Retaking Rationality: How Cost Benefit Analysis Can Better Protect the Environment and Our Health.
Retaking Rationality: How Cost-Benefit Analysis Can Better Protect the Environment and Our Health, by Richard L. Revesz and Michael A. Livermore, aims to convince those who favor more government regulation--in particular environmental groups--that they should embrace cost-benefit analysis and turn it to their purposes. Coauthored by a prominent law school dean and a recent student with a background in environmental advocacy,' the book is a jarring combination of roundhouse political polemics and careful academic argument. Sweeping pronouncements are followed by qualifications that leave the sweep of the pronouncements in doubt--rather like the give-and-take of the law school classroom where the work may have originated. The result cannot be judged a success either as polemic or scholarship, but the effort is highly instructive--often unintentionally--regarding the uses, limitations, and politics of cost-benefit analysis in regulatory policymaking. (2)
The argument of Retaking Rationality is that cost-benefit analysis is a valuable tool for calibrating health, safety, and environmental regulation and making bureaucratic decisionmaking more transparent and accountable, but has been perverted into a tool for weakening regulation, thereby degrading the environment and endangering public health and safety. (3) In the book's account, Republican administrations--those of Ronald Reagan, George H.W. Bush, and George W. Bush--are "conservative" and "antiregulation" and use cost-benefit analysis to block regulatory protections at the behest of corporate interests, while Democratic administrations--those of Bill Clinton and, the authors adumbrate, Barack Obama--are "progressive" and "proregulation" and use cost-benefit analysis in a "neutral" fashion to improve the substance of policy. (4) The problem is that the conservatives and business interests have been so adept at twisting cost-benefit analysis into an antiregulation weapon that environmental, consumer, and labor organizations have come to oppose the very idea of assessing regulatory benefits and costs. As a result, when progressives are in office and in a position to set things straight, proregulation groups continue to sit out the cost-benefit debates. Their boycott is shortsighted for two reasons. First, cost-benefit analysis is here to stay, entrenched in many statutes and judicial precedents. Second, when purged of the analytical "fallacies" introduced by conservatives, cost-benefit analysis is a powerful tool for more aggressive regulation--properly understood, it is proregulation not antiregulation. "Retaking rationality" means ending the boycott and restoring the neutral, beneficial principles that have been hijacked:
This is the responsibility of proregulatory interests groups [sic], as well as citizens and voters participating in democratic politics. By holding our leaders to a restored vision of cost-benefit analysis, we can make up for the damage that has been done to the technique, and allow it to serve a useful role in improving regulatory decisionmaking--and therefore our environment, our public health, and our economic prosperity. (p. 51)
Within this stark narrative structure, the intellectual heart of the book is a debunking of "Eight Fallacies of Cost-Benefit Analysis" perpetrated by antiregulation conservatives and a demonstration of how correct analysis will lead to stricter government controls. A final section criticizes the procedure conservatives have used to block worthwhile regulations with their fallacious theories--central review of agency rulemaking by the Office of Information and Regulatory Affairs ("OIRA") within the Office of Management and Budget ("OMB")--and suggests how OIRA review could be made more regulation friendly.
This Review follows the structure of Retaking Rationality. In Part I we criticize the book's narrative (summarized above) as cartoonish and unhistorical--we think it is confusing rather than helpful to understanding recent developments and controversies in cost-benefit analysis and the organization of regulatory decisionmaking within the executive branch. In Part II we consider the book's "Eight Fallacies of Cost-Benefit Analysis." We find that these discussions are generally well informed and interesting but suffer from the effort to squeeze cost-benefit issues into the antiregulation-versus-proregulation narrative; moreover the discussions are often excessively abstract and ambitious concerning the function of cost-benefit analysis, and they entirely fail to support the thesis that cost-benefit fallacies have been used to defeat beneficial regulations. Finally, in Part III we discuss the authors' arguments about the need for and practice of OMB/OIRA oversight of agency rulemaking. Here we criticize as naive the book's argument that there is no need for an institutional counterweight to agency parochialism and that OIRA's role should be recast as one of coordination, calibration, and promotion of a proregulatory agenda against the forces of agency sloth. A concluding Part sums up our arguments.
I. THE NARRATIVE
There are two serious problems with the Retaking Rationality narrative. The first is that it portrays regulatory policymaking as much more partisan than it is. Although government regulation is sometimes the subject of partisan contention during election seasons, (5) in practice it is largely the domain of interest-group politics and interbranch rivalry between the president and Congress rather than of deep partisan or philosophical divides.
