* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download 2. Sergi - Center for Climate and Energy Decision Making
Climate change and poverty wikipedia , lookup
Economics of global warming wikipedia , lookup
2009 United Nations Climate Change Conference wikipedia , lookup
Open energy system models wikipedia , lookup
Energiewende in Germany wikipedia , lookup
Years of Living Dangerously wikipedia , lookup
Politics of global warming wikipedia , lookup
Climate change mitigation wikipedia , lookup
Economics of climate change mitigation wikipedia , lookup
IPCC Fourth Assessment Report wikipedia , lookup
German Climate Action Plan 2050 wikipedia , lookup
Low-carbon economy wikipedia , lookup
Climate change in Canada wikipedia , lookup
Carbon Pollution Reduction Scheme wikipedia , lookup
Business action on climate change wikipedia , lookup
Mitigation of global warming in Australia wikipedia , lookup
Understanding public perceptions of energy tradeoffs in climate, health, and economic cost Brian Sergi Inês Azevedo Alex Davis CEDM Annual Meeting May 23, 2016 Motivation • Increasing agreement on need for climate action • What tradeoffs are individuals willing to make in order to get there? 2 Previous work on tradeoff perceptions • Individuals respond more strongly to attributes of energy use than to source (Ansolabehere & Konisky, 2014) • Individuals willing to make tradeoffs analysis in energy decisions (Fleishman-Mayer et. al., 2014) • Health frames can motivate changes to energy use more than economics (Asensio & Delmas, 2014) 3 Research questions • How do individuals make tradeoffs across the different attributes of electricity generation? – climate change – health related air pollution – economic costs (electricity bills) • What is the relative effect of providing climate change and health information when making these tradeoffs? 4 Discrete choice survey • Individuals respond to 16 comparisons of discrete electricity “futures” with different attribute levels • Well-established method in marketing, transportation research (Train, 2009) • Emerging method in the energy & environment space: – – – – – 5 Climate change and energy security (Longo et. al., 2008) Estimating implicit discount rates for lighting (Min et. al., 2014) Preferences for electric vehicles (Helveston et. al., 2015) Energy efficiency (Davis & Metcalf, 2014) Renewable energy and electricity bills in Germany (Kaenzig, 2013) Example choice screen 6 Example choice screen Electricity portfolio – ways of meeting a state’s generation needs. Levels: five “representative” scenarios -- coal (41%) (baseline) -- renewables (42%) -- natural gas (56%) -- nuclear (50%) -- efficiency (14%) 7 Example choice screen Climate change related emissions – change in annual CO2 emissions from baseline i.e. current emissions levels (as percentages). 8 Levels: -- 70% decrease -- 30% decrease -- no change -- 30% increase -- 70% increase Example choice screen Health related air pollution – change in annual SO2 emissions from baseline i.e. current emissions levels (as percentages). Levels: -- 70% decrease -- 30% decrease -- no change -- 30% increase -- 70% increase 9 Example choice screen Monthly electricity bill – change in monthly electricity bill levels for consumers from baseline i.e. individuals’ current bill payments (as percentages). Levels: -- 20% decrease -- 10% decrease -- no change -- 10% increase -- 20% increase 10 Effect of emissions information • Randomized controlled trial with different emissions attributes shown in the task. • Respondents see… – – – – – Group 1: all four attributes (portfolio, bill, CO2, and SO2) Group 2: portfolio, bill, and CO2 only (no information on SO2) Group 3: portfolio, bill, and SO2 only (no information on CO2) Group 4: portfolio and bill only (no information on CO2 or SO2) Group 5: all attributes + monetized damages for CO2 and SO2 • CO2 – social cost of carbon of $40 per ton • SO2 – state averaged marginal damage values from AP2 (Muller, 2014) 11 Example choice screen No SO2 emissions information 12 Example choice screen 13 Modeling • Random utility mixed logit model (Train, 2009) – Model the probability that respondents pick any given scenario conditional on its attributes and