 Aggregated defenses create
no common fund, but rather require expenditures to avoid greater
expenditures.300 To overcome the free-riding problem for aggregated
defendants, Hamdani and Klement proposed one-way fee-shifting,
294. See Ford, supra note 156, at 119-22 (suggesting resolving invalidity first for
substantive reasons).
295. Claim construction is the single most important event in patent cases and is often
case-dispositive. See Greg Reilly, Judicial Capacities and Patent Claim Construction: An
Ordinary Reader Standard, 20 MICH. TELECOMM. & TECH. L. REV. (forthcoming 2014)
(manuscript at 24-25), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_
id=2310026.
296. See supra Part IV.B.
297. An individual defendant may be able to free ride if it is sued after another defendant has prevailed on a common defense, see Parchomovsky & Stein, supra note 23, at 1518,
or is part of a joint defense group sharing information and work product, see Erichson,
supra note 20, at 401-08. Moreover, a rational defendant in individual litigation may focus on
individual defenses to avoid creating a common good. See Ford, supra note 156, at 109-12.
298. See Hamdani & Klement, supra note 4, at 712.
299. See Judith Resnik, Money Matters: Judicial Market Interventions Creating Subsidies and Awarding Fees and Costs in Individual and Aggregate Litigation, 148 U. PA. L.
REV. 2119, 2139-40 (2000).
300. See Hamdani & Klement, supra note 4, at 715.
1062
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
where a losing plaintiff would pay the defendants’ fees.301 However,
most cases settle.302 A settlement for aggregated plaintiffs still creates a common fund to pay the lawyers, but a settlement for aggregated defendants neither creates a common fund nor results in a losing plaintiff. Hamdani and Klement would have the plaintiff pay fees
when the defendants prevail or settle,303 even though the settlement
could reflect the strength of the plaintiff’s case, not its weakness.
This could deter plaintiffs from bringing meritorious claims or reduce
settlements, increasing costs and congestion in the courts.
Free riding may not pose as great an obstacle to inter-district coordination as may be first thought. Practitioners involved in multidefendant litigation, at least in higher stakes cases like patent litigation, suggest free riding is less of a problem than “over-riding,” that
is, each defendant insisting on pursuing its preferred strategies and
being involved in or taking the lead on every aspect of the litigation.304 The incentive for defense lawyers to over-ride is clear: it maximizes their billable hour fees. Over-riding may be rational for defendants themselves when aggregation occurs via non-representative
procedures and each defendant can settle separately. If other defendants settle, a free-riding defendant would be left exposed to liability
without having developed a defense. In fact, practitioners report
“plaintiffs often offer the most favorable settlements to the defendants that are best prepared, to remove them from the case and focus
on the more vulnerable parties.”305 Even if not perfectly rational, clients may permit over-riding because of monitoring deficiencies or the
influence and self-interested advice of the lawyers. Finally, because
of loss aversion, defendants will tend to pay more in defense than
they rationally should, and because of self-serving bias, defendants
will overvalue their own strategies and abilities (or the wisdom of
their choice of counsel) and be unwilling to sacrifice their defense to
perceived “weaker” defendants or lawyers.
Even in lower stakes cases, like the DirecTV and internet filesharing cases, one or more defendants have been willing to take on
301. Id. at 715-17; see also Netto, supra note 21, at 112-16 (also supporting fee-shifting
in part to motivate attorneys to present defenses instead of settling).
302. Jonathan D. Glater, Study Finds Settling Is Better than Going to Trial, N.Y.
TIMES (Aug. 7, 2008), http://www.nytimes.com/2008/08/08/business/08law.html?_r=1&
(“The vast majority of cases do settle — from 80 to 92 percent by some estimates . . . .”).
303. Hamdani & Klement, supra note 4, at 715-17.
304. See, e.g., Irene C. Warshauer et al., Methods to Effectively Manage Complex MultiParty Disputes, 14A ROCKY MTN. MIN. L. FOUND. (SPECIAL INST. ON RESOL. & AVOIDANCE
OF DISPUTES) 3, 3-15, 20 (1984); Baghdassarian & Frankel, supra note 151.
305. Baghdassarian & Frankel, supra note 151; see, e.g., Warshauer et al., supra note
304, at 3-19.
2014]
AGGREGATING DEFENDANTS
1063
common issues on behalf of all defendants.306 Aggregation probably
makes it more likely that the common defense will be developed in
low-stakes cases because it offers greater opportunities to fund defense lawyers than individual litigation. For example, aggregation
may encourage involvement by public interest groups whose substantive interests line up with the defendants—such as the open-access
movement in intellectual property cases—by lowering the required
costs and effort and raising the stakes and prominence.307 Alternatively, defense lawyers could collect an inventory of cases, charging
each defendant a small amount but creating a respectable war
chest.308 For example, if a defense lawyer can sign up 100 clients for a
flat fee of $1000,309 the $100,000 total should allow a better common
defense than any defendant could put on in individual litigation.
Finally, free riders in aggregate litigation do not make participating defendants worse off than in individual litigation, as participating defendants would likely litigate the same common issues at virtually the same cost without aggregation.310 To the extent that free
riding is seen as unfair or creating under-investment in common defenses, mechanisms could be used to encourage or force all defendants to contribute where feasible.311
2. Statutory and Constitutional Obstacles
Much of the proposed inter-district coordination procedure can be
implemented under the JPML’s existing rulemaking authority,312 including the procedures for choosing a coordinating judge and the development of general and substance-specific protocols. However, inter-district coordination cannot be fully implemented without
306. See, e.g., Arista Records, LLC v. Does 1–11, No. 1:07–CV–2828, 2008 WL 4823160,
at *1-2 (N.D. Ohio Nov. 3, 2008) (noting that Defendant Doe #9 had entered an appearance
via counsel and filed a motion “to dismiss all Doe Defendants except Defendant Doe # 1” for
improper joinder); In re DIRECTV, Inc., No. C-02-5912-JW, 2004 WL 2645971, at *1 (N.D.
Cal. July 26, 2004) (noting that attorneys appeared on behalf of defendants at hearing on
joinder and DirecTV’s required prima facie showing).
307. Cf. Hamdani & Klement, supra note 4, at 717-18 (noting that class defense mechanisms could provide additional funding for public interest organizations).
308. See DirecTV, Inc. v. Beecher, 296 F. Supp. 2d 937, 939 (S.D. Ind. 2004) (“Many of the
defendants in the several cases in this court have banded together for a common defense.”).
309. People with minimal resources regularly pay $1000 or more to defend DUI charges. See, e.g., Lisa Ellis, What It Costs for a DUI Attorney, WHAT IT COSTS, http://business.
whatitcosts.com/dui-attorney.htm (last visited July 25, 2014).
310. See Defendant Class Actions, supra note 20, at 648. Coordination costs are unlikely to be significant if other defendants are free riding.
311. See Ford, supra note 156155, at 123 (proposing accounting action for defendants
who prevail on a common defense to recover from other potential defendants); Parchomovsky & Stein, supra note 23, at 1520-24.
312. See 28 U.S.C. § 1407(f) (2012) (“The panel may prescribe rules for the conduct of its
business not inconsistent with Acts of Congress and the Federal Rules of Civil Procedure.”).
1064
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
amendments to the multi-district litigation statute to permit issueonly coordination since the JPML has interpreted the statute as limited to coordination of entire claims.313 Likewise, statutory amendments would be necessary to implement automatic coordination procedures, as the statute currently requires notice and a hearing before
transfer,314 and to allow trial of common issues in the consolidated
proceedings, since the statutory language and Supreme Court precedent require remand at the end of pre-trial proceedings.315
Inter-district coordination does not appear to raise any novel constitutional issues. Although the coordinating forum may not otherwise have personal jurisdiction and venue over all defendants, “Congress may, consistent with the due process clause, enact legislation
authorizing the federal courts to exercise nationwide personal jurisdiction,” which it did in the multi-district litigation statute.316 Likewise, issue-only aggregation does not run afoul of the Seventh
Amendment’s guarantee that “no fact tried by a jury, shall be otherwise re-examined in any Court of the United States . . . .”317 Common
legal issues do not implicate the Seventh Amendment, and even for
common issues subject to a jury trial, the Seventh Amendment only
bars revisiting of the same issue, not simply an overlap in evidence
presented to two juries.318 If common issues are properly defined by
the coordinating judge and properly given preclusive effect in indi-
313. See supra Part V.A.3.
314. See § 1407(c).
315. See § 1407(a); Lexecon Inc. v. Milberg Weiss Bershad Hynes & Lerach, 523 U.S.
26, 40 (1998).
316. In re “Agent Orange” Prod. Liab. Litig. MDL No. 381, 818 F.2d 145, 163 (2d. Cir.
1987) (absent class members); see also In re FMC Corp. Patent Litig., 422 F. Supp. 1163,
1165 (J.P.M.L. 1976) (defendants). This is clearly true for cases subject to federal question
jurisdiction. In these cases, the Due Process Clause of the Fifth Amendment applies and
requires only minimum contacts with the United States as a whole. See 4 CHARLES ALAN
WRIGHT & ARTHUR R. MILLER, FEDERAL PRACTICE AND PROCEDURE § 1068.1, at 597-602,
612 (3d ed. 2002). By contrast, in diversity cases, the Due Process Clause of the Fourteenth
Amendment applies; allowing a federal, nationwide jurisdiction statute to trump the statespecific minimum contacts analysis required in state courts would seem to run afoul of the
Erie Doctrine. Id. § 1068.1, at 592. Yet, personal jurisdiction has not proven an obstacle to
multi-district litigation, even in diversity cases. See Agent Orange, 818 F.2d at 152, 163.
For aggregated defendants, personal jurisdiction may prove more problematic in diversity
cases because of the heightened due process concerns for defendants as compared to plaintiffs. See Phillips Petroleum Co. v. Shutts, 472 U.S. 797, 810-11 (1985). At the very least,
personal jurisdiction issues are likely to receive more attention than they have received in
multi-district litigation if inter-district related case coordination is used in diversity cases.
317. U.S. CONST. amend. VII.
318. See Romberg, supra note 270, at 324-25. For a detailed consideration of the Seventh Amendment issue in the context of issue class actions, see Patrick Woolley, Mass Tort
Litigation and the Seventh Amendment Reexamination Clause, 83 IOWA L. REV. 499 (1998).
2014]
AGGREGATING DEFENDANTS
1065
vidual litigation, there is no risk that the individual jury would reconsider issues resolved by the coordinated jury.319
VI. CONCLUSION
This Article has explored the overlooked puzzle of defendant aggregation: the opposition of defendants, courts, and policymakers to
aggregation of similarly situated defendants despite the theoretical
benefits it offers courts and the judicial system. Defendants’ opposition is more easily explained, at least for some defendants and some
types of cases. Yet, these explanations are weaker and less broadly
applicable than may be commonly assumed. Systemic opposition likely
results from some combination of improper merits or policy determinations, capture by defendant interests, and a narrow focus on a judge’s
own self-interest rather than that of the judicial system as a whole.
The need for aggregation of similarly situated defendants to date
has largely arisen in technology cases—patent cases, internet
file-sharing copyright litigation, and satellite television piracy—but
is likely to spread to other areas in the coming years as a result
of mass communications that increase the ability of dispersed people
to injure a single plaintiff.320 Others have noted the parallels between
the rise of mass communications and defendant aggregation in the
21st century and the rise of mass production and a national economy
and plaintiff aggregation in the mid-20th century.321 Overlooked,
however, is that the increased demand for plaintiff aggregation in
the 1950s and 1960s sparked procedural innovation that resulted in
new multi-district litigation procedures322 and substantially revised
class action procedures.323 The modern demand for defendant aggregation similarly necessitates procedural innovation that will overcome the obstacles created by defendant and judicial self-interest and
cognitive limitations.
This Article takes a first step in that direction and proposes an
inter-district related case coordination procedure. The debate going
forward should be about how to better tailor a procedure for the
needs of defendant aggregation, not whether, or how well, aggregation of similarly situated defendants fits into the boxes of existing
procedures largely crafted for aggregating plaintiffs.
319. See Romberg, supra note 270, at 324-25 (reaching a similar conclusion in the context of issue class actions).
320. See Hamdani & Klement, supra note 4, at 741; Netto, supra note 21, at 61-62.
321. See Hamdani & Klement, supra note 4, at 741; Netto, supra note 21, at 61.
322. See Resnik, supra note 24, at 30-33 (describing the history of the 1968 enactment
of the multi-district litigation statute).
323. See id. at 9-17 (describing the history of the 1966 revisions to class action procedures).
1066
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1011
IMPAIRED PHYSICIANS AND THE SCOPE OF INFORMED
CONSENT: BALANCING PATIENT SAFETY WITH
PHYSICIAN PRIVACY
SARAH HASTON
I.
II.
III.
IV.
V.
VI.
INTRODUCTION ..................................................................................................
INFORMED CONSENT DOCTRINE ........................................................................
IMPAIRED PHYSICIANS ......................................................................................
IMPAIRED PHYSICIANS AND THE SCOPE OF INFORMED CONSENT ......................
DUTIES OF THE HOSPITAL: A CORPORATE NEGLIGENCE THEORY .....................
CONCLUSION .....................................................................................................
