UUnniivveerrssiittyy ooff CChhiiccaaggoo LLaaww SScchhooooll CChhiiccaaggoo UUnnbboouunndd Coase-Sandor Working Paper Series in Law and Coase-Sandor Institute for Law and Economics Economics 2001 TThhee AArriitthhmmeettiicc ooff AArrsseenniicc Cass R. Sunstein [email protected] Follow this and additional works at: https://chicagounbound.uchicago.edu/law_and_economics Part of the Law Commons RReeccoommmmeennddeedd CCiittaattiioonn Cass R. Sunstein, "The Arithmetic of Arsenic" (John M. Olin Program in Law and Economics Working Paper No. 135, 2001). This Working Paper is brought to you for free and open access by the Coase-Sandor Institute for Law and Economics at Chicago Unbound. It has been accepted for inclusion in Coase-Sandor Working Paper Series in Law and Economics by an authorized administrator of Chicago Unbound. For more information, please contact [email protected]. C H I C A G O JOHN M. OLIN LAW & ECONOMICS WORKING PAPER NO. 135 (2D SERIES) The Arithmetic of Arsenic Cass R. Sunstein THE LAW SCHOOL THE UNIVERSITY OF CHICAGO This paper can be downloaded without charge at: The Chicago Working Paper Series Index: http://www.law.uchicago.edu/Lawecon/index.html The Social Science Research Network Electronic Paper Collection: http://papers.ssrn.com/paper.taf?abstract_id=285171 Preliminary draft 8/31/01 All rights reserved The Arithmetic of Arsenic Cass R. Sunstein* Abstract What does cost-benefit mean, or do, in actual practice? When agencies are engaging in cost-benefit balancing, what are the interactions among law, science, and economics? This Article attempts to answer that question by exploring, in some detail, the controversy over EPA’s proposed regulation of arsenic in drinking water. The largest finding is that science often can produce only “benefit ranges,” and wide ones at that. With reasonable assumptions based on the existing science data, the proposed arsenic regulation can be projected to save as few as 0 lives and as many as 112. With reasonable assumptions, the monetized benefits of the regulation can range from $0 to $560 million. In these circumstances, there is no obviously right decision for government agencies to make. These points have numerous implications for lawyers and courts, suggesting the ease of bringing legal challenges, on grounds specified here, and the importance of judicial deference in the face of scientific uncertainty. There are also policy implications. Agencies should be given the authority to issue more targeted, cost-effective regulations. They should also be required to accompany the cost-benefit analysis with an effort to identify the winners and losers, so as to see if poor people are mostly hurt or mostly helped. Americans may disagree about a lot of things, but arsenic isn’t one of them. When you turn on the kitchen sink, you ought to be able to drink what comes out, without worrying about being poisoned.1 “What we know is a drop, what we do not know—an ocean” (Isaac Newton). In spite of significant gains in knowledge, we are still moving mainly in the dark when dealing with the quantitative importance of risk factors in *Karl N. Llewellyn Distinguished Service Professor of Jurisprudence, Law School and Department of Political Science, University of Chicago. I am grateful to Laura Warren for outstanding research assistance and to Jonathan Baron, Robert Hahn, Lisa Heinzerling, Eric Posner, Richard Posner, and participants in a work-in-progress lunch at the Universty of Chicago Law School for valuable suggestions on a previous draft. 1 Rep. David Bonior, quoted in the Chicago Tribune, July 28, 2001, p. 1. chemical carcinogenesis, the mechanisms of action of chemical carcinogens, and hence their detection and the assessment of their risks to human health. The basic understanding . . . is still missing.”2 Because the shape of the dose-response curve in the low-dose region cannot be verified by measurement, there is no means to determine which shape is correct. . . . [W]hen modeling the risks associated with lower doses, the dose/risk range in which regulatory agencies and risk assessors are most frequently interested, there is a wide divergence in the risk projected by [different models, all of which fit existing evidence.] . . . In fact . . . the risks predicted by these . . . models produce a 70,000-fold variation in the predicted response.3 Additional epidemiological evaluations are needed to characterize the dose-response relationship for arsenic-associated cancer and noncancer end points, especially at low doses. Such studies are of critical importance for improving the scientific validity of risk assessment.4 Anyone who’s read an Agatha Christie mystery knows that arsenic is a poison.5 Within the past two decades, cost-benefit analysis (CBA) has become one of the most widely discussed topics in all of regulatory law.6 Much of the discussion is occurring within the three branches of government. The Office of Management and Budget (OMB) has overseen a series of executive orders calling for cost-benefit balancing,7 and OMB has attempted to give concrete guidance for agencies to follow.8 Courts have adopted a series of cost-benefit default principles, authorizing agencies to engage in cost-benefit balancing unless 2Toxicology (Hans Marquardt et al. eds. 1999). 3Phillip L. Williams et al., Principles of Toxicology 456 (2000). 4National Research Council, Arsenic in Drinking Water 3 (1999). 