those of the alternative – Respondent is utility maximizer – Estimate heterogeneous preferences for emissions and bills 14 Survey demographics (N = 1006 from Amazon Mechanical Turk) 15 Results • Probability of support (conditional on attributes) • Willingness to pay • Not included – Heterogeneity in responses by state, demographics – Verification and consistency checks 16 Probability of support …more likely to support renewables if the are informed about the emissions benefits (even if there are increased costs) Respondents less likely to support renewables if they imply higher electricity bills, but… 17 Conditional probability Support for renewables falls if emissions benefits do not manifest in groups with emissions info Increased support when climate and health emissions benefits are presented (even in the face of higher bills) 18 Willingness-to-pay Implicit WTP in $ per ton of emissions reduced: $100-130 per ton of CO2 $60,000-110,000 per ton of SO2 19 Results summary • Preferences for lower bills, emissions – Outcomes (monthly bill, CO2, SO2) seem to matter more than means (portfolio of sources) • Climate & health emissions reductions on similar level – Reductions in both pollutants slightly increases support • Limitations of stated choice studies – Hypothetical choices, survey design can affect results (Louviere, 2006) – Cognitive biases in stated preference preference studies (Fischhoff, 2005) 20 Policy implications • Technology “neutral” policies for emissions reductions? – More information on portfolio implications may also be needed • Communicate information on emissions reductions, particularly health information – EPA Clean Power Plan promoted as a CO2 policy but with massive anticipated health “co-benefits” (EPA, 2014) • Consider co-optimizing climate mitigation policies across multiple health and climate objectives – Many possible ways to achieve CO2 reductions (Driscoll et. al., 2015) – Those with health benefits likely to gain more support 21 Acknowledgements • Funding sources: CEDM, Steinbrenner Institute, CBDR small grants program • Special thanks to: – My advisors Inês Azevedo & Alex Davis – EPP colleagues – Pre-testers and survey participants 22 Backup slides 23 Heterogeneity by state CO2 and SO2 random effects plotted for respondents seeing damage information 24 Political party 25 Probability of Support 26 Future work • Expand on modeling technique • Run additional surveys – Nationally representative sample – Local sample (in-depth interviews) – Compare to other regions (e.g. China) • Repeat experiment using different structures – Change attribute ranges, include other attributes (employment) 27 Electricity portfolios 28 Electricity portfolios Coal portfolio Renewables portfolio 29 Natural gas portfolio Nuclear portfolio Efficiency portfolio Choice attributes & levels 30 Efficiency incorporation • Step 1: Current cumulative annual efficiency savings by state from the EPA – Input to Integrated Planning Model analysis for Clean Power Plan (Technical Support Document 2014) – Top-down engineering estimates – State based savings estimated to range from 0-2.2% annually • Step 2: Add EE savings to total generation to calculate total “demand • Step 3: Recalculate percentages using this total demand 31 Energy mix by state 32 Attribute levels and format 33 Monetized emissions damages • What is the effect of providing information on monetized emissions damages on CO2 and SO2 tradeoffs? • Monetary damages for attribute levels for each state – emissions x marginal damages (per ton) = total damages CO2 34 SO2 Well mixed global pollutant – location of emissions does not affect marginal damage Marginal damage dependent on population, meteorology, etc. – spatial heterogeneity by state Social cost of carbon – Interagency Working Group mean value estimate (3% discount rate) of $40/ton AP2 model (Muller, 2014) – state averages of marginal damage (mortality & morbidity only) Damage calculations • Step 1: State emissions data for 2014 (EPA CEMS) • Step 2: Estimate per ton damages – CO2 social cost of carbon • $40 in $2014 (average 2015 value given 3% discount) • (IWGSCC Technical Support Document – SO2 AP2 model • State averaged values • Step 3: emissions x