1125
1127
1132
1134
1139
1144
I. INTRODUCTION
A thirty-eight-year-old orthopedic surgeon comes home to his empty house following a typical ten-hour day at the hospital. Recently
divorced and without children, he drops his keys at the front door
and heads towards the living room with the take-out dinner he
picked up on his way home. He turns on the news and pours himself
a glass of red wine. Ever since his divorce, he has been drinking wine
more frequently than he used to, perhaps to mask his loneliness. After his third or fourth glass of wine, he turns off the television,
brushes his teeth, and goes to bed.
At five o’clock in the morning, he awakes to his alarm clock and
throws on his running shoes. After his usual five-mile run, he grabs a
piece of toast and heads to the hospital. Today is just like any other
day for the surgeon; at eleven o’clock, he reviews the chart of a patient scheduled for a total knee-replacement. The surgeon goes into
the operating room to discuss with the patient, one more time, the
risks and alternatives associated with this surgery, along with its
potential benefits. He asks the patient whether she has any further
questions before anesthesia is administered. After the patient signs
the standard informed consent form and acknowledges that she
wishes to proceed with the surgery, the orthopedic surgeon leaves the
room to prepare for the procedure.
Shortly after the knee-replacement surgery, the patient begins
developing complications, ultimately requiring an additional surgery.
Frustrated with the outcome of the original procedure, the patient
chooses to have a different physician perform the second surgery. Despite all efforts during the second surgery, the patient loses full function of her leg.
J.D. Candidate 2015, Florida State University College of Law. I would like to thank
my parents, Laurie and Jim Haston, for their love and support throughout all of my educational endeavors. I would also like to thank the incredible faculty at the Florida State University College of Law for their support, feedback, and encouragement.
1126
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1125
The patient begins hearing rumors about the orthopedic surgeon’s
recent, nasty divorce. She also sees him at local restaurants, sitting
alone at the bar with a drink in front of him. After the patient calls
some mutual friends of the surgeon’s ex-wife, she learns that the orthopedic surgeon has developed a drinking problem since his divorce.
She decides to file a medical malpractice suit, alleging that the surgeon did not obtain her informed consent before the procedure because he failed to inform her of his alcoholism.
Situations like the one described above present an important and
undefined question to the medical field: whether a physician has a
duty to inform his patient about his alcoholism, drug abuse, and/or
mental illness (such as depression) in gaining a patient’s informed
consent before a procedure.1 On one hand, we live in a country with a
medical system that values patient autonomy.2 We believe a patient
has the right to make her own medical decisions based on a physician’s adequate disclosure of the potential risks and benefits involved
in a given procedure.3 The large amount of medical malpractice suits
and medical regulations also demonstrate the importance we place on
patient safety.4 On the other hand, however, we also recognize the
shortcomings of the informed consent doctrine.5 We recognize an individual’s right to privacy6 and that alcoholism, mental illness, and
drug-related impairments are private matters deserving protection
from disclosure to the public to effect recovery.7
This Note will argue that a physician, in gaining a patient’s informed consent, does not have a duty to disclose to the patient
whether he suffers, or has suffered, from alcoholism, drug abuse, or
mental illness. Physicians are still human and thus deserve the right
to privacy regarding such personal matters as are enjoyed by the
1. See Mary Anne Bobinski, Autonomy and Privacy: Protecting Patients from Their
Physicians, 55 U. PITT. L. REV. 291, 376 (1994) (“The types of personal characteristics . . .
requiring disclosure are undefined.”).
2. See Canterbury v. Spence, 464 F.2d 772, 780 (D.C. Cir. 1972) (“The root premise is
the concept, fundamental in American jurisprudence, that ‘[e]very human being of adult
years and sound mind has a right to determine what shall be done with his own body.’ ”
(quoting Schloendorff v. Soc’y of N.Y. Hosp., 105 N.E. 92, 93 (1914))).
3. See id. at 781.
4. See, e.g., Frederick Levy et al., The Patient Safety and Quality Improvement Act of
2005: Preventing Error and Promoting Patient Safety, 31 J. LEGAL MED. 398-406 (2010).
5. See Sheldon F. Kurtz, The Law of Informed Consent: From “Doctor is Right” to
“Patient has Rights”, 50 SYRACUSE L. REV. 1243, 1245 (2000).
6. See Barry R. Furrow, Data Mining and Substandard Medical Practice: The Difference Between Privacy, Secrets and Hidden Defects, 51 VILL. L. REV. 803, 803 (2006) (“We
properly value privacy: it is a desirable end state and a precondition for identity, allowing
individuals to achieve goals such as autonomy and solidarity with peers . . . .”).
7. See Rebecca Sara Feinberg, The Impaired Physician: Medical, Legal, and Ethical
Analysis with a Policy Recommendation, 34 NOVA L. REV. 595, 605 (2010) (“Confidentiality
is [a] key element to successful intervention and treatment.”).
2014]
SCOPE OF INFORMED CONSENT
1127
public generally. This Note will point out that such conditions do not
have as material of an effect on the performance of medical procedures as people generally assume. Yet, if these impairments were to
rise to a level such that patient safety would be compromised, liability must fall somewhere. Thus, instead of requiring physicians to disclose their personal impairments during the informed consent process, the emphasis should be placed on the hospital’s duty to properly
monitor its physicians’ credentials. Further, disclosure through the
informed consent doctrine is not the proper way to protect patients
from physicians who are incapable of adequately performing medical
procedures due to alcoholism, drug abuse, or mental illnesses. Rather
than pointing the finger at the physician for lack of disclosure, the
issue of impaired physicians involves a bigger picture that the hospital system is, itself, deficient.
Hospitals should implement methods that will incentivize physicians to get the treatment they need without the attached stigmatization. Hospitals must implement and properly maintain confidential
yet effective programs designed to identify and rehabilitate physicians suffering from alcoholism, drug-dependency, and mental illnesses to ensure that those physicians practicing medicine are safely
capable of doing so. This emphasis on a hospital’s duty to monitor
and qualify physicians provides the ideal balance of physician privacy
and patient safety.
This Note will begin by exploring the history and evolution of the
informed consent doctrine in the American medical field. A discussion of the characteristics of “impaired physicians” will follow. This
Note will then examine the relationship between impaired physicians
and the scope of the informed consent doctrine. Finally, I will articulate that the duties of a hospital to monitor and credential physicians
via a corporate negligence theory strikes the proper balance between
protecting a physician’s privacy and patient’s safety.
II. INFORMED CONSENT DOCTRINE
The informed consent doctrine is a relatively recent phenomenon
principled on the American concept of autonomy.8 Informed consent
was seen as one tool in response to the imbalance of power in
the provider-patient relationship.9 The idea was that “requiring physicians to provide more information to their patients [would] help to
8. See Barbara L. Atwell, The Modern Age of Informed Consent, 40 U. RICH. L. REV.
591, 594 (2006).
9. See MARK A. HALL ET AL., MEDICAL LIABILITY AND TREATMENT RELATIONSHIPS
197-98 (Vicki Been et al. eds., 3d ed. 2013).
1128
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1125
redress the power imbalance problems created by the inequality
of knowledge.”10
The doctrine was originally an outgrowth of the tort of battery.11
Claims for an informed consent violation amounted to an intentional
and unauthorized touching of another. 12 Beginning in the 1950s,
courts “viewed risk nondisclosure situations as analogous to and an
extension of those battery cases where consent was fraudulently obtained.”13 Bringing an informed consent claim under a battery theory
had significant advantages over bringing a negligence-based claim:
First, and most importantly, battery did not require the patient to
prove by an expert medical witness that the defendant had deviated from an accepted medical standard of care. Second, battery did
not require the patient to prove that he would have refused the
treatment if he had been given the proper information. Third, battery would entitle the patient to recover damages even if the operation did not have a “bad result.” Fourth, as an intentional tort,
battery might have entitled the patient to an instruction on punitive damages, which would not be available in an action based upon negligence.14
Recognizing the possibility for unnecessary liability, courts gradually began pulling away from the battery theory of informed consent
and, instead, started moving toward a standard negligence theory of
medical malpractice.15 The battery-based claims generally dealt more
with procedures in which no consent was given, while the new negligence-based claims more commonly involved situations in which the
consent given was inadequate.16 Indeed, basing an informed consent
claim on a theory of negligence “more closely comports with the reality of medical practice.” 17 “[D]octors’ informed consent practices require the exercise of judgment. . . . Since the treating physician retains some flexibility in determining how much information to disclose, cases based on a failure to obtain informed consent, like other
medical malpractice claims, are based on negligence.”18
10. Id. at 198.
11. William J. McNichols, Informed Consent Liability in a “Material Information”
Jurisdiction: What Does the Future Portend?, 48 OKLA. L. REV. 711, 714-16 (1995).
12. See Atwell, supra note 8, at 593 (“[The informed consent doctrine] is a natural
outgrowth of the common law tort of battery that prohibits intentional unauthorized
bodily contact.”).
13. McNichols, supra note 11, at 714.
14. Id. at 715.
15. See id.
16. See Laurel R. Hanson, Note, Informed Consent and the Scope of a Physician’s Duty
of Disclosure, 77 N.D. L. REV. 71, 72, 75 (2001).
17. Atwell, supra note 8, at 595.
18. Id. at 595-96.
2014]
SCOPE OF INFORMED CONSENT
1129
Autonomy is a fundamental American concept that is furthered by
the informed consent doctrine.19 “Competent adults exercise that autonomy by deciding whether or not to consent to medical treatment.” 20 Judge Cardozo, in one of the pioneer cases involving informed consent, famously stated: “Every human being of adult years
and sound mind has a right to determine what shall be done with
his own body.” 21 Thus, the informed consent doctrine ensures that
patients “have the material information with which to make an
informed choice.”22
There are two main standards that courts apply to determine a
physician’s required level of disclosure: the “reasonable patient” 23
standard and the “professional malpractice”24 standard.25 Under the
reasonable patient standard, a physician’s required level of disclosure
is that which a reasonable patient would consider “material” in making a medical treatment decision.26 Some courts have argued that determining the materiality of a risk from a reasonable patient’s point
of view is the best way to respect autonomy.27
The professional malpractice standard defines the scope of disclosure as the information which a reasonable physician in similar circumstances would provide to a patient. 28 Today, about half of the
states follow this standard.29 The professional malpractice standard
19. See id. at 594-95.
20. Id. at 594.
21. Schloendorff v. Soc’y of N.Y. Hosp., 105 N.E. 92, 93 (1914), abrogated by Bing v.
Thunig, 142 N.E.2d 3, 7-9 (1957), superseded by statute, Act of May 21, 1975, ch. 109, 1975
N.Y. Laws 134-35.
22. See Atwell, supra note 8, at 596.
23. This is also known as the “material risk” or “patient-centered” standard.
24. This is also known as the “professional disclosure” or “reasonable physician” standard.
25. See HALL ET AL., supra note 9, at 205; Hanson, supra note 16, at 75.
26. See HALL ET AL., supra note 9, at 205; see also Hanson, supra note 16, at 75.
27. See, e.g., Canterbury v. Spence, 464 F.2d 772, 786 (D.C. Cir. 1972) (“[T]he patient’s
right of self-decision shapes the boundaries of the duty to reveal. . . . The scope of the physician’s communications to the patient, then, must be measured by the patient’s need, and
that need is the information material to the decision.” (emphasis added)); see also Kurtz,
supra note 5, at 1251 (“The Canterbury patient-based standard seems more consistent with
the purpose of 'informed consent' which includes a respect for a patient’s autonomy interest
and a desire to have patients participate in the decision making process.”). “A small number of jurisdictions take an even more protective approach, requiring disclosure of information that a particular patient (as contrasted with a ‘rational patient’) would have wanted to make his or her decision.” HALL ET AL., supra note 9, at 205.
28. HALL ET AL., supra note 9, at 205; see also Hanson, supra note 16, at 75.
29. HALL ET AL., supra note 9, at 214. “In adopting the reasonable physician standard,
courts seemed to be assuming that a competent physician acting in good faith would convey
to patients information essential to their understanding of relevant risks, benefits, and
alternatives, rather than simply provide the traditionally meager disclosures . . . .” Ben A.
Rich, Medical Paternalism v. Respect for Patient Autonomy: The More Things Change the
More They Remain the Same, 10 MICH. ST. U. J. MED. & L. 87, 106-07 (2006).
1130
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1125
does not require physicians to speculate as to what the reasonable
patient would want to know; instead, it applies the appropriate medical standard of care to utilize the physician’s professional judgment.30
Testimony from medical experts is necessary, however, because the
professional malpractice standard measures disclosure according to
the appropriate physician’s standard of care.31
Under either standard of informed consent, various situations
may excuse disclosure, 32 including emergencies, 33 areas of common
knowledge, 34 areas of individualized knowledge, 35 situations where
disclosure would “ ‘foreclose rational decision’ or ‘pose psychological
damage’ to the patient,”36 and occasions when a patient has waived
disclosure.37 Such exceptions to the doctrine are generally set up as
affirmative defenses.38
Regardless of which standard courts apply, the informed consent
doctrine has received significant criticism.39 Many studies suggest a
dichotomy between disclosure and comprehension or retention.40 As
such, even when physicians disclosed sufficient information, “few patients understood or remembered what they had been told about
their medical condition and treatment options.”41 Additionally, there
is a valid concern that excessive disclosure, often due to a fear of litigation, will result in patients foregoing medically necessary treat30. See Culbertson v. Mernitz, 602 N.E.2d 98, 103 (Ind. 1992) (“[A physician] should
not be called upon to be a ‘mind reader’ with the ability to peer into the brain of a prudent
patient to determine what such patient ‘needs to know,’ but should simply be called upon
the discuss medical facts and recommendations with the patient as a reasonably prudent
physician would.”).