Among academics, there is much less disagreement between liberals and conservatives over regulatory policy than over taxation, entitlement, and health-care and insurance policy. Across the spectrum, academic students of regulation tend to favor deregulation when it comes to price and entry controls in competitive markets (the sorts of things the Civil Aeronautics Board used to do before its abolition at the hands of Edward Kennedy, Stephen Breyer, and Alfred Kahn), and "regulatory reform" when it comes to environmental, health, and safety regulation. (6) "Regulatory reform" means, inter alia, preferring market-incentive over command-and-control techniques, regulating outputs rather than inputs, and using cost-benefit analysis to set regulatory standards and priorities; many distinguished academic liberals as well as conservatives have contributed to this literature, as Revesz and Livermore acknowledge at various points. (7) The purpose of these reforms is to make regulation more effective and productive--to counter the influence of narrow interest groups in bending rules to their selfish advantage, to avoid policies that are wasteful or counterproductive, and to get more environmental bang for the policy buck. There are important differences among academics concerning health and safety standards in competitive markets--such as pharmaceutical, automobile safety, and workplace health and safety controls--where conservatives tend to see redundancy with market incentives and harmful "unintended consequences" and liberals tend to see needed adjustments and backstops to the marketplace. (8) But the primary focus of Retaking Rationality is environmental regulation, where the government is providing public goods for which there is little effective market demand and where there is little academic disagreement about the need for regulation. (9)
Among practicing politicians, Republicans are no doubt more attentive to business interests, more sympathetic to private markets, and more skeptical about the need for regulation, while Democrats are more attentive to labor interests, more skeptical about private markets, and more sympathetic to regulation. But in environmental and health regulation, the record across administrations is much more one of continuity than of the lurches that characterize tax and health-care policy; and disagreements between the White House (not only OMB/OIRA but also the Council of Economic Advisors ("CEA")) and the regulatory agencies have been a bipartisan phenomenon. Revesz and Livermore emphasize what they regard as Republican misdeeds and Democratic good deeds, but they could have told a different story. Richard Nixon created the Environmental Protection Agency ("EPA") by Executive Order, banned DDT, and proposed a national tax on sulfur dioxide emissions ignored by a Democratic Congress. (10) Jimmy Carter launched the Regulatory Analysis Review Group ("RARG")--the White House precursor to OIRA--which tangled with the EPA, the Occupational Safety and Health Administration ("OSHA"), and other agencies over new regulations, and signed the legislation creating OIRA (the Paperwork Reduction Act of 1980 (11)--brainchild of Democratic Senators Lawton Chiles and Lloyd Bentsen) over the vehement opposition of his cabinet officers. (12) Ronald Reagan was the most "antiregulation" of recent presidents--and his EPA (during the tenures of William Ruckelshaus and Lee Thomas, following the Anne Gorsuch Superfund scandal) tightened the gasoline lead-phasedown and many other EPA rules and was notably aggressive in environmental enforcement. (13) George H.W. Bush championed and won many improvements to the Clean Air Act, including the sulfur dioxide cap-and-trade program that Revesz and Livermore heartily approve. (14) Bill Clinton was the most "proregulation" of recent presidents--and his OIRA and CEA inveighed forcefully (although with only modest effect) against many EPA actions, notably the Agency's National Ambient Air Quality Standards for ozone and particulates, which they regarded as excessively strict and costly, and the EPA's official reports on the benefits and costs of air pollution rules. (15) George W. Bush championed and won stringent new energy conservation standards including the phased abolition of the incandescent light bulb, (16) and his EPA issued new or tighter standards for lead, mercury, diesel engine exhaust, and other pollutants. (17) OIRA administrators of both parties, ourselves included, can tell war stories of standing bravely for right policy in the face of strenuous opposition from powerful administration constituencies. More generally, aggregate health, safety, and environmental regulatory activity--according to such rough measures as agency budgets, agency enforcement budgets, numbers of new regulations, and estimated costs and benefits of new regulations--has not differed much from administration to administration. (18) Nor, evidently, was the vigor or effectiveness of OIRA review much different in the Clinton and W. Bush Administrations. (19)
Selective anecdotalism is a weakness of debate over regulatory policy-an inescapable weakness because regulation operates by piecework (rulemaking involving widely varying issues and circumstances) and resists aggregation and generalization. The growing use and sophistication of regulatory cost-benefit analysis since 1980 has improved our ability to characterize policy trends, but the analyses have themselves varied widely in quality and methodology. (20) The important point is that environmental protection is a consensus issue in American politics, something that virtually all political leaders subscribe to and would ignore at their peril. That consensus has produced hundreds of billions of dollars of public investment and tremendous improvements in environmental quality from one administration to the next over a period of forty years. (21) The serious policy disagreements concern means not ends, and debates over means have been robust within the administrations of both parties.