per ton damages = total damages – Approach by Muller 2014, EPA, and others 35 CEMS data (CO2) 36 CEMS data (SO2) 37 State emissions • TX – 226 million tons CO2 in 2013 from the electric power sector alone – 340 thousand tons SO2, but marginal damages are much lower – ~35% coal and 40% natural gas • PA – 100 million tons CO2, 260 thousand tons SO2 • WV – 7 million tons CO2, 90 thousand tons SO2 • CA – 44 million tons CO2, 3 thousand tons SO2 Source: EPA CEMS data & “State Energy CO2 Emissions” (http://www3.epa.gov/statelocalclimate/resources/state_energyco2inv.html) 38 Social cost of carbon Taken from EPA: http://www3.epa.gov/climatechange/EPAactivities/economics/scc.html 39 AP2 Integrated assessment model for valuing emissions damages. Source: (Muller, 2015) 40 41 AP2 per ton SO2 damages Source: Muller (2014) 42 AP2 state averages • AP2 marginal damages by county averaged by state – Counties weighted by current share of state emissions (CEMS) 43 East Coast coal states: OH, PA, IN, KY Source: author calculations using AP2 and CEMS data 44 Relatively high CO2 costs Relatively high SO2 costs 45 Monetized emissions damages • Electricity dominates social damages of emissions from energy production (Jaramillo & Muller, 2016) – Mortality accounts for 95% these costs • SO2 accounts for around 70% of all social costs from the power sector (Heo, 2015) • 2005 2011 total electricity sectors damages decreased from $154 billion to $100 billion (Jaramillo & Muller, 2016) – SO2 emissions ~ 5 million tons in 2011 (3 million tons in 2014) 46 Electricity sector damages Source: Jaramillo & Muller, 2016 47 * Estimate for 2011. Adjusted to $2014 Also includes damages from VOCs, PM2.5, NOx, NH3. 48 EPA use of damages 49 Respondent checks • Attention checks – 2 tasks with “dominated” alternatives (second & last task) • Transitivity check – 3 tasks with related alternatives (A, B, C) • Linearity check – 2 tasks with different scenarios but identical differences between choices 50 Attention checks • 2 question with dominated alternatives (2nd and last task) – Portfolio question – “Coal, natural gas, renewables, nuclear, and energy efficiency are all electricity sources or reductions considered in this survey.” – Bill question – “Assuming the amount of electricity you use does not change, higher electricity prices would lower your monthly electricity bill.” 51 Attention checks • 2 tasks with dominated alternatives (2nd and last task) 52 Attention checks 53 Transitivity checks • Series of 3 tasks – Choice 1: A vs. B – Choice 2: B vs. C – Choice 3: C vs. A 54 Transitivity checks 55 Transitivity checks • 3 choices 8 possible choice combinations • ABC, ABA, ACC, ACA, BBC, BBA, BCC, BCA • 6 of these 8 consistent with transitive preferences – ABC, BCA not consistent with transitivity – 75% chance of randomly picking choices consistent with transitivity 56 A B C Portfolio Coal Renew. Gas CO2 0 -30 70 SO2 0 -30 70 Bill 0 10 20 57 Transitivity checks • Some thoughts – Relatively small emissions reductions via renewables may not be enough to entice coal lovers – For people with renewables as a first preference, gas preferred to coal (group 1) unless it brings high emissions (other groups) • Suggests renewable people support emissions reductions, gas as a “bridge” strategy • Will support coal if cheaper + has lower emissions – “ABA” combination consistent with choosing on bill alone 58 Linearity check • Two tasks with different baseline levels but identical differences between the two options • Linear preferences if respondents choose AB or BA – 50% chance of responding this way by just guessing 59 Linearity checks Group Linearity (%) Significant difference by group 1 2 3 * 1 90 - * 2 75 * - 3 68 * 4 85 5 78 * * 4 * * - * * - * Significant differences tested using Welch Two Sample Test (confidence level 95%) 60 5 * - Passed all checks • Probability of … – Passing attention checks 25% – Transitive preferences 75% – Linear preferences 50% – All consistency checks ~ 9 % • Actual success rates All 61 Group1 Group 2 Group 3 Group 4 Group 5 0.82 0.71 0.64 0.82 0.74 Conditional probability check Comparison of predicted and actual probability estimates for 1 task 62 Group1 Group 2 Group 3 Group 4 Group 