31. HALL ET AL., supra note 9, at 215; see also Hanson, supra note 25, at 75 (“Under this
standard, a patient usually must present expert testimony to demonstrate that a physician,
following acceptable medical practice, would have disclosed the injury-causing risk.”).
32. See Hanson, supra note 25, at 73-75.
33. HALL ET AL., supra note 9, at 221 (“There is no duty to disclose information in an
emergency situation where the patient is not competent, immediate treatment is required
to prevent more serious harm, and no substitute decisionmaker is available.”).
34. Id. at 221 (“There is no duty to disclose risks ‘of which persons of average sophistication are aware.’ ”).
35. Id. (“The patient cannot recover for the physician’s failure to disclose a risk already known by the patient.”).
36. Id. at 221. This is known as the “therapeutic privilege,” which courts have applied
only in a few circumstances. Id. at 221-22.
37. Id. at 222.
38. See id. (“The defendant generally has the burden of proving that an exception to
the duty to inform is present.”).
39. See, e.g., Atwell, supra note 8, at 597-98 (“In practice, informed consent is far from
perfect.”); Kurtz, supra note 5, at 1245 (“[T]he doctrine falls short of its intended purpose.”).
40. See also, e.g., David A. Herz et al., Informed Consent: Is It a Myth?, 30
NEUROSURGERY 453 (1992); Alan Meisel & Loren H. Roth, Toward an Informed Discussion of
Informed Consent: A Review and Critique of Empirical Studies, 25 ARIZ. L. REV. 265 (1983).
41. HALL ET AL., supra note 9, at 202.
2014]
SCOPE OF INFORMED CONSENT
1131
ments.42 Other scholars suggest that the doctrine of informed consent
doctrine does not fulfill its “goals of patient protection” in the modern
practice of medicine.43 Further, the elements needed to succeed in a
cause of action for lack of informed consent make recovery difficult;
thus, many claims are never litigated.44
In bringing an informed consent claim, the plaintiff will generally
have to prove the following: “(1) that the medical procedure carried a
specific risk that was not disclosed, (2) that the physician violated the
applicable standard of disclosure, (3) that the undisclosed risk materialized, and (4) that the failure to disclose the information caused
the patient’s injury.”45
Proving materialization of the undisclosed risk and causation are
often the most difficult barriers for plaintiffs. 46 One can imagine
many circumstances where physicians might fail to disclose a risk
without the risk ever materializing. For example, if a physician fails
to disclose to a patient that a particular surgery could result in paralysis (arguably a very material risk), and paralysis never materializes, the patient would not have a valid claim against the physician.
When such a situation occurs, it seems the informed consent doctrine
serves the purpose of protecting patients from injury rather than its
stated purpose of respecting patient autonomy.47 To prove causation,
a plaintiff must convince the court that a reasonable patient in similar circumstances would have refused the treatment had the disclosure been sufficient.48 This requirement is often a very difficult burden of proof to satisfy and tends to raise a range of issues.49 While the
causation requirement is an objective standard, the jury is permitted
to consider the plaintiff’s particular circumstances.50 The law is still
fuzzy and inconsistent regarding how much of a plaintiff’s subjective
circumstances should be considered.51
42. See Barry R. Furrow, Doctors’ Dirty Little Secrets: The Dark Side of Medical Privacy, 37 WASHBURN L.J. 283, 293 (1998) (“Disclosure creates the risk that a patient will
refuse the physician’s care because of the status or the addiction, rather than looking at
the particular case and the risks posed.”). Ironically, this trend would actually compromise
patient safety.
43. Atwell, supra note 8, at 611.
44. See, e.g., HALL ET AL., supra note 9, at 217.
45. Id.
46. See id.
47. “Vindicating patient autonomy is at the heart of all inadequate consent cases,
whether the theory is based on battery or negligence.” McNichols, supra note 11, at 715.
48. HALL ET AL., supra note 9, at 217.
49. See id. at 217-18.
50. See id. (“[A] fact finder applying objective causation rules may take into account characteristics of plaintiff, including ‘idiosyncrasies, fears, age, medical condition, and religious
beliefs.’ ” (quoting Ashe v. Radiation Oncology Assocs., 9 S.W.3d 119, 123-24 (Tenn. 1999))).
51. See id.
1132
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1125
III. IMPAIRED PHYSICIANS
“As a result of their status [doctors] are expected to be above human
failings, while proving immune to the ailments that afflict the
general population. Nonetheless, to err is human and in being
human doctors inevitably make mistakes in the normal course of
providing care.”52
An “impaired physician is a medical doctor who suffers from alcoholism, drug addiction, or mental illness.”53 The American Medical
Association (AMA) uses the phrase “impaired physician” as a term of
art and defines it as “ ‘the inability to practice medicine adequately
by reason of physical or mental illness, including alcoholism or drug
dependency.’ ”54 The AMA’s designation of a physician as being impaired can potentially result in the suspension or revocation of the
physician’s medical license.55
“Given the everyday work stress that medical professionals experience, it is not surprising that many physicians may turn to alcohol
or substance abuse.”56 The rate of impairment for physicians is consistent among various specialties, regions, and age ranges; “[n]o
group of physicians are immune.”57 The problem is real and the numbers are frightening. The Medical Board of California estimates that
eighteen percent of the physicians in California abuse alcohol or other drugs at some point during their careers.58 The sad fact remains:
physicians have a higher rate of impairment than non-physicians, yet
they have more trouble securing treatment. 59 The fear that physicians will receive some negative stigma from their peers, along with
the fear of potential punishment, often causes physicians to conceal
their alcoholism, substance abuse, or mental illness.60 This pattern
52. A. Mavroforou et al., Alcohol and Drug Abuse Among Doctors, 25 MED. & L. 611,
612 (2006).
53. Feinberg, supra note 7, at 598.
54. Id.
55. Id. at 601, 613. For purposes of this note, the phrase “impaired physician” shall
refer to the concept more generally (that of a physician suffering from alcoholism, drug
abuse, or mental illness) rather than the term of art, as used by the AMA.
56. Ila S. Rothschild, Law for Physicians: An Overview of Medical Legal Issues, 21 J.
LEGAL MED. 437, 440 (2000) (book review).
57. Roger S. Cicala, Substance Abuse Among Physicians: What You Need to Know,
HOSPITAL PHYSICIAN, July 2003, at 39.
58. See Lucian L. Leape & John A. Fromson, Problem Doctors: Is There a System-Level
Solution?, 144 ANNALS INTERNAL MED. 107, 108 (2006).
59. See Feinberg, supra note 7, at 599 (“Physicians have both a higher prevalence of
impairment and more difficulty obtaining treatment than non-physicians.”).
60. Id. (“[T]he impaired physician often conceals his addiction because the stigma
attached to physician impairment makes seeking help significantly more difficult
than for the general population. Potential punitive responses may also play a role in
incenting concealment.”).
2014]
SCOPE OF INFORMED CONSENT
1133
means that physicians have a much lower likelihood of receiving the
treatment they need.61
Alcoholism and substance abuse have been so stigmatized among
medical professionals that discussion of the topic is taboo.62 Despite
the high incidences of alcoholism and drug abuse among physicians,
little talk is present at professional meetings, and coverage in medical school curricula is sparse.63
Unsurprisingly, there is a good deal of literature arguing that
physician impairment has a significant, negative impact on a physician’s medical performance.64 One study reveals that physicians who
reported making a major medical mistake within the last three
months were more likely to suffer from depression or struggle from
alcohol or drug dependence.65 However, as one professor at the University of Tennessee College of Medicine also suggests: “Most physician substance abusers continue to function quite well until the problem is far advanced.”66 Further, “[a]lcoholics can often remain sober
during working hours for many years, even though they drink large
quantities at night and on weekends.”67 Other medical literature often compares alcoholism in physicians to sleep-deprivation. Indeed,
studies have revealed that physicians who have not slept in twentyfour hours (as is not entirely uncommon) have impaired “cognitive
psychomotor performance to the same degree as having a 0.1% blood
alcohol level.”68 Yet, there is a lack of empirical evidence demonstrating a correlation between sleep deprivation and medical errors.69 This
61. See id. at 600.
62. See Cicala, supra note 57, at 39; see also Feinberg, supra note 7, at 597 (“Physician
addiction is taboo, but this silence injures both the physician and his patients.”).
63. Cicala, supra note 57, at 39.
64. See, e.g., Bobinski, supra note 1, at 298-99 (“The abuse of drugs and alcohol is in
turn associated with higher risks of work-related performance deficits. Physicians with
substance abuse problems may present several different sorts of risks to their patients.”);
Feinberg, supra note 7, at 597 (“Alcohol and drug addiction interfere with multiple functions in daily life. There is no doubt, when the addicted individuals are physicians, the
interference affects their ability to practice medicine, causing their patients to receive a
lower standard of care than they would otherwise receive.”).
65. Looking at Alcohol Use Disorders Among Surgeons, PHYSICIAN’S WEEKLY (Aug.
23, 2012), http://www.physiciansweekly.com/surgeons-alcohol-use-disorders/ [hereinafter
Alcohol Use]. However, “the study did not identify the nature of the errors or
determine the cause and effect.” Carolyne Krupa, 15% of Surgeons Struggle with Alcohol
Problems, AM. MED. NEWS (Mar. 6, 2012), http://www.amednews.com/article/20120306/
profession/303069998/8/.
66. Cicala, supra note 57, at 39.
67. Id.
68. Merit Buckley, Imposing Liability in the United States Medical Residency
Program: Exhaustion, Errors, and Economic Dependence, 12 DEPAUL J. HEALTH CARE L.
305, 313 (2009).
69. See id. at 314.
1134
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1125
logically seems to suggest that a similar correlation between alcoholism and medical errors would also be lacking.70
Regardless, the fact remains that there is a higher percentage of
physicians suffering from alcoholism or drug abuse than there is in
the general population, and something must be done to address this.
IV. IMPAIRED PHYSICIANS AND THE SCOPE OF INFORMED CONSENT
Under either a professional malpractice standard or a reasonable
patient standard, the exact scope of required disclosure remains gray.
It is clear that informed consent requires, at a baseline, the benefits
and risks of a particular medical procedure and potential medical alternatives.71 But even at this minimal level, how much does a physician have to disclose about the benefits, and how significant must a
risk be to warrant disclosure? The reasonable patient standard
attempts to answer this question by applying the materiality concept:
whether the information would be material to the reasonable
patient’s decision about the treatment. 72 Yet, this also leaves little
guidance. The professional malpractice standard judges the scope
of disclosure based on the applicable standard of care in the particular circumstances, but this often and simply yields a battle of
expert witnesses.73
The bottom line is that the exact scope of required disclosure, under either standard, remains unknown.74 Given this level of uncertainty, coupled with increased medical litigation and attention to
patient autonomy,75 it is not surprising that the trend in the informed
consent doctrine seems to be moving toward expanding the scope
of disclosure.
Beyond just the disclosure of risks, benefits, and alternatives to
the medical treatment, recent attempts to raise this “floor” of disclosure have been made.76 Attention has been drawn towards disclosure
of a medical treatment’s economic implications and a physician’s per-
70. See Alcohol Use, supra note 65 (“Although actual injury to patients from impaired
physicians is incredibly rare, alcohol abuse and dependence are important factors to consider when thinking about patient safety in surgery . . . .”).
71. See Bobinski, supra note 1, at 293 (“In the past, both patients and courts have
focused attention on the benefits and risks of a particular treatment . . . .”).
72. See Hanson, supra note 16, at 75.
73. See Atwell, supra note 8, at 596.
74. See Bobinski, supra note 1, at 342-43.
75. I say this hesitantly. See supra Part II for a critique of the informed consent doctrine.
76. See Kurtz, supra note 5, at 1255-56 (“In recent years, the preoccupation of the
informed consent rules with disclosure of risks inherent in a proposed treatment has expanded to the disclosure of risks of not having a proposed treatment and risks associated
with having treatment by a particular physician.”).
2014]
SCOPE OF INFORMED CONSENT
1135
sonal characteristics. 77 The personal characteristics of physicians
that have been discussed as potentially warranting disclosure before
medical procedures include a physician’s surgical experience,78 HIV
status, 79 deterioration of skills, 80 conflicts of interest, 81 and impairments.82 Though the majority rule does not require physicians to disclose personal characteristics, a growing minority of states are requiring such disclosures.83
Although the scope of disclosure obligations continues to expand,
the line must be drawn somewhere. Requiring physicians to disclose
to patients that they have suffered from, or are currently suffering
from, alcoholism, drug abuse, or mental illness, opens the door to
a flood of other potential disclosure obligations. Sure, one could argue
that alcoholism, drug abuse, or mental illness has the potential
to negatively affect a physician’s medical performance, and thus
should be disclosed.84 But if we are willing to say that these personal
impairments require disclosure, where do we end? Conceivably,
almost every personal attribute of a physician might be relevant
77. See id. at 1257-58.
78. See Johnson v. Kokemoor, 545 N.W.2d 495, 498, 506-07 (Wis. 1996) (proposing
that a physician’s lack of experience with intricate and difficult procedures may be material to a patient’s decision to proceed with the particular physician, rather than the actual
procedure). However, the Johnson principle of disclosure has not been widely expanded, as
the case is seen as unique and fact-specific. See Barry R. Furrow, Patient Safety and the
Fiduciary Hospital: Sharpening Judicial Remedies, 1 DREXEL L. REV. 439, 454 (2009); see
also Emmanuel O. Iheukwumere, Doctor, Are You Experienced? The Relevance of Disclosure of Physician Experience to a Valid Informed Consent, 18 J. CONTEMP. HEALTH L. &
POL’Y 373, 402-07 (2002).