5Cong. Rec. H4751 (July 27, 2001) (Rep. Anna Eshoo). 6For an overview, see Cost-Benefit Analysis: Legal, Philosophical, and Economic Perspectives (Mathew Adler & Eric Posner eds., 2001). 7See, e.g., Executive Order 12291, 46 Fed. Reg. 13193 (1980); Executive Order 12498, 50 Fed. Reg. 1036 (1985); Executive Order 12866, 58 Fed. Reg. 51735 (1993). 8See best practice guidelines, available at http://www.whitehouse.gov/omb/inforeg/ index.html. A useful overview is Richard B. Stewart, A New Generation of Environmental Regulation?, 29 Capital U. L. Rev 21, 29–34 (2001). 2 Congress has required otherwise.9 Congress itself has shown considerable interest in requiring agencies to compile information on the costs and benefits of regulation. 10 At the same time, there has been renewed academic interest in CBA, exploring the technique from a variety of perspectives.11 In all of these contexts, however, the discussion has tended to be quite abstract. Within the legal culture, there has been little exploration of what CBA specifically entails, or of how it might be used or improved by agencies.12 To date, there appears to be no sustained investigation of any regulation in which CBA proved pivotal to the outcome. In this Article, I hope to begin to fill the gap. I do so by exploring one of the most contested early decisions of the Environmental Protection Agency (EPA) under President George W. Bush: the suspension of the EPA regulation of arsenic in drinking water.13 Much of the contest over that decision has involved a debate about the relevant costs and benefits. As we will see, it is possible to draw a range of general lessons from the arsenic controversy. My principal finding is simple: Sometimes the best that can be done is to specify an exceedingly wide “benefits range,” one that does not do a great deal to discipline judgment. Much of the discussion will be devoted to establishing this insufficiently appreciated point, with some effort to specify the judgments that must be made in order both to identify the health benefits and to monetize them. As a result of this finding, it would be wrong to have confidence that the EPA’s proposed rule, in the Clinton Administration, was either right or wrong.14 At the same time, I offer three more positive suggestions. First, cost-benefit analysis, even with wide ranges, provides an important improvement over the “intuitive toxicology” of ordinary people, in which general “affect” helps to determine judgment.15 This intuitive toxicology can lead people to large blunders in thinking about risk, not excluding the excessive reaction to the Bush Administration’s decision to suspend the arsenic rule issued by the Clinton Administration.16 Second, considerable progress could be made by authorizing 9See Corrosion Proof Fittings v. EPA, 947 F2d 1201 (5th Cir 1991). 10See, e.g., 42 U.S.C. § 300f et seq. (Safe Drinking Water Act). 11See Adler & Posner supra n. 6. 12A superb general discussion in Richard Revesz, Environmental Regulation, Cost-Benefit Analysis, and the Discounting of Human Lives, 99 Col L Rev 941 (1999). 1366 Fed. Reg. 28,342 (May 22, 2001). 14See infra. 15See Paul Slovic, The Perception of Risk 285 (2001). 16Id. 3 EPA both to use market incentives and to target drinking water controls to areas where they would do the most good. Third, the EPA should be required to provide a distributional analysis, showing who, exactly, would be helped and hurt by regulation. In its voluminous materials on the effects of the new arsenic rule, for example, the EPA does not say a word about whether poor people would bear the sometimes significant costs of the regulation. It would be easier to assess the new rule with a clearer sense of the benefited and burdened classes. More particularly, I suggest that an understanding of the arsenic controversy offers seven general lessons. 1. CBA can sometimes produce an illusion of certainty.17 Even where, as in the arsenic case, science has a great deal to offer, the most that the agency can be expected to do may be to specify a range, sometimes a wide range, without assigning probabilities to various “points” along the spectrum. This suggestion should be taken as an attack not on CBA, but on what might be described as the false promise of CBA: the thought that science and economics, taken together, can produce “bottom lines” to be mechanically applied by regulatory agencies. “There is wide recognition among experts—but not necessarily in the public opinion—that current approaches to the regulation of most agents remain judgmental.”18 2. With respect to health benefits, plausible assumptions can lead in dramatically different directions. In the case of arsenic, it would be to conclude that the annual number of lives saved from EPA’s proposed regulation would be as low as 5 or as high as 112—and that the annual monetized benefits of the proposed standard would be as high as $1.2 billion or as low as $10 million! It is worthwhile to pay special attention to the dose-response curve, on which direct information is typically absent; I will make a particular effort to connect the legal and economic issues involved in cost-benefit balancing to the underlying scientific questions. 