The second problem is that "antiregulation" versus "proregulation"--a formulation employed throughout both the narrative and analytical sections of Retaking Rationality--is an impossibly crude way to characterize differences over regulatory policy or cost-benefit analysis. Regulation is a form of government action, with advantages and disadvantages from case to case relative to other forms of action or no action at all: there is no principled, coherent sense in which one can be "for" or "against" regulation tout court. To adopt this dichotomy is to leave the world of the analyst and enter the world of the lobbyist. Especially in environmental regulation, where regulated firms are paying (in the first instance) the costs of pure or nearly pure public goods, firms will characteristically favor the most lenient plausible standards while environmental groups will characteristically favor the most stringent plausible standards. Both sides realize they are in something like a zero-sum lobbying game and adopt positions based on expectations of opposing positions--they will indeed be "pro" and "anti" in the particular case, whatever concessions they make for tactical reasons. (22) But a book written by academics and published by a prestigious university press is not, presumably, a how-to-succeed manual for lobbyists--that would be appropriate for a group such as the Natural Resources Defense Council. Revesz and Livermore have set themselves to persuading environmental activists that cost-benefit analysis can advance their cause; in characterizing groups, public officials, and schools of thought as antiregulation or proregulation, they may have decided to adopt the Manichaean ethos of their primary audience at the expense of clarity and accuracy. But when they come to discussing issues of cost-benefit method and philosophy in the same terms, they have entered intellectually treacherous territory.
Fortunately, as the cost-benefit discussions proceed, the authors reveal themselves to be members in good standing of the liberal wing of the academic regulatory reform movement: ultimately, they are for right policy, not lobbying scalps. Their antiregulation cost-benefit "fallacies" are not complete shams but have some basis in reason and disinterested analysis. Their proregulation correctives are subject to qualifications and caveats, some of them quite substantial. They allow that antiregulation conservatives have occasionally introduced reforms--such as marketable permits, which allow firms to calibrate their pollution abatement efforts according to differential abatement costs--that have spurred environmental progress. And, in the end, even the unsullied, proregulation cost-benefit analysis they favor will counsel priority setting and something short of zero pollution in each and every case.
We hope the authors do not lose their audience when the going gets nuanced. But the difficulties run deeper than thematic dissonance.
II. THE COST-BENEFIT FALLACIES
Retaking Rationality's "Eight Fallacies of Cost-Benefit Analysis" are summarized in the book's chapter titles as follows: Ali Unintended Consequences are Bad, Wealth Equals Health, Older People are Less Valuable, People Cannot Adapt, People Always Want to Put Off Bad Things, We are Worth More than Our Children, People Value Only What They Use, and Industry Cannot Adapt. At the book's outset, we are told that the antiregulation monopoly on cost-benefit issues "has had profound negative consequences." (23) Later we are told that the eight fallacies "together ... amount to a virtual Berlin Wall blocking good regulations" (p. 145). But the chapters themselves tell a vividly different story. The titles are fallacies alright but nobody actually believes any of them--they are the authors' own waggish slogans, written with chapter-title license. The actual propositions turn out not to be fallacies at all but rather questions of method, measurement, and philosophy on which, the authors acknowledge, reasonable people can disagree. Can and do: each one is the subject of robust debate in academic journals and government councils. Those debates (certainly the academic ones) are generally aimed at producing better regulation rather than merely more or less regulation, and feature many positions that are either intermediate or indeterminate on the score of regulatory volume and stringency. Most strikingly, the authors' proregulation positions have been doing at least as well as their antiregulation positions in government rulemaking. The evidence that beneficial rules have been throttled by the forces of antiregulation bias is essentially nil.
Revesz and Livermore's accounts of the cost-benefit debates are well informed and instructive. Although the authors are not above selective anecdotalism, for the most part they balance forceful presentations of their own positions with fair presentations of contrary positions. But what we found most instructive were two features of their approach that we regard as serious deficiencies. First, they consider questions of method and philosophy at too high a level of abstraction--arguing that certain positions are definitively correct or incorrect when instead they are more or less useful in different circumstances. Second, they regard regulatory cost-benefit analysis as a device for social engineering. The social-engineering approach treats people and institutions as relatively passive and regulation dependent--as cogs in a machinery of government optimization and control--and wants cost-benefit analysis to subsume many important political questions. Our view of cost-benefit analysis is much more modest. We regard it as a device for gauging public interventions in a world that is governed mainly by the decisions of individuals and markets, and where the political process generates profuse demands for intervention that vary greatly in their merits and cannot all be accommodated in any event. The Revesz-Livermore approach is certainly proregulation, but to our way of thinking it is mainly unrealistic. We think there are natural limits to the possibilities of effective government intervention--especially in a nation as individualistic and heterogeneous as the United States--which cannot be reduced much by techniques of analysis and control. And we think that many important political questions--such as the just distribution of wealth, the desirability of government paternalism, and the obligations of current generations to future generations--cannot be effectively decided by cost-benefit analysis. (24)
The problems of excessive abstraction and excessive ambition are related: both derive from the conviction that pure reason--the "rationality" of the book's title--is capable of resolving or displacing a great many political and intellectual differences. (25) This impulse may, ironically, lead one to chalk up opposing views to bad motives rather than legitimate disagreement, as sometimes happens in Retaking Rationality. Looking at cost-benefit issues in the context of specific cases rather than pure principle tends to suggest practical answers to questions of method, and also to suggest practical limitations to cost-benefit analysis itself. As an added benefit, the narrower, more pragmatic approach tends to eliminate the clutter of partisan polemics and the effort to judge cost-benefit methods according to their tendency to produce more regulation rather than more social welfare.