79. See Iheukwumere, supra note 78, at 396.
80. See Semeraro v. Connolly, Civ. A. No. 92-4636, 1992 WL 392621, at *1 (E.D. Pa.
Dec. 14, 1992).
81. Compare Furrow, supra note 78, at 452 (“Courts have been less willing to impose
an obligation on physicians to disclose putative economic conflicts of interest.”), with Moore
v. Regents of the Univ. of Cal., 793 P.2d 479, 485 (Cal. 1990) (“[A] physician who is seeking
a patient’s consent for a medical procedure must, in order to satisfy his fiduciary duty and
to obtain the patient’s informed consent, disclose personal interests unrelated to the patient’s health, whether research or economic, that may affect his medical judgment.”).
82. See, e.g., Hidding v. Williams, 578 So. 2d 1192, 1196 (La. Ct. App. 1991)
83. See, e.g., id. (holding that physician’s failure to inform patient of physician’s
chronic alcohol abuse violated obligations of informed consent disclosure); Gaston v.
Hunter, 588 P.2d 326, 351 (Ariz. Ct. App. 1978) (stating that a physician’s surgical experience should be disclosed to a patient during the informed consent process).
84. See Hidding, 578 So. 2d at 1196 (holding that physician’s chronic alcohol abuse
“create[d] a material risk associated with the surgeon’s ability to perform, which if disclosed would have obliged the patient to have elected another course of treatment . . . .”);
see also Iheukwumere, supra note 78, at 402 (“Certainly, the fact of a physician’s alcoholism would be of significant concern to a reasonable patient, since an alcohol impaired physician, or one likely to be impaired during an invasive procedure would impact on the likelihood of a mistake during the procedure.”).
1136
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1125
to a patient in making a decision on whether to proceed with a
medical intervention.85
For example, consider a physician who skips breakfast the morning of a surgery. Should he have to inform his patient of this personal
detail? If a physician skips breakfast, he will certainly become hungry (if he is not already). Surely, hunger could potentially have a
negative effect on a physician’s medical performance.
Even more extreme yet under the same logic, consider a physician
who gets into a heated argument with his wife the night before performing surgery and is now in a bad mood. Should he be required to
inform his patient of this in gaining consent before the surgery? One
could also see how a bad mood might negatively affect a physician’s
medical performance.
And finally, consider a physician who has a ritual of listening to a
certain song before every surgery for good luck. He has listened to
this song before every successful surgery for the past fifteen years of
his practice. The morning of his next surgery, however, he is in a
rush and does not get a chance to listen to his “lucky song.” Cognizant that he skipped this important step, he becomes concerned.
Should the physician have to disclose this information to his patient?
As outrageous as the scenario might sound, requiring disclosure of
this idiosyncrasy logically follows from the current, expanding scope
of the informed consent doctrine.
As demonstrated above, expanding the disclosure obligations to
include a physician’s alcoholism, drug abuse, or mental illness has
the potential to create unintended bounds.86 The process of obtaining
informed consent from a patient will likely transform from an explanation of the potential risks associated with a procedure to a laundrylist-reading of the physician’s personal life.87 In fact, for this exact
reason, the Supreme Court of Pennsylvania refused to expand the
doctrine of informed consent to include mandatory disclosure of a
physician’s alcoholism.88 As explained in Kaskie v. Wright, “[t]o do so,
where the absent information consists of facts personal to the treating physician, extends the doctrine into realms well beyond its original boundaries.”89 Such an expansion of the doctrine will likely make
85. See Kaskie v. Wright, 589 A.2d 213, 217 (Pa. Super. Ct. 1991) (affirming summary
judgment for defendant physician in an informed consent claim, rhetorically asking: “Are
patients to be informed of every fact which might conceivably affect performance in the
surgical suite?”).
86. See id.
87. See Bobinski, supra note 1, at 295 (“A seemingly infinite number of provider characteristics combine to determine the physician’s ability to deliver appropriate care to a
particular patient at any given moment.”).
88. Kaskie, 589 A.2d at 217.
89. Id.
2014]
SCOPE OF INFORMED CONSENT
1137
any attempt at line-drawing very difficult,90 and patients will likely
become so overwhelmed with information to the point that they decline many important medical procedures.
Though physicians are undoubtedly held to higher expectations
than the general public and are bound by fiduciary obligations to
their patients,91 it is easy to forget that physicians are still human.
Physicians make mistakes and have struggles in their personal lives
just like everyone else. Because alcoholism, drug abuse, and mental
illnesses are such personal and sensitive issues, physicians will have
even more incentive to conceal or deny any such problems if they
know they will have to share these details with their patients. The
result will be even more physicians who are unable to seek the
treatment they need due to conscious denial or concealment.92 Further, because physicians are even more likely to keep their impairments invisible, as has been indicated, requiring physicians to disclose such impairments to patients in the informed consent process
will likely be a fruitless expansion of the doctrine.
This is not to say that a patient who receives substandard care
from a physician who suffered from alcoholism, drug abuse, or mental illness is without legal recourse against that physician.93 However, such recourse does not create a separate legal issue under the informed consent doctrine.94 Legal remediation for substandard care is
still an option for patients who are injured by their physicians’ negligence.95 As the Court of Appeals of Arizona properly held as a matter
of law in Ornelas v. Fry:
[T]he fact that [the physician] may have been an alcoholic at the
time of the surgery on [the patient] does not create in and of itself
a separate issue or claim of negligence. It is only when that alcoholism translates into conduct falling below the applicable standard of care that it has any relevance.96
90. See id. (explaining that expansion of the informed consent doctrine to include disclosure of a physician’s alcoholism will make limitations not easily definable).
91. See Furrow, supra note 78, at 444.
92. See Feinberg, supra note 7, at 598 (“It is [the] risk to patients that makes physician impairment incompatible with the practice of medicine. And it is this incompatibility
that leads the impaired physicians to further conceal their addiction . . . .”).
93. See Duttry v. Patterson, 771 A.2d. 1255, 1259 (Pa. 2001) (“[I]nformation personal
to the physician, whether solicited by the patient or not, is irrelevant to the doctrine of
informed consent. Our holding should not, however, be read to stand for the proposition
that a physician who misleads a patient is immune from suit. . . . [We do not] see a need to
expand this doctrine into a catchall theory of recovery since other causes of action provide
avenues for redress to the injured patient.”).
94. See Ornelas v. Fry, 727 P.2d 819, 823 (Ariz. Ct. App. 1986).
95. See Duttry, 771 A.2d at 1259 (“For example, it is conceivable that a physician’s lack
of experience in performing an operation would support a plaintiff’s case in negligence.”).
96. Ornelas, 727 P.2d at 823.
1138
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1125
Additionally, it is undisputed that the doctrine of informed consent is based on the concept of autonomy. Under the most patientfriendly standard of the doctrine, a physician informs a patient of the
material risks associated with the procedure in order to facilitate an
informed decisionmaking process.97 However, to allow practicability
of the doctrine, the materiality of a risk must be limited to risks specifically related to the medical procedure itself. 98 Determining the
materiality of a risk specifically related to the medical procedure has
already proven to be a difficult task.99 If we further consider the physician’s personal characteristics to be a material risk, the doctrine
will eventually prove unworkable because of undefined boundaries.100
By disclosing the risks “specifically germane to surgical or operative treatment,”101 a patient is able to exercise his “right to determine
what shall be done with his own body.”102 This limitation to the concept of materiality accomplishes the doctrine’s intended goal: autonomy.103 However, considering a physician’s impairment to be a material risk is not concerned with autonomy, it concerns something entirely separate: patient safety.
The arguments in favor of expanding the informed consent doctrine to impose a disclosure obligation of the physician’s alcoholism,
drug abuse, or mental illnesses all implicitly center around the same
general argument: such physician impairments negatively affect a
physician’s performance of medical procedures, thereby compromising the safety of the patient.104 Yet, the informed consent doctrine
97. See Atwell, supra note 8, at 594-95.
98. The informed consent doctrine should not be expanded “to include matters not
specifically germane to surgical or operative treatment.” Kaskie v. Wright, 589 A.2d 213,
217 (Pa. Super. 1991); see also Semeraro v. Connolly, Civ. A. No. 92-4636, 1992 WL 392621
(E.D. Pa. Dec. 14, 1992).
99. See HALL ET AL., supra note 9, at 217-18; Iheukwumere, supra note 78, at 392
(“Subsequent to Canterbury many important informed consent decisions came down addressing the materiality of information in divergent ways . . . .”); see also Feinberg, supra
note 7, at 621 (“There is no clear consensus in the courts or in the medical literature about
physician disclosure and the informed consent process.”).
100. See Kaskie, 589 A.2d at 217 (refusing to expand the informed consent doctrine to
include matters not specifically germane to the surgical treatment because such expansion
would “extend[] the doctrine into realms well beyond its original boundaries” and limitations would not be easily definable).
101. Kaskie, 589 A.2d at 217.
102. See Schloendorff v. Soc’y of N.Y. Hosp., 105 N.E. 92, 93 (1914), abrogated by Bing
v. Thunig, 142 N.E.2d 3, 7-9 (1957), superseded by statute, Act of May 21, 1975, ch. 109,
1975 N.Y. Laws 134-35.
103. Kurtz, supra note 5, at 1251.
104. See, e.g., Hidding v. Williams, 578 So. 2d 1192, 1198 (La. Ct. App. 1991) (determining that the physician’s abuse of alcohol “increased [the] potential for injury during
surgery”); Feinberg, supra note 7, at 597 (stating that physician impairment results in
“patients . . . receiv[ing] a lower standard of care than they would otherwise receive”); Furrow, supra note 42, at 293 (“Certainly an alcoholic surgeon may have an impairment that
might seriously affect performance and thus success rate.”); Iheukwumere, supra note 78,
2014]
SCOPE OF INFORMED CONSENT
1139
was not developed with the purpose of ensuring patient safety.105 The
doctrine was simply intended to allow patients to weigh the potential
risks against the potential benefits of a specific medical procedure in
order to informatively decide for themselves whether they wish to
proceed with the procedure.106 However, within this informed decisionmaking process, the patient must be able to assume that the
physician will operate in a reasonably safe manner; the patient must
be able to proceed under an assumption that the physician’s alcoholism, drug abuse, or mental illness will not pose a significant risk.107
The solution to such a dilemma is not an expansion of the informed consent doctrine. If a physician is truly impaired by alcoholism, drug abuse, or a mental illness such that it will significantly and
negatively affect his performance of medical procedures to a degree
that would compromise patient safety, the physician should not be
able to simply disclose this information to a patient and avoid any
potential liability. The informed consent doctrine should not be expanded such that it can be used strategically as a liability “out”
where physicians are truly unfit to perform medical procedures. Instead, stepping back and considering the bigger picture is necessary.
V. DUTIES OF THE HOSPITAL: A CORPORATE NEGLIGENCE THEORY
“[With impaired physicians,] the goal is to eliminate the risk to the patient, not simply to inform the patient and let the patient choose.”108
Despite various opinions in medical literature, something we
know for sure is that physicians who are currently suffering or have
suffered from alcoholism, drug abuse, or mental illness may pose
threats to patient safety by performing medical operations.109 If such
impairment reaches a level where patient safety would be compromised, the solution is not disclosure through expanded informed con-
at 402 (“[A]n alcohol impaired physician . . . would impact on the likelihood of a mistake
during the procedure.”); see also Judice v. Hosp. Serv. Dist. No. 1, 919 F. Supp. 978, 984
(E.D. La. 1996) (holding that the discovery of alcohol in a physician’s system just prior to
surgery raised valid concerns about the risk to public safety).
105. “[P]atient autonomy is at the heart of all inadequate consent cases . . . .” McNichols, supra note 11, at 715.
106. See Atwell, supra note 8, at 596 (“Informed consent, then, is designed to protect
patients by ensuring that they have the material information with which to make an
informed choice.”).
107. See Bobinski, supra note 1, at 302 (discussing other personal physician conflicts:
“we want to ensure that patients receive good quality care of sufficient amount, duration,
and scope regardless . . . .”).
108. Furrow, supra note 42, at 293.
109. See Feinberg, supra note 7, at 598.
1140
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1125
sent obligations, but rather a bigger-picture solution: placing a duty
on hospitals to properly credential and monitor physicians.110
A hospital’s duty to non-negligently credential and monitor physicians is an obligation that is part of a larger conception known as
the “corporate negligence theory.”111 Under the corporate negligence
theory, a hospital can be held directly liable for injuries to patients
resulting from substandard medical care.112 The reasons for such a
theory are obvious: as blame for sub-quality medical treatment
continues to be placed directly on the treating physicians, hospitals
will be free of any duty (and thus any incentive) to try and prevent
medical errors.113
Consider the nature of the modern hospital. Hospitals are big
businesses, spending millions marketing themselves through “expensive advertising campaigns.” They provide a range of health services,
and the public expects emergency care, radiological and other testing
services, and other functions as a result of hospitals’ self-promotion.