3. If literate in some basic science and economics, an adroit lawyer, on either side, might mount apparently reasonable challenges to any EPA decision about whether and how to regulate arsenic in drinking water. An industry lawyer should be able to urge, with some force, that any new regulation of arsenic is too severe, because the costs exceed the benefits. An environmental lawyer should be able to urge, with some force, that nearly 17 This point is also pressed in Lisa Heinzerling, Regulatory Costs of Mythic Proportions, 107 Yale L.J. 1981 (1998). 18 Toxicology 1145 (Hans Marquardt et al. eds., 1999). 4 any imaginable regulation of arsenic is too lenient, because the benefits of further regulation would exceed the costs. Both challenges would be plausible for a simple reason: It is easy to identify assumptions that would drive the numbers up or down. Hence one of my principal goals is to provide a kind of primer on how informed lawyers can integrate science, economics, and law in order to challenge regulatory outcomes. 4. In part because of point 3, and in light of an understanding of the scientific and economic complexities, courts should play an exceedingly deferential role in overseeing CBA at the agency level. To say the least, judges are not specialists in the relevant topics, some of which are highly technical, and because good lawyers will be able to raise so many plausible doubts, the best judicial posture is one of deference. In the arsenic case, and in many other contexts, agencies must decide in the midst of considerable scientific uncertainty and on the basis of judgments of value on which reasonable people can differ. If agencies have been both open and reasonable, the judicial role is at an end. It follows, for example, that the Clinton Administration’s arsenic rule, if it had been finally issued and challenged, should have survived judicial review. It also follows that a much less stringent regulation, if chosen by the Bush Administration, should survive judicial review too. The claim for judicial deference, in both cases, is rooted in institutional considerations, and above all a sense of the likely problems of intensive judicial review – not in approval of any particular agency decision. 5. The false precision of CBA is a significant cautionary note, but it should not be taken as a fundamental attack on the method itself, at least if CBA is understood as a way of compiling relevant information. In the arsenic case, an assessment of costs and benefits cannot determine the outcome. But even so, the assessment is indispensable to inform the inquiry and to ensure that discretion is exercised in a way that is transparent rather than opaque. Without some effort to ascertain the effects of regulation, agencies are making a stab in the dark. At the very least, an understanding of the data helps show exactly why the decision about how to regulate arsenic is genuinely difficult – and why, and where, reasonable people might differ. This is itself a significant gain. 6. The Safe Drinking Water Act (SDWA), designed to control pollution in drinking water, has been amended to require cost-benefit balancing, partly in order to permit the EPA to relax regulatory requirements where the benefits are low and the costs are high.19 At the same time, however, 19 42 USC § 300f et seq. (Safe Drinking Water Act). 5 the SDWA continues to have a high degree of rigidity. The EPA is not authorized to impose regulation selectively and in those areas in which regulation would do the most good; it is required to proceed with a uniform, national regulation. The EPA is also forbidden to create trading programs, which might well make best sense for some pollutants. Statutory amendments would be sensible here, especially under a statute dedicated to cost-benefit balancing. Regulatory statutes generally should authorize agencies to target regulations to areas where the benefits exceed the costs, and should also allow agencies to use market incentives where appropriate. 7. It would be extremely valuable to assemble information about the distributional consequences of regulation. The benefits of some regulations are enjoyed disproportionately by people who are poor and members of minority groups.20 The burdens of some regulations are imposed disproportionately on exactly the same groups. To assess the arsenic rule, it would be highly desirable to know whether poor people are mostly helped or mostly hurt. Would they bear high costs? Would the regulation operate as a regressive tax? Unfortunately, the EPA has not answered that question, though it would almost certainly be easy for it to do so. My own preliminary analysis suggests that the most significant financial burdens would be imposed on people with annual incomes well below the median21 -- a point that is certainly relevant to overall evaluation. Existing Executive Orders, calling for CBA, should be amended to require a careful distributional analysis as well.22 This Article comes in several parts. Part I offers a general overview of the movement toward cost-benefit balancing, a movement for which the SWDA stands as the most dramatic legislative endorsement. It also gives a brief description of the public outcry over President Bush’s decision to suspend the regulation, in a way that is intended to fortify the case for CBA. Part II provides a brief outline of the SDWA and of the EPA’s rationale in its regulation of arsenic. Part III explores the very different analysis coming from the American Enterprise Institute-Brookings Joint Center on Regulation. Part IV, in many ways the heart 20See Matthew E. Kahn., The Beneficiaries of Clean Air Act Regulation, 24 Regulation 22 (2001), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=267073. 