These arguments are developed in the following brief reviews of Retaking Rationality's cost-benefit arguments. We begin with more extended discussions of the books first three fallacies; then, with the themes of our criticisms established, we extend them more briefly to the last five fallacies.
A. All Unintended Consequences are Bad
Regulations that reduce health and safety risks may produce offsetting increases in risks. When regulations ban or restrict certain risky products or activities, or increase their prices or costs or reduce their utility, people may substitute other products or activities that involve risks of their own. Thus, banning an effective pharmaceutical drug because of its harmful side effects may lead people to substitute drugs that are less effective or have their own harmful side effects. The authors do not object to taking account of such "risk tradeoffs" in assessing the costs and benefits of regulations--indeed they applaud it so long as the causal link is not excessively attenuated. But they think that proponents of risk-tradeoff analysis "tend to look only at negative collateral consequences" and to ignore collateral benefits (p. 55). Their principal example of collateral benefits is controls on conventional air pollutants such as sulfur dioxide, nitrogen oxide, and particulate matter-they are imposed to provide immediate health and aesthetic benefits but have the additional benefit of reducing emissions of carbon dioxide and other greenhouse gasses, thereby reducing the risks of global warming (and vice versa--controls on greenhouse gases would yield additional reductions in conventional air pollutants) (pp. 59, 63-64). Another example is controls on carbon monoxide ("CO") exhaust from motor vehicles: they have been intended to provide health benefits from pollution reduction but have had the incidental, and much larger, benefit of reducing accidental deaths and suicides from CO inhalation. (26)
Revesz and Livermore want to give equal consideration to ancillary benefits, and think this will lead to more and more stringent regulation, because "there is no reason to think they are any less common than countervailing risks" (pp. 58, 61). The authors do not explain this proposition. We think it is mistaken, and that both the mistake and its assertion as a simple identity (like a rule of arithmetic) illustrate the perils of thinking about cost-benefit analysis in terms of comprehensive social control. If one conceives of regulation as dispensing benefits and assigning costs around an essentially compliant society, then the ancillary effects will appear to be random and unsystematic in relation to regulatory purposes. But if one conceives of regulation as an intervention into a market (or other social arrangement) where individuals and organizations have durable interests and purposes of their own, then the ancillary effects will not appear random. A regulation will be intended to achieve a purpose that, by definition, is not being achieved privately--but the affected individuals and organizations will continue to pursue their independent purposes after the regulation is imposed, and will have myriad means of doing so that are beyond the scope of the regulation. In such a world, private responses to regulation will tend to compromise regulatory purposes systematically rather than to compromise and amplify those purposes randomly or equally. (27)
Our authors offer several examples of academics, policymakers, and judges concerning themselves with risk tradeoffs, and attribute the failure to give equal time to ancillary benefits to antiregulatory bias rather than to anything inherent in regulation itself. They offer reductions in greenhouse gasses from controls on conventional air pollutants as an example of how ancillary benefits could justify stricter controls, but do not offer a single example of the neglect of ancillary benefits blocking or weakening an actual regulation. In fact, the EPA and other agencies frequently include ancillary benefits in their benefit estimates (Retaking Rationality acknowledges a few such cases and complains that the practice "appears to be spotty" (p. 60 n. 183)). And OIRA itself recommends that agencies account for ancillary benefits as well as countervailing risks (Retaking Rationality applauds this but complains that the point "takes up but half a page of the guidelines" (p. 60 n.184)). The closest we get to an actual biased result is a truck fuel-economy standard whose cost-benefit analysis failed to account for the ancillary benefit of reduced carbon emissions--but, the authors note, the standard was rejected by an appeals court for precisely that reason (p. 64).