And yet the legal relationships in the hospital are byzantine, creating
two strongly autonomous management structures side by side: a hospital administrative structure in parallel with the hospital medical
staff, which operates as a staff of independent contractors. The very
existence of this odd structure shields hospitals from liability under
agency law rules for the errors of their physicians, even when it is the
hospital systems that have allowed the physicians to fail.114
The concept of placing liability directly on a hospital under a corporate negligence theory has long been argued as an abdication of the
medical practice; “only an individual properly educated and licensed,
and not a corporation, may practice medicine.”115 Traditionally, hospitals were seen as merely the structure under which physicians operated. 116 Nonetheless, in 1965, the Supreme Court of Illinois announced its decision to impose liability directly on a hospital for inju110. See Furrow, supra note 42, at 293 (“[D]isclosure of [the physicians’] limitations
may be of questionable efficacy . . . . It risks destroying provider privacy while reducing the
pressure on state authorities and hospitals to monitor their physicians and set proper and
reasonable standards for practice.”).
111. See Judith M. Kinney, Tort Law—Expansion of Hospital Liability Under the Doctrine of “Corporate Negligence”—Thompson v. Nason Hospital, 591 A.2d 703 (Pa. 1991), 65
TEMP. L. REV. 787, 792 (1992).
112. HALL ET AL., supra note 9, at 481 (“Direct or ‘corporate’ liability contrasts with
vicarious liability in that it imposes on hospitals a duty of care owed directly to patients
with respect to medical judgment.”).
113. See Furrow, supra note 42, at 459.
114. Id. (emphasis added).
115. Darling v. Charleston Cmty. Mem’l Hosp., 211 N.E.2d 253, 256 (Ill. 1965).
116. See Pedroza v. Bryant, 677 P.2d 166, 169 (Wash. 1984) (“The hospital’s role is no
longer limited to the furnishing of physical facilities and equipment where a physician
treats his private patients and practices his profession in his own individualized manner.”).
2014]
SCOPE OF INFORMED CONSENT
1141
ries resulting from physician error.117 Darling v. Charleston Community Memorial Hospital has since been seen as the seminal case for
imposing direct liability on a hospital via a corporate negligence theory for substandard medical care.118
A hospital’s duty to non-negligently credential its physicians is the
critical first step to addressing the issue of impaired physicians. This
duty “entails reviewing physicians’ competency and performance history before admission to the medical staff and periodically (typically
every two years) thereafter.”119 The Supreme Court of Wisconsin in
Johnson v. Misericordia Community Hospital described the minimum requirements for a hospital (or the hospital’s credentialing
committee) in investigating and evaluating an applicant’s qualifications for hospital privileges:
[A] hospital should, at a minimum, require completion of the application and verify the accuracy of the applicant’s statements . . . . Additionally, it should: (1) solicit information from the
applicant’s peers, including those not referenced in his application,
who are knowledgeable about his education, training, experience,
health, competence and ethical character; (2) determine if the applicant is currently licensed to practice in this state and if his licensure or registration has been or is currently being challenged;
and (3) inquire whether the applicant has been involved in any adverse malpractice action and whether he has experienced a loss of
medical organization membership or medical privileges or membership at any other hospital. The investigating committee must
also evaluate the information gained through its inquiries and
make a reasonable judgment as to the approval or denial of each
application for staff privileges.120
Additionally, in 1986, the National Practitioner Data Bank
(NPDB) was established as part of the Healthcare Quality Improvement Act121 and was designed to provide licensing and credentialing
entities with a more uniform body of information regarding discipli-
117. Darling, 211 N.E.2d at 258.
118. HALL ET AL., supra note 9, at 481. “Conventional forms of direct liability entail
primarily administrative, not medical, functions such as maintaining safe premises, sterile
equipment, and adequate rules and regulations. Darling is recognized as extending direct
corporate liability to substandard medical care rendered by independent doctors. Hospitals
thus can be found liable for some act of negligence on their part with respect to patient care
decisions made by independent doctors . . . .” Id.
119. Id. at 482.
120. Johnson v. Misericordia Cmty. Hosp., 301 N.W.2d 156, 174-75 (Wis. 1981).
121. Healthcare Quality Improvement Act of 1986, 42 U.S.C. §§ 11101–11152 (2006).
1142
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1125
nary actions taken against physicians.122 Especially following the establishment of this databank, the duty to properly credential physicians makes clear that a hospital cannot simply plead ignorance with
respect to a physician’s information in granting that physician privileges.123 Imposing this duty on hospitals also means that if a hospital
chooses to grant privileges to a physician who has suffered from alcoholism, drug abuse, or mental illness in the past, the hospital bears
the risk of such impairments resurfacing. Through the credentialing
process, the hospital should be regarded as having been put on notice
about potential relapses that may occur and thus should be charged
with taking necessary, continual steps to prevent such occurrences.
Further, a hospital should be free to grant privileges to a physician
currently undergoing treatment for alcoholism, drug abuse, or mental
illness; however, it should be presumed that the hospital underwent
investigative measures to ensure that this physician is able to safely
practice medicine.124 The AMA has made clear that it opposes the discrimination against otherwise capable physicians solely because “the
physician is either presently, or has in the past, been under the supervision of a medical licensing board in a program of rehabilitation.”125
Hospitals must be charged with granting privileges to only those
physicians capable of safely practicing medicine. This may entail a
standardized screening evaluation for current alcoholism, drug
abuse, or mental illness. When this evaluation is conducted at the
commencement of the credentialing process, it will be regarded as
nondiscriminatory and can be a step towards removing the stigmatization of such impairments in the medical profession. In a case for
negligent credentialing, evidence to show a breach of this duty would
122. “The comprehensive reporting system requires medical boards to report licensure
revocations, suspensions, restrictions, censures, reprimands, probation, and licenses surrendered relating to the physician’s professional competence or professional conduct. The
statutory restrictions of the ADA do not insulate impaired physicians from being reported
to the NPDB.” Yuri N. Walker, Protecting the Public: The Impact of the Americans with
Disabilities Act on Licensure Considerations Involving Mentally Impaired Medical and
Legal Professionals, 25 J. LEGAL MED. 441, 447 (2004).
123. See Elisabeth Ryzen, The National Practitioner Data Bank: Problems and Proposed Reforms, 13 J. LEGAL MED. 409, 411, 419 (1992) (“Hospitals are required to query the
Data Bank before granting physicians privileges. . . . Failure to request information [from
the NPDB every two years] means the hospital will be presumed to have knowledge of any
information in the Data Bank concerning the applicant in any subsequent lawsuit.”). Nonetheless, the consistency of the NPDB regarding the reporting of physician impairments
needs improvement. See POLICIES RELATED TO PHYSICIAN HEALTH § H-355.992 (Am. Med.
Ass’n 2011) (“Our AMA will continue to monitor the issue of reporting impaired physicians
to the National Practitioner Data Bank and will seek further clarification of ambiguities or
misinterpretations of the reporting requirements for impaired physicians.”).
124. See Katharine A. Van Tassel, Blacklisted: The Constitutionality of the Federal
System for Publishing Reports of “Bad” Doctors in the National Practitioner Data Bank, 33
CARDOZO L. REV. 2031, 2057 (2012); see also Kinney, supra note 111, at 793-95 (discussing
actual or constructive notice triggering a hospital’s duties to act).
125. POLICIES RELATED TO PHYSICIAN HEALTH, supra note 123, § H-275.949.
2014]
SCOPE OF INFORMED CONSENT
1143
include the hospital’s own bylaws on credentialing physicians, state
licensing regulations, the Joint Commission’s Standards for Hospital
Accreditation,126 common law, and other relevant authorities.127
The hospital’s duty to non-negligently credential physicians is
generally accepted by the courts as part of the obligations of hospital
administration. 128 However, an affirmative duty on the hospital to
monitor its physicians has been seen as more controversial.129 Several
courts have rebuffed any duty to contemporaneously monitor treatment decisions made by its physicians, reasoning that it would constitute an unlawful interference of the physician-patient relationship.130
Starting at the local level, under a corporate negligence theory, a
hospital must be much more active in monitoring its physicians to
ensure those practicing medicine are safely capable of doing so. 131
Given that the hospital should have non-negligently credentialed its
physicians, it will be charged as being on notice of any potential red
flags in the physician’s record of past or current alcoholism, drug
abuse, or mental illness. The hospital should have a heightened affirmative duty to monitor these physicians.
In order to adequately monitor physicians, general awareness of
the prevalence of impaired physicians must increase. 132 Hospitals’
risk management committees should go through training and education programs on the potential signs of alcoholism, drug abuse, or
mental illnesses. Similar information should be distributed to all
hospital staff so that everyone is aware of the symptoms to look for in
potentially impaired physicians. The frequent monitoring of these
high-risk physicians, then, should include frequent in-person interviews, self-evaluations, and anonymous peer-review programs.
To ensure that such monitoring programs are effective, however,
hospitals also need to have a system in place that guarantees confi126. See JOINT COMMISSION, www.jointcommission.org (last visited June 29, 2014).
127. See Kinney, supra note 111, at 790-91.
128. See HALL ET AL., supra note 9, at 482. But see Paulino v. QHG of Springdale, Inc.,
386 S.W.3d 462, 467 (Ark. 2012) (refusing to recognize a claim for negligent credentialing
on the basis that regulatory oversight of hospitals is sufficient).
129. See HALL ET AL., supra note 9, at 482.
130. See, e.g., Gafner v. Down East Cmty. Hosp., 735 A.2d 969, 979-80 (Me. 1999).
131. Id. at 979 (quoting Pedroza v. Bryant, 677 P.2d 166, 168-69 (Wash. 1984)) (“The
doctrine of corporate negligence has . . . been utilized by courts to require hospitals to exercise reasonable care to insure that the physicians selected as members of hospital medical
staffs are competent. [Those courts] have also held that hospitals have a continuing duty to
review and delineate staff privileges so that incompetent staff physicians are not retained.”). “Health care’s casual approach to monitoring physician performance contrasts
markedly to that of other professions whose conduct affects the public welfare. Commercial
pilots, for example, must pass both physical and performance examinations every year.”
Leape, supra note 58, at 109.
132. See Feinberg, supra note 7, at 597 (“Physician addiction is taboo, but this silence
injures both the physician and his patients.”).
1144
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1125
dentiality and includes penalty waivers.133 Because many physicians
already conceal any impairment they may have for fear of potential
penalties and stigmatization from their peers, hospitals must ensure
that all monitoring policies are conducted with the utmost respect for
the physician’s confidentiality.134 Upon discovering alcoholism, drug
abuse, or mental illnesses in physicians, hospitals should motivate
those physicians to seek recovery through participation in treatment
programs.135 Every state has a form of a physician health program
designed specifically for physicians suffering from alcoholism or substance abuse.136 Completion of such a program should be a prerequisite for waiving the penalty of a complete revocation of hospital privileges, which will encourage physicians to obtain the treatment they
need but all so often do not get.
Imposing liability on hospitals for negligent credentialing and
monitoring of physicians, rather than expanding the doctrine of
informed consent, will place the task of determining the competency
of physicians with the party in the more capable position to do so:
the hospital.
VI. CONCLUSION
“The real issue is whether the physician is likely to perform well,
regardless of an impairment or other personal characteristic.” 137
Physicians suffering from alcoholism, drug abuse, or mental illnesses may be perfectly capable of performing medical treatments to
the necessary standard of quality.138 However, should such impairments rise to a level such that the safety of the patient is compromised, liability must fall somewhere in order to ensure proper preventative measures are taken and injured patients are not left without legal recourse.
The prevalence of physicians suffering from alcoholism, drug
abuse, or mental illnesses is a sad reality that cannot continue to be
ignored.139 However, attempting to solve such a large-scale issue by
expanding the scope of disclosure obligations under the informed
consent doctrine would be both impracticable and counterintuitive.
Such an expansion would be inconsistent with the original purpose of
133. See Feinberg, supra note 7, at 608.
134. See id.
135. See id. at 607-09.
136. See Cicala, supra note 57, at 43-44.
137. Furrow, supra note 42, at 293.
138. See id. (“[A disclosure obligation] raises a difficult causation question, since alcoholism may not always impair a physician’s performance.”).
139. See generally Feinberg, supra note 7, at 605 (discussing the prevalence of alcoholism, drug abuse, and similar impairments in the medical field).
2014]
SCOPE OF INFORMED CONSENT
1145
informed consent and would render the doctrine so lacking in certainty that it would eventually prove completely unworkable.
Further, confidentiality is an essential element for the proper
identification of, treatment of, and recovery from alcoholism, drug
abuse, and mental illness. Patients want to be sure that impaired
physicians, incapable of safely practicing medicine, are identified and
obtain the treatment they need. Yet, requiring physicians to disclose
to patients that they suffer from, or have suffered from, such impairments would be entirely illogical to achieve such a goal.
Addressing the issue of impaired physicians is part of a much
larger web of actors than can be remedied through the informed consent doctrine. Awareness to the problem of impaired physicians must
increase generally within the medical profession. State medical
boards and the Joint Commission on Accreditation of Health Care
Organizations must cooperate with hospitals so that they have the
necessary tools and information to begin the process.
Rather than expanding the informed consent doctrine, liability
should be placed on the hospital under a corporate negligence theory
when a hospital negligently credentials and monitors its physicians.