5 Predicted probability 0.73 0.71 0.72 0.79 0.76 Actual probability 0.63 0.64 0.63 0.76 0.73 Number of respondents 204 192 205 221 184 Pilot test demographics *Data 63 from U.S. Census, 2015 Gallup Polls Pilot test demographics 64 Pilot test results 65 Sampling by state 66 Sampling by state 67 Demographics by group 68 Monthly electricity bill 69 Survey completion time 70 Choice set assumptions • Alternatives must be… – Mutually exclusive • Specific generation levels of portfolios other attribute levels – Exhaustive • “representative” extreme values • non-included fuels (e.g. biomass, geothermal) not likely to be large sources of generations – Finite number of alternatives • Infinite number of portfolios, simplified by representative scenarios 71 Choice modeling • Logit model – Random utility model: “Observed” component 72 “Unobserved” component Observed component of utility • Often assumed to be a additively separable linear function – X = levels of the attributes 73 Unobserved component of utility • Assumed iid Gumbel distribution (extreme value) • Standard Gumbel: μ = 0, β = 1 74 Gumbel distribution 75 Difference in two errors ~ logistic • Only differences in utility matter – Gumbel assumption does not affect utility estimation so long as unobserved components have the same mean • Difference of two Gumbel random variables is distributed logistically 76 Logistic distribution • Similar to assumption of normally distributed errors but with fatter tails – Allows for more variability (Train, 2009) • Outcome of a well specified model – Unobserved error should be “white noise” • Model used in logistic regression/binary logit models 77 Choice modeling • Logit model – Random utility model: “Observed” component 78 “Unobserved” component Choice modeling • Logit model – Assumed that respondents are utility maximizers – Choices made based on difference in utility (Train, 2009) • Choose left if Ui,left > Ui,right – Since we don’t observe Uij, need to model probabilistically – Coefficients for Vij estimated using Maximum Likelihood Estimation 79 Logit Assumptions • Multinomial logit – Cannot represent random taste heterogeneity – Independence of Irrelevant Alternatives, – Cannot deal unobserved correlation over time • Mixed logit – Does not require these assumptions – Only requires choice proper set construction + error distribution 80 List of models Multinomial logit Mixed logit Base model Individual random effects Base model, no intercept State random effects State fixed effects (large states only) Individual + state random effects Single model with dummies for group Individual + state random effects (SO2 & CO2 at state level) SO2 indicator Individual + state random effects, no intercept (SO2 & CO2 at state level) Individual random effects + SO2 indicator Error variance • 2 sources of bias: – Different scale parameter (bias downward) – Traditional omitted variable bias for any attribute based on covariance (bias up or down, depending on perception) • Scale parameter issue neutralized by taking ratio of coefficients (e.g. WTP) 82 Nonprice Incentives (Asensio & Delmas, 2014) • “Environment and health-based information treatments motivated 8% savings vs. control” – Families with children achieved ~19% energy savings – Monetary group increased consumption ~ 3% (rebound) – Did not address long-term persistence Table S1. Treatment Messages 83 Group Monetary Savings Group Treatment Message “Last week, you used 66% more/less electricity than your efficient neighbors. In one year, this will cost you (you are saving) $34 dollars extra.”* Health Group “Last week, you used 66% more/less electricity than your efficient neighbors. You are adding/avoiding 610 pounds of air pollutants which contribute to health impacts such as childhood asthma and cancer.”