There appear to be no legal, political, or intellectual (certainly not from us) impediments to treating ancillary benefits and countervailing risks equally in cost-benefit analysis and regulatory design. The progress of policy research and debate should therefore provide a fair measure of whether the two phenomena arise equally, as Revesz and Livermore believe, or one of them is far more characteristic and important, as we believe.
B. Wealth Equals Health
Typically, the costs of complying with regulations are paid in the first instance by business firms (e.g., to install pollution-control equipment or adjust production methods) but ultimately borne by individuals in the form of reduced returns on investment to owners, reduced wages to employees, or increased prices to consumers. Compliance costs thereby reduce individuals' incomes and wealth, an effect that is separate from the risk-substitution effect discussed in the previous Part. Wealth, however, is correlated with health--wealthier people tend to be healthier. Should a cost-benefit assessment of a regulation aimed at improving health or safety take account of offsetting reductions in health occasioned by reductions in wealth? Revesz and Livermore say emphatically no, because correlation does not amount to causation. They note that it is easy to imagine causation running in the opposite direction: people who are less healthy may be less wealthy because of their poorer health. And they point to a recent study finding that wealth and health are not highly correlated after adjusting for education--more highly educated people tend to be both wealthier and healthier.
Here the authors mostly stick to their polemical guns about cost-benefit fallacies and antiregulatory skullduggery. The health-wealth tradeoff is "pseudo logic," a statistical "rookie error," and "a sham, devoid of meaningful content, designed to shut down regulations"; those who embrace it believe that "health regulations should be abolished because they kill people"; it "has been used by courts, OMB, and Congress to overturn regulation [and] circumvent statutory prohibitions against taking costs into account" (pp. 67, 70, 145-46). The rhetoric is strangely overwrought. The statistical rookies who have perpetrated the health-wealth sham turn out to include Professors W. Kip Viscusi and John Graham, Justice Stephen G. Breyer, and Judges Frank H. Easterbrook and Richard A. Posner. That lineup should give the reader, and should have given the authors, some pause. The only account of the health-wealth tradeoff in action is an OIRA rejection of a proposed OSHA regulation on that ground in 1992. The rejection, however, incited immediate condemnation by congressional leaders and newspapers, a Senate hearing, and an investigation by the General Accounting Office; OIRA promptly backed down and approved the rule for issuance and has not been heard from on the matter since. (28) Beyond this nonshutdown, we are pointed only to references to health-wealth tradeoff issues in a few judicial opinions--all but one in concurring or dissenting opinions and that one in an opinion upholding the regulation in question. (29)
The health-wealth tradeoff is not a sham but a perfectly respectable variant of the risk-tradeoff phenomenon approved by Revesz and Livermore. That poorer health may lead to less wealth, and that more education may lead to both more wealth and better health, does not mean that less wealth cannot also lead to poorer health. It is not a matter of one or the other or right or wrong, but of degree and of the circumstances of the case at hand. In particular, it may be true both that accidents and serious illnesses result in loss of income and wealth and also that losses in jobs and employment opportunities result in poorer health. (30)
Consider the opinion of Judge Posner cited by Revesz and Livermore. Writing for the court in approving (with an exception not pertinent here) a costly OSHA rule designed to limit the exposure of dental and medical care providers to blood-borne AIDS and Hepatitis B viruses, Judge Posner noted that:
OSHA ... exaggerated the number of lives likely to be saved by the rule by ignoring lives likely to be sacrificed by it, since the increased cost of medical care, to the extent passed on to consumers, will reduce the demand for medical care, and some people may lose their lives as a result. (31)
This is probably not a health-wealth tradeoff at all--the reduced demand for medical care seems to be the result of higher price, which would be a risk tradeoff. But the reduced demand might instead or additionally be the result of lower wealth, which would be a health-wealth tradeoff, so the example really illustrates the similarity of the two phenomena. (32)
In any event, the important point is that the pursuit of health is not the exclusive preserve of government regulation: health is a private good as well as a public good, something that people value and pursue all on their own. With greater resources, individuals are in a position to pursue better health through more or better medical care and many other means. (They are also of course in a position to pursue things that have little or nothing to do with better health or that are risky to health.) When regulations direct resources toward health-improvement projects the government deems important, they also direct them, to some degree, away from health-improvement projects individuals deem important. The academic studies cited by Revesz and Livermore, and the OIRA assessment of the OSHA rule that stirred congressional outrage, may or may not have estimated the relationship correctly, (33) but it is not in principle different from risk tradeoffs where risk-reducing rules divert people to activities carrying risks of their own.