The hospital is in the best position to ensure the competency of the
physicians performing medical treatments, not the patient. Such a
structure would protect the privacy of the physician and encourage
detection and rehabilitation while having the effect of increasing patient safety on a broader scale.
1146
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1125
DURBIN AMENDMENT TO THE DODD FRANK ACT:
TWO CAPS ARE BETTER THAN ONE FOR DEBIT CARD
INTERCHANGE FEES
MAUREEN KANE
I. INTRODUCTION ..................................................................................................
II. DODD-FRANK WALL STREET REFORM AND CONSUMER PROTECTION ACT ........
A. History of the Financial Crisis and Implementation of the Dodd-Frank
Act ..............................................................................................................
B. Durbin Amendment ...................................................................................
1. Reasonable and Proportional...............................................................
2. Considerations .....................................................................................
3. Fraud Prevention .................................................................................
4. Small Bank Exemption ........................................................................
5. Final Rule ............................................................................................
III. DEBIT CARDS AND INTERCHANGE FEES ............................................................
A. Debit Card Transactions ...........................................................................
B. Debit Card Market .....................................................................................
C. The Hidden Fee ..........................................................................................
IV. EFFECTS ON THE PARTIES .................................................................................
A. Cost Savings for Merchants .......................................................................
B. Lost Profits on Low-Priced Items ...............................................................
C. Decrease Revenue Generation for the Banks..............................................
D. Effect on Exempted Banks .........................................................................
E. Use of Other Methods of Payment ..............................................................
V. SOLUTIONS ........................................................................................................
A. Leave the Durbin Amendment “As Is”—Cap on the Total Fee...................
B. Repeal the Durbin Amendment—No Caps ................................................
C. No Interchange Fee ....................................................................................
D. Cap the Percentage of the Transaction ......................................................
E. Double Cap: Cap Percentage of the Transaction Price and Total
Interchange Fee ..........................................................................................
VI. CONCLUSION .....................................................................................................
1147
1150
1150
1150
1151
1151
1152
1152
1153
1153
1154
1156
1157
1158
1158
1159
1162
1163
1163
1164
1165
1165
1166
1167
1167
1171
I. INTRODUCTION
Ignorance is not always bliss. Every day in the United States,
there are over one hundred million debit card transactions completed to purchase goods or services;1 however, most consumers do not
know or even care how they work–they just care that they work.
This ignorance allows the card-issuing banks to charge “hidden”
fees, known as interchange fees. Every time a debit card is swiped,
J.D. 2014, magna cum laude, Business Law Certificate with High Honors, Florida
State University College of Law; B.A., Biology and minor in Business Administration,
2011, cum laude, University of Florida. I would like to thank Professor Manual A. Utset,
Jr. for his guidance and comments. I would also like to thank my family for their unconditional love and support.
1. See Debit Card Interchange Fees and Routing, 75 Fed. Reg. 81,722, 81,723 (proposed Dec. 28, 2010) (to be codified at 12 C.F.R. pt. 235). There were 37.9 billion transactions in 2009. Id.
1148
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1147
fees are deducted from the amount taken out of the consumer’s bank
account,2 and the remainder is deposited into the merchant’s bank
account. Since, normally, neither merchants nor consumers are fully
aware of the exact cost of the debit card interchange fees, the cardissuing bank can abuse these fees.3 Even though merchants are
aware of the fees, they likely will not stop accepting debit cards because they need to remain competitive in the market, and because
consumers want and expect to be able to use them.
Before the Durbin Amendment, interchange fees were a percentage of the transaction cost and could be negotiated between the merchants and the networks, such as Visa and MasterCard.4 Obviously,
larger merchants with a greater number of transactions were able to
negotiate lower percentage fees than smaller merchants.5 Senator
Richard Durbin and other members of Congress saw a problem with
this, so they proposed an amendment to the Dodd-Frank Wall Street
Reform and Consumer Protection Act (Dodd-Frank Act), seeking to
“help small businesses, merchants, and consumers by providing relief
from high interchange fees for debit card transactions.”6 The Durbin
Amendment gave the Federal Reserve Board (Board) the rulemaking
power to create regulations for debit card interchange fees so that
they “shall be reasonable and proportional to the cost incurred by the
issuer with respect to the transaction.”7 The Board was tasked to
take into consideration the cost per transaction for the card issuer.8
The final rule adopted by the Board provides that the interchange
2. Fees are not added to the purchase price stated when the consumer swipes his or
her card, but instead are deducted from the price stated, so the merchant does not get the
full transaction price. See Patrick C. McGinnis, Misguided Regulation of Interchange Fees:
The Consumer Impact of the Durbin Amendment, 25 LOY. CONSUMER L. REV. 285, 286 (2013).
3. This was true before and after the Durbin Amendment. See infra Part III.C.
4. See Richard A. Epstein, The Constitutional Paradox of the Durbin Amendment:
How Monopolies are Offered Constitutional Protections Denied to Competitive Firms, 63
FLA. L. REV. 1307, 1315 (2011). The interchange fees were not negotiated between merchants and the card-issuing bank. Id. Networks also had the ability to create categories
based on the type of merchant or the transaction volume. See Debit Card Interchange Fees
and Routing, 75 Fed. Reg. at 81,723.
5. A merchant is able to negotiate the fees with its bank, and the merchant’s bank
and the card-issuing bank can negotiate their fees. Interchange Myths and Facts,
MASTERCARD, www.mastercard.com/us/merchant/pdf/021208 MythsFacts.pdf (last visited
Apr. 6, 2014). To avoid thousands of separate negotiations, MasterCard sets “default”
interchange rates. Id.
6. 156 CONG. REC. S4839 (daily ed. June 10, 2010) (statement of Sen. Richard Durbin) (“My amendment sought to give small businesses and merchants and their customers
across America a real chance in the fight against the outrageously high swipe fees charged
by Visa and MasterCard credit card companies.”); see also 156 CONG. REC. S3695 (daily ed.
May 13, 2010) (statement of Sen. Richard Durbin).
7. Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank Act),
Pub. L. No. 111-203, § 1075, 124 Stat. 1376, 2068 (2010) (emphasis added).
8. See id.; Debit Card Interchange Fees and Routing, 75 Fed. Reg. at 81,733.
2014]
DURBIN AMENDMENT
1149
fees charged cannot exceed the sum of $0.21, plus five basis points of
the value of the transaction, as a fraud-prevention adjustment.9
This Note does not suggest changing the goals of the Durbin
Amendment, but instead suggests an alteration to the regulations
created by the Board. The regulations suggested in this Note aim to
better match the Amendment’s goals of being “reasonable and proportional” while taking into consideration the card issuer’s cost per
transaction. This Note suggests placing a double cap on interchange
fees.10 The suggested fees would be reasonable because they would
cap the total amount that can be charged based on the cost per
transaction, and would be proportional because they would base the
fees on a capped percentage of the purchase price. This Note does not
suggest an exact value of either cap, because it would require more
information to ensure that the caps capture the actual cost per
transaction while taking into account the number of transactions
with percentages that are potentially below the actual cost per transaction. This Note focuses primarily on the three parties participating
in the two-sided market: (1) the merchant, (2) the consumer, and (3)
the card-issuing bank. There are other parties involved in the debit
card transfer, but their role in the realm of interchange fees is minimal.11 These three parties are in a two-sided market because the
card-issuing bank’s goal is to make sure that both merchants and
consumers want to use its card. This is a careful balancing act because
both merchants’ and consumers’ decisions are affected by fees. If the
fees are too high then merchants will not accept the card, and if merchants do not accept the card, then consumers will not want the card.
Part II examines the economic background and implementation of
the Dodd-Frank Act and then explains the Durbin Amendment’s key
parts. Part III explains how a debit card transaction takes place and
the two-sided nature of these transactions. It also discusses the hidden nature of the debit card interchange fees. Part IV discusses the
effects that the Durbin Amendment has on each of the parties, including: merchants’ cost savings or lost profits on low-priced items,
banks’ loss of revenue generation, the effect on exempted banks, and
the possibility of switching to other methods of payment.
Part V discusses the advantages and disadvantages of some potential alternatives. This Note discusses five possible alternatives to
this problem: (1) leave the Durbin Amendment “as is;” (2) repeal the
Durbin Amendment and return to a pre-amendment status with no
regulation; (3) have no interchange fees; (4) change the Amendment
9. Debit Card Interchange Fees and Routing, 76 Fed. Reg. 43,394, 43,404 (July 20,
2011) (to be codified at 12 C.F.R. pt. 235).
10. See infra Part V.E.
11. See infra Part III.A.
1150
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1147
to cap the fees on the percentage of the sale; or (5) change the
Amendment to cap fees on the percentage of the sale price while capping the total interchange fees. Ultimately this Note will suggest an
alteration to the Durbin Amendment that caps both the fees based on
a percentage of the sale price and the total amount allowed in interchange fees, so that the regulation more closely tracks the Amendment’s goal of being “reasonable and proportional.”
II. DODD-FRANK WALL STREET REFORM AND CONSUMER PROTECTION
ACT
A. History of the Financial Crisis and Implementation of the DoddFrank Act
The U.S. Government passed the Glass-Steagall Act after the
Great Depression, which created banking regulations.12 However, in
1999, President Clinton repealed many of the banking regulations
that were established under the Glass-Steagall Act.13 Many people
believe that one cause of the 2008 financial crisis was the deregulation of the banking industry and the fact that new regulations were
needed to rehabilitate the economy.14
On July 21, 2010, the Dodd-Frank Act became law.15 The DoddFrank Act’s goal, as stated in its short title, is “[t]o promote the financial stability of the United States by improving accountability
and transparency in the financial system, to end ‘too big to fail’, to
protect the American taxpayer by ending bailouts, to protect the consumers from abusive financial services practices, and for other purposes.”16 Whether the Dodd-Frank Act is meeting this goal is a topic
for many other articles.
B. Durbin Amendment
A few months after the Dodd-Frank Act was passed, Senator Durbin recommended an addition, seeking to help small businesses.17
This addition can be found in section 1075 of the Dodd-Frank Act,
entitled “Reasonable Fees and Rules for Payment Card Transactions,” but is commonly referred to as the Durbin Amendment.18 The
12. See McGinnis, supra note 2, at 292.
13. Id.
14. See id. at 292-93.
15. Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank Act),
Pub. L. No. 111-203, § 1, 124 Stat. 1376, 1376 (2010); McGinnis, supra note 2, at 293.
16. Dodd-Frank Act § 1.
17. See 156 CONG. REC. S4839 (daily ed. June 10, 2010) (statement of Sen. Richard Durbin).
18. Dodd-Frank Act § 1075.
2014]
DURBIN AMENDMENT
1151
Durbin Amendment, which took effect on October 1, 2011, adds section 920 to the Electronic Fund Transfer Act (EFTA).19
Before the Dodd-Frank Act, debit networks incorporated an ad
valorem percentage to determine the interchange fees for each
transaction.20 Networks also had the ability to create categories
based on the type of merchant or the transaction volume.21 Before
the Durbin Amendment, the average interchange fee was between
1.15% and 1.35% of the transaction cost, and averaged approximately
$0.47 per transaction.22
1. Reasonable and Proportional
The Durbin Amendment gave the Board the rulemaking power to
create regulations “regarding any interchange transaction fee that
an issuer may receive or charge with respect to an electronic
debit transaction.”23 The principle provided as guidance to the Board
was that the interchange transaction fees “shall be reasonable and
proportional to the cost incurred by the issuer with respect to
the transaction.”24
2. Considerations
The Board was directed to consider the “functional similarity” between debit transactions and checking transactions that clear at par,
as well as “the incremental cost incurred by an issuer for the role of
the issuer in the authorization, clearance, or settlement of a particular electronic debit transaction.”25 However, the Board was not to
consider “other costs incurred by an issuer which are not specific to a
particular electronic debit transaction.”26
The Board requested public comment, as is required in the rulemaking process, to determine an interchange fee for debit transaction
that shall be “reasonable and proportional to the cost[s] incurred.”27
19. See 15 U.S.C. § 1693 (2012); Dodd-Frank Act § 1075.
20. Debit Card Interchange Fees and Routing, 75 Fed. Reg. 81,722, 81,724 (proposed
Dec. 28, 2010) (to be codified at 12 C.F.R. pt. 235).
21. See id. at 81,723.
22. Richard A. Epstein, The Dangerous Experiment of the Durbin Amendment: Congress’ Interchange Fee Limit is a Reckless Exercise in Price Regulation, REG.: CATO REV.
BUS. & GOV’T, Spring 2011, at 24, 25; CADY NORTH, BUSINESS IMPACT OF THE DODD-FRANK
DEBIT FEE CAP, BLOOMBERG GOV’T STUD. 12 (Lisa Getter ed., 2011).
23. Dodd-Frank Act § 1075.
24. Id. (emphasis added).
25. Id. See infra Part III for an explanation of a check clearing at par.
26. Dodd-Frank Act § 1075.
27. Debit Card Interchange Fees and Routing, 75 Fed. Reg. 81,722, 81,722 (proposed
Dec. 28, 2010) (to be codified at 12 C.F.R. pt. 235). See infra Part II.B.1.