* Control Group None. * ‘Efficient neighbors’ in this context means households in the top 10th percentile of household weekly average kWh consumption (households with the lowest electricity use) for similar size apartments in the community. What motivates public opinion on energy? • Cultural model – Cultural cognition/motivated reasoning; e.g. community identity (Wildavsky, 1981), post-materialism (Ingleheart 1990) – Demographic identities and social groupings (race, gender, income, etc.) – Some challenge to the useful explanatory power of demographic variables (Smith 2002) • Psychological model – Dread and nuclear power (Slovic, Fischhoff, and Lichtenstein 1981) • Political model – Institutional trust and perceived risk (Slovic 1993) – Political affiliation and attitudes; e.g. Republicans and nuclear power (Greenberg and Truelove 2011) • Other personal values and norms 84 What motivates public opinion on energy? • Consumer model – People evaluate energy choices based on the attributes of the options (Ansolabehere and Konisky 2014) – Individuals think about options in terms of tradeoffs between attributes (Fleishman-Mayer et. al. 2010) – Most important attributes: economic cost and environment impact (Ansolabehere and Konisky 2014) • In this model, two key elements: – The public’s knowledge of different technologies’ attributes – The public’s valuation of the tradeoffs between different options 85 Discrete choice: pros & cons Pros Cons Established theory of choice behavior Random utility theory assumes articulated values Complex choices decomposed Hypothetical Control of attribute levels Subject to cognitive biases (e.g. scope insensitivity, anchoring, prospect theory, inattention, etc.) May not include all relevant factors Error variance assumption not theoretically justified 86 Mixed Logit 87 Multinomial Logit 88 Multinomial Logit (dropped data) 89 Moulton Factors 90 Random effects • Standard deviations for state level random effects 91 Equations Logit definition Conditional probability (at baseline) 92 Odds ratio calculation 93 Odds ratios 94 Conditional probabilities 95 Group 5 results 96 Group 4 results 97 Willingness-to-pay ratio • Derived from ratio of attribute and price (i.e. bill coefficients) CO2 SO2 Stated CO2 WTP* No emissions info - - 8% CO2 info 18% (14-22) - 10% SO2 info - CO2 & SO2 info (no damages) 16% (13-20) 21% (17-27) 15% (12-19) 9% 12% CO2 & SO2 info 14% 12% 11% (with damages) (11-130) (10-160) WTP as percentage increase in monthly electricity bill * Respondents were directly asked how much they would be willing to pay for 30% CO2 reductions in percent increase of electricity bills 98 Implicit per ton WTP calculation • Step 1: modeled WTP – % bill for 30% reduction in CO2 / SO2 • Step 2: multiply by U.S. average monthly electricity bill – Survey average: $124 – EIA average: $114 • Step 3: multiply by U.S. population (national WTP) – Could also use number of households • Step 4: divide by tons emissions reduced 99 Implicit per ton WTP results Calculated results Results from other studies 100 In each screen we will ask you to choose between 2 energy scenarios just like the ones you see here. Each scenario is described by 4 characteristics. They are: • Electricity portfolio • Climate change related emissions • Health related air pollution • Monthly electricity bill The next screens will provide indepth information on these characteristics. Electricity portfolio shows how much of your state's electricity generation would come from coal, natural gas, nuclear power, and renewable energy (renewables include wind, solar, and hydropower). In addition, energy efficiency programs (using more efficient appliances or weatherizing houses, for example) can help reduce the total electricity needed, and these savings are included in the electricity portfolio. Climate change related emissions shows the percentage change in yearly carbon dioxide (CO2) emissions compared to today’s levels. CO2 is a greenhouse gas that contributes to climate change, and more CO2 will result in increased average global temperature, more intense storms, more floods and droughts, and rising sea levels. These effects occur on a global scale. An estimate of the additional costs to the planet from these climate change effects is also provided. Health related air pollution shows the percentage change in yearly sulfur dioxide (SO2) emissions compared to today’s levels. SO2 can form small particles that can get into the lungs, meaning that people who live in areas with higher SO2 can have an increased risk of heart attacks, asthma, and other respiratory problems. These effects are most severe in the areas near