In the end, our authors concede that the issue of health-wealth tradeoffs is "debatable" (34) and proceed to argue that, to the extent such tradeoffs exist, the proper response is not less costly regulations but rather redistribution. They suggest without elaboration that regulations be designed so that compliance costs fall on people with higher rather than lower incomes, or that they be accompanied by transfer payments, job retraining, and the like (pp. 73-75). We have no idea how the costs of complying with a health regulation could be directed to particular income groups, and greatly doubt the feasibility of linking the government's many income-transfer and training and development programs to regulatory decisions in any coherent way. These are additional examples of the authors' unbounded conception of cost-benefit analysis and the possibilities of regulatory social engineering. They criticize "antiregulatory groups" that subscribe to the health-wealth tradeoff for not following through with proposals for more income redistribution to compensate for regulatory costs (pp. 73-75), but it seems much more realistic to say that health regulators should do their best at reducing health risks without attempting to resolve other social and economic issues along the way.
The authors also suggest that the proper role of health-wealth tradeoffs is as "an additional justification for health and safety regulation," because healthy citizens are more productive and increase social wealth (pp. 75-76). Those effects, however, are already accounted for in cost-benefit analysis, and indeed are the heart of the benefit estimate for a health or safety regulation. The health-wealth tradeoff proponents are simply proposing, so far without success to judge by the evidence in Retaking Rationality, that the counterbalancing effect be considered as well--as an addition on the cost side of the ledger where statutes permit consideration of cost, or as a subtraction on the benefits side where statutes command the pursuit of health benefits regardless of cost.
C. Older People Are Less Valuable
In estimating the benefits of regulations that reduce health and safety risks, cost-benefit analyses traditionally counted savings of "statistical lives," meaning that they estimated the number of fatalities from illness or accidents that the regulations would avert. More recently, a more refined technique has been to estimate years-of-life saved as well. The refinement is an attempt to account for the fact that regulations do not really "save lives," and that measuring benefits in this manner gives an imperfect account of their function. We all die sometime, and our lives are "saved" innumerable times along the way (because things that might have killed us did not); all that risk-reducing regulations can do is extend expected life spans by reducing a subset of the risks we face. But substituting actuarial years-of-life saved for fatalities averted may suggest that regulations addressed to risks facing younger people (say, rules requiring car seats for infants and toddlers) are more beneficial than those addressed to risks facing older people (say, rules aimed at reducing the incidence of cancers and other diseases of older age)--and therefore justify greater expenditures.
Revesz and Livermore believe that the years-saved refinement "must be abandoned" because it is "theoretically incoherent and empirically ungrounded" and "leads to the fallacy that older people are less valuable than younger people" (pp. 84, 146). Their first argument is that an additional year of life is more valuable to older people because they have actuarially fewer years left to live in any event--getting to 76 from 75 is a bigger benefit to someone who is 70 than to someone who is 20, and the 70-year-old is likely to have more discretionary wealth and therefore more willingness to pay for the additional year. That is a weak argument, because the issue is not the purchase of an additional year of life but the purchase of a reduction in a particular health or safety risk against a background of other risks (upside and downside) that will vary with age. On this score, the authors direct us to several theoretical models and empirical studies. The studies point every which way--to greater willingness to pay for risk reduction among younger people, older people, and people in their middle years when earnings and obligations are greatest. Because these studies are inconclusive, and because life-years proponents simply estimate years saved without estimating the differential values of risk reduction to older and younger people, the authors conclude that cost-benefit analyses should be limited to estimating numbers of lives saved--"the gold standard for estimating the value of life-saving regulations" (p. 146).
The authors' plenary opposition to measuring regulatory benefits by life-years saved is puzzling. For one thing, the issue fits particularly poorly into their storyline of bricks in a Berlin Wall blocking good regulations. We are given the hypothetical example of EPA ozone rules, which primarily benefit people in ill health with lower life expectancy, and which therefore might be scaled back if benefits were measured by years saved instead of lives saved. But that has not happened, and various efforts during the George W. Bush Administration to introduce years-saved estimates were as additions (not substitutes) to lives-saved estimates. And even these efforts, we learn, were abandoned in the face of objections from senior citizens' groups to a "senior death discount." The outcry arose over an EPA cost-benefit test of an air pollution rule similar to, and perhaps the source of, our authors' hypothetical. But the EPA's capitulation was even more abject than OIRA's on the health-wealth tradeoff issue discussed in the previous Part of this Review--because, entirely unmentioned by the authors, the EPA rule passed the test according to the years-saved approach as well as the lives-saved approach. (35) Once again we are without any evidence of antiregulation forces rolling back regulation; on the contrary, this episode is said to demonstrate that "if progressive groups get involved in the conversation over how cost-benefit analysis is conducted, they can prevail, especially if they stand on the moral and rational high ground, as they did on this issue" (p. 84).