1152
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1147
The Board originally proposed two alternatives.28 Alternative
one combined a cap of $0.12 per transaction and a safe-harbor level
of $0.07 per transaction.29 Alternative two was simply a cap of
$0.12 per transaction.30
The Board received over 11,500 comments from “issuers, payment
card networks, merchants, consumers, consumer advocates, trade
associations, and members of Congress.”31 The merchants suggested
lowering the cap so that the savings could be passed along to consumers because the costs were significantly below the $0.12 cap.32
Merchants also suggested prohibiting exclusivity arrangements to
increase competition.33 On the other hand, payment card networks,
such as Visa and MasterCard, were concerned that the cap would
decrease revenue for card-issuing banks, which would increase cardholder fees and decrease benefits for merchants and consumers.34 After considering all the comments, the Board adopted a modified version of alternative two.35
3. Fraud Prevention
In addition, the Durbin Amendment allows the Board to create an
adjustment to the interchange fees if the “adjustment is reasonably
necessary to make allowance for costs incurred by the issuer in preventing fraud in relation to electronic debit transactions involving
that issuer; and the issuer complies with the fraud-related standards
established by the Board.”36 The goal of this addition is to encourage
the card-issuing bank to take necessary actions to prevent fraud.37
4. Small Bank Exemption
The Durbin Amendment does not apply to any card issuer that
has less than $10 billion in assets.38 These “small issuers” will not
28. Debit Card Interchange Fees and Routing, 75 Fed. Reg. at 81,736.
29. See id. at 81,736, 81,755.
30. Id. at 81,738, 81,756.
31. Debit Card Interchange Fees and Routing, 76 Fed. Reg. 43,394, 43,394 (July 20,
2011) (to be codified at 12 C.F.R. pt. 235).
32. See id. at 43,402.
33. See id.
34. Id.
35. Id. at 43,404.
36. Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank Act),
Pub. L. No. 111-203, § 1075, 124 Stat. 1376, 2069 (2010).
37. Id.; Debit Card Interchange Fees and Routing, 76 Fed. Reg. 43,394, 43,404 (July
20, 2011) (to be codified at 12 C.F.R. pt. 235).
38. 15 U.S.C. § 1693o-2 (2012). The value of the assets was raised from $1 billion to
the $10 billion in order to get enough votes. Epstein, supra note 4, at 1326; see also, M.
Pierce Sandwith, Note, Debit Card Interchange Fees and the Durbin Amendment’s Small
Bank Exemption, 16 N.C. BANKING INST. 223, 234-36 (2012).
2014]
DURBIN AMENDMENT
1153
have to prescribe to the Board’s regulations of debit card interchange
transaction fees and can instead charge whatever fee they need to or
can.39 Only three credit unions and approximately sixty banks have
enough assets to be affected by the Durbin Amendment, while approximately 7500 credit unions and 7000 banks are exempt.40 One
reason for this exemption to the cap on interchange fees is that
smaller banks have fewer transactions, and thus, they have fewer
transactions across which they can spread their fixed costs.41
5. Final Rule
The final rule adopted by the Board is an adaptation to alternative two and provides that the interchange fees charged cannot exceed the sum of $0.21, plus five basis points of the value of the transaction, as a fraud-prevention adjustment.42
III. DEBIT CARDS AND INTERCHANGE FEES
When consumers want to purchase goods or services, they have a
variety of payment options, including but not limited to: cash,43
check,44 debit card, or credit card. However, consumers are increasingly shifting to electronic payment methods such as debit and credit
cards.45 This is good news for merchants, because they have an informational asymmetry problem when consumers pay with checks,
39. Sandwith, supra note 38, at 235-36.
40. Epstein, supra note 22, at 24.
41. See David C. John, The Durbin Debit Card Interchange Fee Cap Hurts Consumers,
THE HERITAGE FOUNDATION (Mar. 17, 2011), http://report.heritage.org/wm3194.
42. See Debit Card Interchange Fees and Routing, 76 Fed. Reg. 43,394, 43,404 (to be
codified at 12 C.F.R. pt. 235). See supra Part II.B for a more detailed explanation of the
Durbin Amendment.
43. In cash payments the purchaser provides the merchant with cash equal to the
amount of purchase price. The risk that the merchant will not get paid does not exist here
because goods and services are not exchanged without the full value of the cash unless
otherwise agreed upon (under certain circumstances such as contracts or the delivery of
goods, payment might not be tendered until after the goods are received). The downside of
cash payments for merchants is there is a risk of theft from employees or outsiders. Additionally, the payments are not automatically deposited, so the merchant or merchant’s
employees have to manually transfer the cash to the merchant’s bank.
44. In a check payment, the purchaser provides a check for the amount of the purchase price. After the merchant receives the check, the merchant has to manually transfer
the checks to the merchant’s banks because the deposit is not automatic. In the past few
years there have been improvements in check depositing, such as being able to deposit
checks by using the camera on your mobile phone. See, e.g., Press Release, Sun National
Bank, Sun National Bank Launches SunMobile Deposit (Oct. 8, 2013), available at
http://online.wsj.com/article/PR-CO- 20131008-906737.html. This only eliminates the deposit problem and not the risk of adequate funds. Furthermore, check writing is timeconsuming and will slow down the checkout process, which may lead to merchants needing
to staff more employees. See Epstein, supra note 4, at 1317.
45. Debit Card Interchange Fees and Routing, 75 Fed. Reg. 81,722, 81,723 (proposed
Dec. 28, 2010) (to be codified at 12 C.F.R. pt. 235).
1154
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1147
since merchants have no information about the consumers’ financial
positions.46 Additionally, consumers often make larger purchases
with debit cards and credit cards than they would with cash.47
Checks normally clear “at par,” and therefore, the merchant receives the full value of the check without fees.48 On the other hand,
credit and debit cards both have interchange fees, but these higher
fees are paid in exchange for greater merchant and consumer benefits, such as convenience and assuring the merchant that there are
available funds.49 Merchants are essentially paying for the option to
have more information about the consumers’ financial positions, and
to avoid making irreversible decisions without having all of the information.50 Both debit cards and credit cards also eliminate the hassle and risk of having and processing cash or check payments.51
A. Debit Card Transactions
Debit card usage has rapidly increased since the mid-1990s and is
the fastest-developing form of electronic payment.52 The Durbin
Amendment defines a debit card as “any card, or other payment code
or device, issued or approved for use through a payment card network to debit an asset account (regardless of the purpose for which
the account is established), whether authorization is based on signature, PIN, or other means.”53
46. In check transactions, the merchant assumes the risk that there might not be
enough money in the checking account to cover the cost of the purchase. See Epstein, supra
note 4, at 1317. The merchant also has no additional knowledge about the consumer’s financial position; therefore, the merchant has less information in these transactions than it
does in debit card transactions. Id.
47. John, supra note 41, at 1.
48. Epstein, supra note 4, at 1316. At first glance, it seems like merchants would prefer paying zero fees. However, there are some added risks, one being that the merchant
bears the risk that there might not be sufficient funds to cover the cost of the purchase,
and another being that the merchant takes this risk without any knowledge of the consumers’ financial status. Id. at 1316-17.
49. Epstein, supra note 22, at 29. Often the fees for credit cards are higher than the
debit card fee because the benefits for credit card users are generally greater. Id.
50. This is very similar to other option agreements or contracts because merchants
are purchasing the option to get more information before making a decision. See BLACK’S
LAW DICTIONARY 544 (4th pocket ed. 2011) (defining option agreement as “[a] share-transfer
restriction that commits the shareholder to sell, but not the corporation or other shareholders
to buy, the shareholder’s shares at a fixed price when a specified event occurs”).
51. See supra notes 43-44.
52. Debit Card Interchange Fees and Routing, 75 Fed. Reg. 81,722, 81,723 (proposed
Dec. 28, 2010) (to be codified at 12 C.F.R. pt. 235).
53. Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank Act),
Pub. L. No. 111-203, § 1075, 124 Stat. 1376, 2073 (2010).
2014]
DURBIN AMENDMENT
1155
Most debit card transactions are processed in the following manner, commonly referred to as the “open-loop” or “four-party” system.54
The parties in this system are: the consumer; the merchant; the merchant’s bank; the debit card network (e.g., Visa or MasterCard); and
the entity that issued the debit card to the consumer, that is, the
card-issuing bank.55 When a debit card is swiped, the consumer’s
identity and bank account information is made available; however, a
signature or a personal identification number (PIN) is needed to initiate the transaction.56 The transaction information is then sent to
the merchant’s bank.57 If the transaction is authorized, then the merchant’s bank sends the transaction information to a debit-card network.58 This debit-card network connects the merchant’s bank with
the consumer’s bank.59 A network is used to relay information
through the bank that issued the consumer’s debit card.60 The consumer’s bank then verifies that there are sufficient funds and that
the card has not been reported lost or stolen.61 If there is enough
money in the consumer’s debit account, then notice of a complete
transaction is sent to the merchant through the processors, and the
purchase is finalized.62 The consumer’s bank then transfers the
transaction amount from the consumer’s account to the merchant’s
account; however, the card-issuing bank assesses interchange fees
and other processing fees at this point.63 The bank that issued the
consumer’s debit card will transfer the funds to the merchant’s account at the merchant’s bank.64 The merchant will pay their bank for
interchange fees and other processing charges at a later time.65
Merchants pay three types of fees: (1) to their bank; (2) to card
companies, e.g., Visa and MasterCard; and (3) to the card-issuing
54. This Note focuses on the four-party, or open-loop system; however, it is important
to acknowledge that there is a less commonly used “three-party” or “closed-loop” system.
See Debit Card Interchange Fees and Routing, 75 Fed. Reg. at 81,723, n.6. The three-party
system is different because the merchant and consumers contract directly with the network, instead of the network setting the fees. See Lisa Farrell, A Step in the Right Direction: Regulation of Debit Card Interchange Fees in the Durbin Amendment, 15 LEWIS &
CLARK L. REV. 1077, 1082 (2011).
55. Debit Card Interchange Fees and Routing, 75 Fed. Reg. at 81,723.
56. NORTH, supra note 22, at 8.
57. Id.
58. Id.
59. Id.
60. Id.
61. Debit Card Interchange Fees and Routing, 75 Fed. Reg. 81,722, 81,723 (proposed
Dec. 28, 2010) (to be codified at 12 C.F.R. pt. 235).
62. NORTH, supra note 22, at 8.
63. See id. The card-issuing bank takes the interchange fee before transmitting to the
merchant’s bank, so the merchant does not receive the full value of the transaction. Id.
64. McGinnis, supra note 2, at 288.
65. NORTH, supra note 22, at 8.
1156
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1147
bank, i.e., the consumer’s bank.66 These fees are often one of the largest portions of merchants’ costs of doing business.67 The Durbin
Amendment only regulates the interchange fees to the consumer’s
card-issuing bank.68 In an open-loop system, the interchange fees are
set by the debit card-issuing bank69 and are often the largest portion
of the fees for each transaction.70 The debit card-issuing bank and the
merchant’s bank both receive a portion of the fees.71
Under their contractual obligations to networks such as Visa and
MasterCard, merchants are not allowed to assess a surcharge on consumers who use debit cards or credit cards instead of other payment
methods.72 The consumer does not directly see the effects of the interchange fees; instead, the fees are incorporated into the price of the
goods or services as a cost to the merchant.73
B. Debit Card Market
Electronic payment tools such as debit cards and credit cards operate in a two-sided market.74 A two-sided market occurs during a
three-party transaction, when the middle party has an incentive to
make sure that the other two sides want to do business together.75
“[T]he ability to satisfy one side of the market depends on the continued participation of the other.”76 In debit card transactions, cardissuing banks need both merchants to accept and consumers to use
debit cards in order to stay in business. Merchants will not accept a
company’s debit card if the fees are too high or if there are no consumers using that company’s card. Conversely, consumers will not
obtain or use a company’s card if few merchants accept it or if the
66. Id.
67. Gregory D. Wasson, the CEO of Walgreens, stated that interchange fees are his
company’s fourth-largest cost after salaries, mortgages and rent, and healthcare. Epstein,
supra note 22, at 24.
68. Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank Act),
Pub. L. No. 111-203, § 1075, 124 Stat. 1376, 2068 (2010).
69. See Debit Card Interchange Fees and Routing, 75 Fed. Reg. 81,722, 81,723 (proposed Dec. 28, 2010) (to be codified at 12 C.F.R. pt. 235); John, supra note 41, at 1.
70. Farrell, supra note 54, at 1083.
71. Id.
72. Id. Merchants can increase the cost of their goods or services and therefore spread
the cost of these fees among all payment methods.
73. See id.
74. See id. at 1087.
75. Epstein, supra note 4, at 1323. Examples of two-sided market scenarios include
the following: (1) Newspapers put a greater-cost burden on advertisers because advertisers
will only purchase ads if enough people read the newspaper, so the cost for readers has to
be lowered to encourage more readers, Marc Rysman, The Economics of Two-Sided Markets,
23 J. ECON. PERSP. 125, 128 (2009); and (2) Groupon tries to match businesses with customers, and typically, the fees are borne by the businesses, Epstein, supra note 4, at 1324.
76. Epstein, supra note 22, at 26.
2014]
DURBIN AMENDMENT
1157
fees are too high.77 Therefore, card-issuing banks have to carefully
balance the interests of both parties when setting the amount of fees
and when determining who will bear the burden of those fees.78
In most two-sided markets, the fees are not shared equally.79 Typically, the party that is inelastic, i.e., less sensitive to changes in
price, will pay a greater portion of the costs.80 A merchant’s demand
for consumers is often less sensitive to changes in price than a consumer’s demand for the goods and services provided by the merchants.81 Therefore, merchants are more likely to bear a majority of
the costs associated with interchange fees.