where the pollution is emitted. An estimate of the additional costs to polluted areas from these health effects is also provided. Monthly electricity bill shows the percentage change in your electricity bill relative to what you pay now. Different scenarios can be more or less expensive and can affect how much you pay for electricity. Throughout the choices you can click on the yellow “Learn more” box to see this information again. Learn& more& Learn& more& Learn& more& Learn& more& You can also click the text above to find out more about the combinations you are seeing. Finally, you can indicate the scenario you prefer using these buttons below. When you’re ready, click the blue arrow below to finish the guided example. References • • • • • • • • • • • Adamowicz, W., Boxall, P., Williams, M., & Louviere, J. J. (1998). Stated preference approaches for measuring passive use values: Choice experiments and contingent valuation. American Journal of Agricultural Economics, 80(1), 64–75. http://doi.org/10.2307/3180269 Alriksson, S., & Öberg, T. (2008). Conjoint Analysis for Environmental Evaluation: A review of methods and applications, 15(3), 244–257. Ansolabehere, S., & Konisky, D. M. (2014). Cheap and clean: how Americans think about energy in the age of global warming. MIT Press. Apt, J., & Fischhoff, B. (2006). Power and People. The Electricity Journal, 19(9), 17–25. http://doi.org/10.1016/j.tej.2006.09.008 Aravena, C., Martinsson, P., & Scarpa, R. (2014). Does money talk? - The effect of a monetary attribute on the marginal values in a choice experiment. Energy Economics, 44, 483–491. http://doi.org/10.1016/j.eneco.2014.02.017 Asensio, O. I., & Delmas, M. A. (2015). Nonprice incentives and energy conservation. Proceedings of the National Academy of Sciences, 112(6), E510–E515. http://doi.org/10.1073/pnas.1401880112 Ben-Akiva, M., & Morikawa, T. (1990). Estimation of switching models from revealed preferences and stated intentions. Transportation Research Part A: Policy and Practice, 24(6). Bergmann, A., Hanley, N., & Wright, R. (2006). Valuing the attributes of renewable energy investments. Energy Policy, 34(9), 1004–1014. http://doi.org/10.1016/j.enpol.2004.08.035 Berinsky, A. J., Margolis, M., & Sances, M. W. (2014). Separating the shirkers from the workers? Making sure respondents pay attention on internet surveys. American Journal of Political Science, 58(3), 739–753. Buhrmester, M. (2011). Amazon’s Mechanical Turk a new source of inexpensive, yet high-quality, data? … on Psychological Science. Retrieved from http://pps.sagepub.com/content/6/1/3.short Davis, C., & Fisk, J. M. (2014). Energy Abundance or Environmental Worries? Analyzing Public Support for Fracking in the United States. Review of Policy Research, 31(1), 1–16. http://doi.org/10.1111/ropr.12048 Davis, L. W., & Metcalf, G. E. (2014). Does Better Information Lead to Better Choices? Evidence from EnergyEfficiency Labels. 108 References • • • • • • • • • • Douglas, M., & Wildavsky, A. (1983). Risk and culture: An essay on the selection of technological and environmental dangers. Univ of California Press. Fann, N., Baker, K. R., & Fulcher, C. M. (2012). Characterizing the PM2.5-related health benefits of emission reductions for 17 industrial, area and mobile emission sectors across the U.S. Environment International, 49, 141– 151. http://doi.org/10.1016/j.envint.2012.08.017 Fann, N., Fulcher, C. M., & Hubbell, B. J. (2009). The influence of location, source, and emission type in estimates of the human health benefits of reducing a ton of air pollution. Air Quality, Atmosphere and Health, 2(3), 169–176. http://doi.org/10.1007/s11869-009-0044-0 Fischhoff, B. (1991). Value elicitation: Is there anything in there? American Psychologist. http://doi.org/10.1037/0003-066X.46.8.835 Fischhoff, B. (2005). Cognitive Processes in Stated Preference Methods. In K.