But the deeper puzzle involves the substance of the issue. Lives saved is not a gold standard; insisting that regulators ignore the years of life saved by their rules is not a rational high ground; the AARP's publicity campaign against the EPA study was not a moral high ground; and taking account of years saved is not "antiregulation." Revesz and Livermore are correct that focusing exclusively on years saved can produce analytical and ethical dilemmas, but an exclusive focus on lives saved can produce parallel dilemmas. (36) If the years-saved approach can be said to yield a "senior death discount" compared to a lives-saved approach, it does so only to the extent that the lives-saved approach yields a "junior death discount." Some rules are addressed to risks of infancy, childhood, and youth, others to risks of older age, and others to risks that do not discriminate by age--and no one has any idea whether lives saved or years saved would result in more or stricter regulations on net. In practice, estimating years saved is inescapable in assessing the benefits of regulations that affect morbidity--the duration and severity of disease under different policies--and the approach has been applied to mortality reduction (often in the same rules) in many cases without political controversy. (37) That has been so even after the EPA recanted its years saved assessment in the case celebrated by Revesz and Livermore--in the Obama as well as Bush Administrations. (38) And years-saved analysis is likely to continue in circumstances where our authors will find it difficult to allege antiregulatory bias. In 1996, when the Clinton Food and Drug Administration ("FDA") attempted to regulate the marketing of cigarettes to children and adolescents, it measured the benefits by years saved as well as lives saved--without anyone complaining that it had introduced a "junior death premium." (39) The rule was vacated on jurisdictional grounds. (40) Now the FDA has been given jurisdiction. (41) The rule will be back soon.
We think Revesz and Livermore object to estimating years of life saved because they are attempting to accomplish too much by cost-benefit protocol. Some things, such as improved visibility and other aesthetic values and the psychic satisfaction of wilderness preservation, cannot be quantified beyond extremely broad and uncertain ranges of value. In such cases, the important analytical task is to set out the costs so that public officials can make informed decisions on how much to purchase. The differential value of reducing risks to people at various stages of the life cycle is certainly one of those things. The solution is not to restrict the amount of readily available information laid out in a regulatory analysis: the better approach is to augment lives-saved estimates with simple unadjusted years-saved estimates in cases where the additional information seems pertinent. Where two equally effective regulations would affect populations of the same size, one of oldsters and one of youngsters, the benefit assessment would show the same number of lives saved for both rules but more years of life saved by the youngsters rule. That is true information, and, according to the empirical and theoretical literature cited by the authors, the analyst doesn't have much more to tell the decisionmaker about what to make of it. (42) The result of the full-disclosure approach may even be proregulation: if controlling the risks facing different age groups is left to informed political choice, we may find that protecting children gets a special boost, and protecting the elderly also gets a special boost, and protecting hard-working moms and dads all across America gets a special boost too.
D. People Cannot Adapt
A further refinement of lives-saved estimates is to consider not only life years saved but the quality of the saved years. The result--quality-adjusted life years ("QALYs")--takes us over a critical threshold. To measure regulatory benefits by lives saved or life years saved is to deal with, in principle, objective facts. To value those benefits by willingness to pay, as revealed by private decisions involving risk, is to gauge public risk-reduction expenditures by similar (albeit usually implicit) expenditures people make in their private lives. But valuing benefits by the quality of years saved---say, years in dialysis or in a wheelchair versus years in good health--is to introduce an irreducible element of subjectivity (and in an area where interpersonal comparisons are often erratic).
Retaking Rationality makes a persuasive case that the government's judging the quality of citizens' lives is generally unsound, unhelpful, and unpopular, and is often ethically problematic to boot. (43) But the discussion is marred, once again, by its abstract, absolutist character and failure to consider that those who are dubious but not abolitionist about QALYs may have a serious point or two. If one were considering the benefits of a regulation to reduce the incidence of infections acquired in hospitals, would it make sense to count reductions in death rates among patients in permanent vegetative states equally with those among mothers and newborns on the maternity ward? The effort to force cost-benefit issues into the antiregulation-proregulation construct is particularly unproductive in the QALY discussion. The authors point to no case where QALY considerations determined or were even considered in a regulatory decision aimed at saving lives. Their warning that QALYs are nevertheless "on the horizon" and "must never get off the ground" is based on a few academic papers and on references--all of them nuanced and tentative--in two OIRA reports and one judicial decision that was reversed on other grounds (pp. 87-89, 146).