21See note infra. 22See note supra. There is a brief reference to “distributive impacts” in Executive Order 12966. See also Environmental Protection Agency, Guidelines for Preparing Economic Analyses, 164–70 (2000) (discussing need to explore equitable and distributional considerations, including effects on poor and minority communities). 6 of the Article, shows how apparently reasonable assumptions lead to a dramatically diverse set of benefit numbers, both monetized and nonmonetized. Part V explores how lawyers and courts might respond to the existing data. Part VI discusses the role of policymakers, explaining that agencies should be permitted to issue targeted regulations and to use economic incentives, and that in keeping with its informational functions, CBA should include a description of the expected winners and losers from regulation. 1. Intuitive Toxicology and the Cost-Benefit State A. Arsenic and the Public My principal topic will be the contest over the appropriate analysis of existing data relating to arsenic; but it will be useful to begin with a puzzle. In April 2001, the Bush Administration suspended the Clinton Administration’s arsenic regulation , calling for further study.23 There seems to be little question that of all the controversial environmental actions of the Bush Administration, the suspension of the arsenic rule produced the most intense reaction. A national survey, conducted between April 21 and April 26, 1001, found that 56% of Americans rejected the Bush decision, whereas only 34% approved of it – and that majorities of Americans opposed the decision in every region of the nation.24 At various points, the public outcry combined concern, certainty, and cynicism. “Arsenic everywhere, and Bush is not helping,” according to one newspaper.25 “You may have voted for him, but you didn’t vote for this in your water,” wrote the Wall Street Journal.26 In an editorial, the New York Times demanded that “Americans should expect their drinking water to be at least as safe as that of Japan, Jordan, Namibia and Laos,” all of which impose a 10 ppb standard.27 A respected journalist asked, “How callous can you get, Mr. Compassionate Conservative?”28 Ridiculing the Bush Administration in a cartoon 23See 66 Fed. Reg. 16,134 (2001) (delaying arsenic rule); 66 Fed. Reg. 20,579 (2001) (ordering subsequent process) 24Mark Barabak, Bush Criticized as Fear of Environment Grows, LA Times, April 20, 1001, available at latimes.com/news/timespoll/national/lat_poll010430.htm. 25Erik Olson, Arsenic Everywhere, and Bush Is Not Helping, Balt. Sun, May 14, 2001, 9A. 26John Fialka, Arsenic and Wild Space: Green Activists from across Spectrum Unite Against Bush, Wall St. J., Apr. 11, 2001, A20. 27Robert K. Musil, Arsenic on Tap, N.Y. Times, Apr. 24, 2001, A18. 28Michael Kinsley, Bush Decision on Arsenic Tough to Swallow, Times Union, Apr. 16, 2001, A9. 7 entitled, “Safety is for Sissies,” Time Magazine epitomized public sentiment by targeting the arsenic decision as a chief example of several environmental foibles.29 The public reaction came to a head during the legislative debates on the issue, particularly within the House of Representatives, which voted to reinstate the Clinton rule on the theory that arsenic “is a poison.”30 Here is the puzzle. With respect to arsenic, the underlying issues are highly technical, and very few people are expert on the risks posed by exposure to low levels of arsenic. What accounts for the public outcry? I believe that the reason is simple: Arsenic was involved, and so was drinking water. These two facts made the controversy seem highly accessible, and it seemed easy to be outraged. Why was the Bush Administration allowing dangerously high levels of arsenic to remain in drinking water? This appeared to be a rhetorical question. By contrast, many environmental problems are both obscure and technical, and people do not have an easy or intuitive handle on them. Is carbon dioxide a serious problem? Most people have no idea. But arsenic is well-known, and it is well-known to be a poison, not least because of the exceedingly popular movie, Arsenic and Old Lace. In fact an influential environmental group, the Natural Resources Defense Council, has exploited exactly this reference with its work on the arsenic problem, under the title, Arsenic and Old Laws.31 Ordinary people seem to be “intuitive toxicologists,” with a set of simple rules for thinking about environmental risks.32 Among those simple rules is a belief that substances that cause cancer are unsafe, and should be banned.33 That intuitive toxicology does not easily make room for issues of degree. It does not accommodate the judgment that low levels of admittedly carcinogenic substances should sometimes be tolerated, because the risks are low and the costs of eliminating them are high. It does not show an understanding of the 29Bruce Handy & Glynis Sweeny, Safety is for Sissies, Time, Apr. 16, 2001, 88. 30Cong. Rec. H4751 (July 27, 2001). 31Natural Resources Defense Council, Arsenic and Old Laws: A Scientific and Public Health Analysis of Arsenic Occurrence in Drinking Water, Its Health Effects, and EPA’s Outdated Arsenic Tap Water Standard (2001), available at http://www.nrdc.org/water/drinking/arsenic/ exesum.asp. 32See Paul Slovic, The Perception of Risk 285–93 (2000). 33Id., p. 291. 8