-G. Mäler & J. R. Vincent (Eds.), Handbook of Environmental Economics (Vol. 2, pp. 938–964). Elsevier. http://doi.org/10.1016/S15740099(05)02018-8 Fischhoff, B., Slovic, P., & Lichtenstein, S. (1983). “The Public” vs. “The Experts”: Perceived vs. actual disagreements about risks of nuclear power. In Analysis of actual vs. perceived risks. Fleishman-Mayer, L., Bruine de Bruin, W., & Morgan, M. G. (2014). Informed public choices for low-carbon electricity portfolios using a computer decision tool. Environmental Science & Technology, 48(7), 3640–8. http://doi.org/10.1021/es403473x Forswall, C. D., & Higgins, K. E. (2005). Clean Air Act Implementation in Houston : An Historical Perspective, (February), 1–95. Freudenburg, W. R., & Gramling, R. (1994). Oil in troubled waters: Perceptions, politics, and the battle over offshore drilling. SUNY Press. Goodman, J. K., Cryder, C. E., & Cheema, A. (2013). Data Collection in a Flat World: The Strengths and Weaknesses of Mechanical Turk Samples. Journal of Behavioral Decision Making, 26(3), 213–224. http://doi.org/10.1002/bdm.1753 109 References • • • • • • • • • • Gosling, S., & Vazire, S. (2004). Should we trust web-based studies? A comparative analysis of six preconceptions about internet questionnaires. American …. Retrieved from http://psycnet.apa.org/journals/amp/59/2/93/ Helveston, J. P., Liu, Y., Feit, E. M., Fuchs, E., Klampfl, E., & Michalek, J. J. (2015). Will subsidies drive electric vehicle adoption? Measuring consumer preferences in the U.S. and China. Transportation Research Part A: Policy and Practice, 73, 96–112. http://doi.org/10.1016/j.tra.2015.01.002 Heo, J. (2015). Evaluation of Air Quality Impacts on Society: Methods and Application. Carnegie Mellon University. Hess, S., & Daly, A. (2010). Choice Modelling: the state of the art and the state of practice. In Proceedings from the Inaugural International Choice Modelling Conference (pp. 1–639). Hoyos, D. (2010). The state of the art of environmental valuation with discrete choice experiments. Ecological Economics, 69(8), 1595–1603. http://doi.org/10.1016/j.ecolecon.2010.04.011 Interagency Working Group on Social Cost of Carbon United States Government. (2015). Technical Support Document: Technical Update of the Social Cost of Carbon for Regulatory Impact Analysis Under Executive Order 12866 - July 2015 Revision, (July 2015), 1–21. Retrieved from http://www3.epa.gov/climatechange/EPAactivities/economics/scc.html Intergovernmental Panel on Climate Change. (2014). Climate Change 2014 Synthesis Report. Fifth Assessment Report (AR5). Kaenzig, J., Heinzle, S. L., & Wüstenhagen, R. (2013). Whatever the customer wants, the customer gets? Exploring the gap between consumer preferences and default electricity products in Germany. Energy Policy, 53, 311–322. http://doi.org/10.1016/j.enpol.2012.10.061 Levy, J. I., Baxter, L. K., & Schwartz, J. (2009). Uncertainty and Variability in Health-Related Damages from CoalFired Power Plants in the United States. Risk Analysis, 29(7), 1000–1014. http://doi.org/10.1111/j.15396924.2009.01227.x Longo, A., Markandya, A., & Petrucci, M. (2008). The internalization of externalities in the production of electricity: Willingness to pay for the attributes of a policy for renewable energy. Ecological Economics, 67(1), 140–152. http://doi.org/10.1016/j.ecolecon.2007.12.006 110 References • • • • • • • • • • • • Louviere, J. J. (2006). What you don’t know might hurt you: Some unresolved issues in the design and analysis of discrete choice experiments. Environmental and Resource Economics, 34(1), 173–188. http://doi.org/10.1007/s10640-005-4817-0 Louviere, J. J. (2014). The Role of the Scale Parameter in the Estimation and Comparison of Multinomial Logit Models. American Marketing Association, 30(3), 305–314. Louviere, J. J., Flynn, T. N., & Carson, R. T. (2010). Discrete choice experiments are not conjoint analysis. Journal of Choice Modelling, 3(3), 57–72. http://doi.org/10.1016/S1755-5345(13)70014-9 Louviere, J. J., Hensher, D. A., & Swait, J. D. (2000). Stated choice methods: analysis and applications. Cambridge University Press. Luce, R. D. (1959). Individual choice behavior: A theoretical analysis. New York: John Wiley. Luce, R. D., & Tukey, J. W. (1964). Simultaneous conjoint measurement: A new type of fundamental measurement. Journal of Mathematical Psychology, 1(1), 1–27. http://doi.org/10.1016/0022-2496(64)90015-X Magidson, J., & Vermunt, J. (2007). Removing the scale factor confound in multinomial logit choice models to obtain better estimates of preference. Sawtooth