C. The Hidden Fee
The debit card interchange fee is similar to a “hidden tax.”82 The
card-issuing banks are able to generate revenue with very little detection, because normally, neither the merchant nor the consumer is
fully aware of how much these fees cost since they are not billed separately to the merchant but instead are deducted from the amount
deposited into the merchant’s account.83 Regulations such as the
Durbin Amendment protect merchants from these hidden fees.
The consumer does not directly see the cost of the interchange
fees; instead, it is incorporated into the price of the goods or services
because it is a cost to the merchant.84 This is similar to a hidden tax
for consumers, because if they are unaware of these fees, their spending and use of the card is not affected by the price of the interchange
fee.85 However, if the listed prices are higher because of this additional merchant cost, then the consumer might be less likely to base
a purchase on the sticker shock than if the fee was added on at the
register.86 This is basic supply-and-demand economics: if the price is
77. See e.g., NORTH, supra note 22, at 15; McGinnis, supra note 2, at 301.
78. See Epstein, supra note 4, at 1323. For example, online dating services want to have
an equal number of men and women, so they charge lower prices to women to attract more of
them, because women are less likely to join an online dating service than men. Id. at 1324.
79. Id.
80. Id. The Ramsey pricing principle states that a price should be set so that the
price-marginal cost margins are inversely related to the sensitivity of demand to the price;
and thus, that more revenue will be recovered from the side with a more inelastic demand.
ORG. FOR ECON. CO-OPERATION & DEV. [OECD], POLICY ROUNDTABLES: TWO-SIDED
MARKETS 12-13, 25 (2009), available at http://www.oecd.org/daf/competition/44445730.pdf.
81. Epstein, supra note 22, at 26.
82. See generally Brian Galle, Hidden Taxes, 87 WASH. U. L. REV. 59 (2009) (explaining that a hidden tax is a tax that goes unnoticed by those who pay them).
83. NORTH, supra note 22, at 8; McGinnis, supra note 2, at 288.
84. See Farrell, supra note 54, at 1083.
85. See Galle, supra note 82, at 75-77.
86. See Raj Chetty et al., Salience and Taxation: Theory and Evidence, 99 AM. ECON.
REV. 1145, 1146 (2009).
1158
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1147
higher, fewer consumers will demand the good or service; however, if
the consumer is ignorant of the higher price—because it is hidden—
then it may not affect the demand.87
Since the interchange fees are taken out of the amount deposited
into the merchant’s bank account, merchants may be unaware of the
exact cost of these fees, unless they calculate the fees themselves.
The merchants never see a bill with the total cost of the interchange
fees.88 This might explain why the cost savings post-Durbin Amendment may not have been passed along to the consumer.
The hidden nature of the interchange fees is beneficial for the
card-issuing banks because they are able to generate revenue without negative effects from either side of the two-sided market. When
banks attempt to issue more obvious fees, they often lose consumers
to other methods of payment.89
IV. EFFECTS ON THE PARTIES
Senator Durbin and other members of Congress saw a problem
with the debit card interchange fee, so they enacted the Durbin
Amendment to help small businesses.90 It remains unclear whether
the Amendment actually helps small businesses;91 however, it does
affect all three parties—the merchants, the consumers, and the card
issuers—in a variety of ways. These effects are interesting because of
the nature of the two-sided market; i.e., something that affects one
party will also affect one of the other parties.92 Keeping merchants
and consumers in the market, while simultaneously allowing the
card issuer to make a profit, is a careful balancing act. This Note will
discuss a few of the big issues and then explain the interwoven effects on the parties involved.
A. Cost Savings for Merchants
In an ideal world, merchants would pass their cost savings along
to consumers; unfortunately, the real world is more complicated than
that. Merchants save over $8 billion per year with the Durbin
87. See, e.g., Galle, supra note 82, at 75-77.
88. See supra Part III.A.
89. On September 29, 2011, Bank of America and other major banks announced a $5
monthly fee for debit cards, but due to consumer outrage, the banks had to abandon the
plan about a month later. NORTH, supra note 22, at 9.
90. 156 CONG. REC. S4839 (daily ed. June 10, 2010) (statement of Sen. Richard Durbin).
91. See e.g., McGinnis, supra note 2, at 285.
92. See supra Part III.B.
2014]
DURBIN AMENDMENT
1159
Amendment interchange fees.93 If merchants do not pass these savings along to consumers, then consumers lose on both ends because
the banks have to make the lost revenue up by charging additional
fees or by cutting services.94 A decrease in debit card interchange fees
would be good for the economy if the savings were passed along to the
consumers and if consumer spending increased.
Scholars and analysts disagree on whether the savings are being
passed along to consumers. Some say that merchants are holding on
to the $8 billion in savings they received from the Durbin Amendment,95 while others argue that the savings are being passed along to
consumers.96 It is possible that merchants are unaware of their cost
savings, which would explain why they have not lowered their prices.97 There is also a possibility that card issuers will raise the interchange fees on credit cards because they are currently unregulated
by the Durbin Amendment.98 If these fees increase, there may not be
any cost savings for merchants to pass along to consumers.99
B. Lost Profits on Low-Priced Items
While high-priced items will create cost savings for merchants
with regard to the interchange fees, low-priced items will cost merchants more in fees.100 Merchants that sell items worth less than $20
will actually pay higher interchange fees than they were paying be-
93. Vicki Needham, Report: Businesses Passing on Swipe Fee Savings to Consumers,
THE HILL (Oct. 1, 2013, 5:16 PM), http://thehill.com/blogs/on-the-money/banking-financialinstitutions/325851-report-businesses-passing-on- swipe-fee-savings-to-consumers. This
also means that banks are losing $8 billion in revenue. See infra Part IV.C.
94. See infra Part IV.C.
95. The Durbin Effect: Two Years Later, Consumers Still Not Benefitting from Durbin
Amendment, ELEC. PAYMENTS COAL., http://wheresmydebitdiscount.com/the-durbin-effect/
(last visited Apr. 6, 2014). This article, however, cites a survey that looks at low-cost items,
such as milk, peanut butter, a hammer, and a Slurpee, to illustrate that the cost savings have
not been passed along. This is flawed logic because low-priced items have a higher burden with
the post-Durbin Amendment interchange fees than they did before the Durbin Amendment.
96. Vicki Needham, Retailers File Appeal Over Credit Card Swipe Fee Settlement, THE
HILL (Jan. 2, 2014, 6:19PM), http://thehill.com/policy/finance/194318-retailers-file-appealover-credit-card (quoting Mallory Duncan, Senior Vice President and General Counsel of
the National Retail Federation).
97. See supra Part III.C. There are also a lot of other factors that could explain the
increase in prices above inflation, such as an increase in other costs or an increase in demand for a good or service.
98. Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank Act),
Pub. L. No. 111-203, § 1075, 124 Stat. 1376, 2068 (2010).
99. This also suggests that if fees and costs can be shifted to credit card interchange
fees, then the mere regulation of debit card interchange fees is a frivolous task; therefore, it
would be beneficial to also regulate the credit card interchange fees, but this idea extends
beyond this Note.
100. See infra Table 1.
1160
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1147
fore the Durbin Amendment.101 Approximately 20% of all debit card
purchases are for low-priced items.102 These small-ticket sales come
from coffee shops, convenience stores, fast food restaurants, and other merchants that sell items or services worth less than $20.
Before the Durbin Amendment, there was special pricing for merchants with low-cost items; however, after the Durbin Amendment,
the networks set interchange fees at the maximum amount allowed.103 Therefore, instead of using the ad valorem interchange fee,
the maximum of $0.21, plus fraud prevention, is being charged.104 As
illustrated in Table 1, when the transaction cost is below approximately $20, merchants are actually paying more in interchange fees
than they were before the Durbin Amendment.
Table 1:
BEFORE
DURBIN
AMENDMENT
AFTER DURBIN
AMENDMENT –
FEE CAP
AFTER DURBIN
AMENDMENT –
FEE CAP PLUS
FRAUD
PREVENTION
MERCHANT
SAVINGS OR
(COSTS)
Interchange fee
of 1.15% per
transaction 105
Interchange fee
of $0.21 per
transaction
Interchange fee of
$0.21 per transaction plus 0.05%
Fee Before
minus Fee
After plus
Fraud
Prevention
$1
$0.0115
$0.21
$0.2105
($0.119)
$5
$0.0575
$0.21
$0.2125
($0.155)
$10
$0.115
$0.21
$0.215
($0.1)
$20
$0.23
$0.21
$0.22
$0.01
$25
$0.2875
$0.21
$0.2225
$0.065
$38106
$0.437
$0.21
$0.229
$0.208
$50
$0.575
$0.21
$0.235
$0.34
$100
$1.15
$0.21
$0.26
$0.89
PRICE OF
TRANSACTION
101. See id.
102. Dakin Campbell & Donal Griffin, Visa, MasterCard Risk “Mom and Pop” Ire with
Debit-Fee Increase, BLOOMBERG (Sept. 23, 2011), http://www.bloomberg.com/news/2011-0923/visa-mastercard-risk-mom-and-pop-ire-with-debit-fee-increase.html.
103. Sandwith, supra note 38, at 241-42.
104. See e.g., Visa U.S.A. Interchange Reimbursement Fees, VISA 2-3, 7-8 (Apr. 20,
2013), http://usa.visa.com/download/merchants/visa-usa-interchange-reimbursement-feesapril2013.pdf.
105. NORTH, supra note 22, at 6. This is just one example of the average interchange fee.
Another source stated the average interchange fee was 1.35%. Epstein, supra note 22, at 25.
106. The average purchase price for all debit transactions is $38. NORTH, supra
note 22, at 12.
2014]
DURBIN AMENDMENT
1161
BEFORE
DURBIN
AMENDMENT
AFTER DURBIN
AMENDMENT –
FEE CAP
AFTER DURBIN
AMENDMENT – FEE
CAP PLUS FRAUD
PREVENTION
MERCHANT
SAVINGS OR
(COSTS)
$250
$2.875
$0.21
$0.335
$2.54
$500
$5.75
$0.21
$0.46
$5.29
PRICE OF
TRANSACTION
Companies like Redbox and Starbucks that sell low-priced items
have had to raise prices to cover the increase in debit interchange
fees for their products and services.107 Redbox states the following on
its website regarding its increase in prices: “The increase is a result
of rising operational costs, including increased debit card fees.”108
Redbox raised the daily rental price from $1 to $1.20 because of the
increased fees. Redbox is also an interesting example because its business plan requires the use of a debit or credit card for all purchases, so
it cannot offset the increased costs by setting a minimum transaction
price or by switching to other payment methods. Instead, Redbox has
to pass the costs along to its consumers.109 Businesses such as coffee
shops, dollar stores, and convenience stores could offset some of the
higher fees by utilizing the minimum-purchase-price-per-transaction.
Merchants can now require a minimum purchase before allowing
a consumer to use a credit card or debit card.110 Before the DoddFrank Act, merchants would have been penalized for such requirements.111 Ideally, this would allow merchants to offset some of the
added costs of accepting debit cards for low-cost purchases. However,
merchants do not want to discourage consumers from making purchases, so merchants have little choice but to accept debit cards in
order to remain competitive.112 This forces merchants to pay the fees,
especially with respect to companies with products and services worth
less than $20.113 Companies, such as Redbox, with business models
that require the use of a debit card or credit card, and without the ability to use a minimum purchase price to offset the higher interchange
fees, cannot easily offset the added costs without raising prices.
107. General Questions: Price Change, REDBOX, http://www.redbox.com/pricechange
(last visited Apr. 6, 2014). There is speculation that Starbucks increased the price of some
of its items due to the increased interchange fees. See Willy Staley, Starbucks Price Hike:
Could Debit Interchange Be the Culprit?, MY BANK TRACKER (Jan. 9, 2012),
http://www.mybanktracker.com/news/2012/01/09/why-did-starbucks-raise-price/.
108. General Questions: Price Change, REDBOX, http://www.redbox.com/pricechange
(last visited Apr. 6, 2014).
109. Id.
110. NORTH, supra note 22, at 12.
111. Id.
112. Epstein, supra note 4, at 1319.
113. Id.
1162
FLORIDA STATE UNIVERSITY LAW REVIEW
[Vol. 41:1147
C. Decrease Revenue Generation for the Banks
The Federal Reserve estimates that banks will lose around $8 billion annually due to the current interchange fees, assuming there is
no change in debit card usage.114 This is approximately half of the
roughly $16 billion that banks made in 2009.115 Card-issuing banks
will want to make up for the lost revenue in other ways. Not only are
the banks making less in interchange fees than they made in previous years, but these fees may not fully cover the cost per transaction,
because the Federal Reserve used a narrow definition of which costs
to consider when determining the cap.116
Card issuers can make up for their lost revenue by: (1) adding fees
in other places, (2) cutting back on rewards programs, (3) imposing
an annual fee, (4) raising other fees, or (5) lowering interest rates on
deposits.117 However, the card issuers have to be careful with how
they generate this revenue because they do not want to upset the
careful balance of this two-sided market.118
Card issuers will have a more difficult time imposing other nonhidden fees.119 For example, on September 29, 2011, Bank of America
announced a $5 monthly fee for debit cards