Software Conference, 1–18. Retrieved from http://www.sawtoothsoftware.com/downloadPDF.php?file=2007Proceedings.pdf#page=147 Mason, W., & Suri, S. (2012). Conducting behavioral research on Amazon’s Mechanical Turk. Behavior Research Methods, 44(1), 1–23. http://doi.org/10.3758/s13428-011-0124-6 McCaffrey, D. P. (1991). The Politics of Nuclear Power: A History of the Shoreham Nuclear Power Plant (Vol. 5). Springer Science & Business Media. McFadden, D. (1974). Conditional logit analysis of qualitative choice behavior. In P. Zarembka (Ed.), Frontiers in Econometrics (pp. 105–142). Academic Press, New York. McFadden, D., & Train, K. E. (2000). Mixed MNL models for discrete response. Journal of Applied Econometrics, 15(5), 447–470. http://doi.org/10.1002/1099-1255(200009/10)15:5<447::AID-JAE570>3.0.CO;2-1 Min, J., Azevedo, I. L., Michalek, J., & de Bruin, W. B. (2014). Labeling energy cost on light bulbs lowers implicit discount rates. Ecological Economics, 97, 42–50. http://doi.org/10.1016/j.ecolecon.2013.10.015 111 References • • • • • • • • • • • • • • Muller, N. Z. (2015). Environmental Benefit Cost Analysis and The National Accounts. In USAEE Conference Proceedings. Muller, N. Z., & Mendelsohn, R. (2007). Measuring the damages of air pollution in the United States. Journal of Environmental Economics and Management, 54(1), 1–14. http://doi.org/10.1016/j.jeem.2006.12.002 Muller, N. Z., & Mendelsohn, R. (2009). Efficient Pollution Regulation: Getting the Prices Right. American Economic Review, 99(5), 1714–1739. http://doi.org/10.1257/aer.99.5.1714 Muller, N. Z., Mendelsohn, R., & Nordhaus, W. (2010). Environmental Accounting for Pollution in the United States Economy ., 1–44. National Research Council. (2010). Hidden Costs of Energy: Unpriced Consequences of Energy Production and Use. Washington, DC. Paolacci, G., & Chandler, J. (2014). Inside the Turk: Understanding Mechanical Turk as a Participant Pool. Current Directions in Psychological Science, 23(3), 184–188. http://doi.org/10.1177/0963721414531598 Phadke, R. (2010). Steel forests or smoke stacks: the politics of visualisation in the Cape Wind controversy. Environmental Politics. Retrieved from http://www.tandfonline.com/doi/abs/10.1080/09644010903396051 Pope, C. A., Burnett, R. T., Thun, M. J., Calle, E. E., Krewski, D., & Thurston, G. D. (2002). Lung Cancer, Cardiopulmonary Mortality, and Long-term Exposure to Fine Particulate Air Pollution, 287(9). Pope, C. A., & Dockery, D. W. (2006). Health effects of fine particulate air pollution: lines that connect. Journal of the Air & Waste Management Association (1995), 56(November), 709–742. http://doi.org/10.1080/10473289.2006.10464485 Regenwetter, M., Dana, J., & Davis-Stober, C. P. (2010). Testing transitivity of preferences on two-alternative forced choice data. Frontiers in Psychology, 1(DEC), 1–15. http://doi.org/10.3389/fpsyg.2010.00148 Slovic, P. (1987). Perception of risk. Science, 236(4799), 280–285. Smith, E. R. A. N. (2001). Energy, the environment, and public opinion. Rowman & Littlefield Publishers. Train, K. E. (2009). Discrete choice methods with simulation. Cambridge University Press. Train, K. E., & Weeks, M. (2005). Discrete choice models in preference space and willingness-to-pay space. Springer. 112 References • • • • • Train, K. E., & Wilson, W. W. (2008). Estimation on stated-preference experiments constructed from revealedpreference choices. Transportation Research Part B: Methodological, 42(3), 191–203. http://doi.org/10.1016/j.trb.2007.04.012 U.S. Energy Information Administration. (2015). Electricity Data Browser. U.S. Environmental Protection Agency. (2014). Regulatory Impact Analysis for the Proposed Carbon Pollution Guidelines for Existing Power Plants and Emission Standards for Modified and Reconstructed Power Plants. U.S. Environmental Protection Agency. (2015a). Regulatory Impact Analysis for the Clean Power Plan Final Rule. Washington, DC. Retrieved from http://www.epa.gov/airquality/cpp/cpp-final-rule-ria.pdf U.S. Environmental Protection Agency. (2015b). U.S. Greenhouse Gas Inventory Report: 1990-2013. Retrieved from http://www3.epa.gov/climatechange/Downloads/ghgemissions/US-GHG-Inventory-2015-Chapter-ExecutiveSummary.pdf 113