E. People Always Want to Put Off Bad Things
Intertemporal discounting is the practice of aligning the values of regulatory benefits and costs, for purposes of cost-benefit comparison, in circumstances where costs are incurred before benefits begin to materialize. An important case, and the focus of the discussion in Retaking Rationality, is health and environmental regulations whose pollution-reduction costs begin immediately and whose benefits begin in later years--because the health effects of the pollutants would have been cumulative and/or because the diseases averted would have had long latency periods. (44) The lower the rate used to discount future benefits to the time when costs are incurred, the higher the estimate of a rule's net benefits. Revesz and Livermore refer to many discount-rate tussles among the regulatory agencies, OIRA, and the courts, and point to one important case, Corrosion Proof Fittings v. EPA, in which the court vacated and remanded the H.W. Bush Administration's asbestos ban. (45) The court criticized the EPA for failing to discount health benefits at all for the (typically lengthy) periods between asbestos exposure and the manifestation of disease, and directed the EPA to employ a reasonable discount rate on remand. (46)
That would be a smoking pistol--an accurate example of an antiregulation fallacy blocking a worthwhile regulation--except that (a) the EPA's failure to discount benefits was not among the grounds for the court's decision, (47) and (b) the authors are not opposed to intertemporal discounting at all--indeed they support it. Their argument is instead that discount rates for latency-period benefits should take account of the postexposure dread that people might experience, in the absence of the regulation, in anticipating the risks of acquiring a disease in the future, and of the fact that occupational and environmental risks may be considered involuntary rather than voluntary (because people are more averse to involuntary risks). But the discussion of these factors is amorphous and provides no idea of how significant the adjustment would be or what difference it would make. The authors object to using "a straight discount rate borrowed from the Wall Street Journal financial pages" (p. 146), but that is essentially never done. The EPA's discount rate for preexposure future benefits in the asbestos case was three percent, and the OMB's current guidance document for regulatory analyses advises that rates of both three percent and seven percent should customarily be employed. (48)
F. We Are Worth More Than Our Children
Where the lag time between regulatory costs and benefits is not merely intertemporal but intergenerational--that is, where regulations impose costs today whose benefits will accrue to future generations now unborn--the discounting issues become trickier and more ethically fraught. Discounting benefits over many decades suggests that averting enormous future catastrophes merits only trivial current efforts, but failing to discount suggests indifference between saving lives today and saving them far in the future.
The Revesz-Livermore discussion of intergenerational discounting is competent but restricted and at points tendentious. On the critical question of the unpredictability of the distant future, they dispute two arguments for intergenerational discounting: that future generations will be much richer and more proficient than the present one if current rates of economic growth and technological progress continue, and that the human race might be wiped out by an asteroid or nuclear war before the problem we are trying to avert comes to pass (pp. 111-113). But these are only two among innumerable possible developments that, taken together, are likely to make the remote future very different from what we can imagine or project today. It is the totality of cumulative contingencies over long periods of time, not technophilia or worries about asteroids, that should give any sensible person pause over large current investments to address one future problem--a problem we can conceive of today but whose nature may change fundamentally over generations to come. The authors' obliviousness to this issue is dramatized by a silly example: "Discounting the value of regulatory benefits that accrue to future generations is as arbitrary as discounting benefits affecting people west of the Mississippi River--unfair and unjustifiable" (p. 146). But we know fully as much about the circumstances of people on the west side of the Mississippi as we do of people on the east side. To eliminate the dimension of time is to eliminate the very issue discounting seeks to address.
Revesz and Livermore agree that we should consider the opportunity costs of current regulation (the alternative uses society would make of the resources regulation would command) and weigh the consequences for future generations of regulatory versus private exertions (pp. 113-14). (49) But their main concern is to prevent these and other considerations from leading to any discounting of future benefits against current costs--"discounting in the intergenerational context amounts to punting one of the most important moral questions of our time" (p. 146). (50) They recommend that moral obligations be addressed through such concepts as sustainable development and intergenerational distributive and corrective justice (pp. 116-17, 146)--yet still they regard it as essential that such inquiries stand alongside undiscounted rather than discounted benefits. (51)
It seems to us that the authors are once again trying to load cost-benefit analysis with too much ethical and political freight. As a practical matter, it probably makes little difference whether the benefits of averting distant problems are discounted fully, partially, or not at all. Current U.S. government deficits--which amount to sending enormous bills to our grandchildren to pay for our own consumption--suggest that the American people, or in any event American politicians, are not overly concerned with our obligations to future generations. Not discounting intergenerational benefits may be intended to generate large, scary numbers to stimulate ethical and political debate, but our deficit and debt numbers are already large and scary. The only live regulatory controversy Revesz and Livermore mention in connection with intergenerational discounting is climate-change policy; they acknowledge that discounting "may not be the driving cause" for the failure to regulate greenhouse gas emissions, and emphasize conventional political causes. (52) But it is an overstatement to say that discounting is "at least a powerful means of justifying that inaction" (p. 109). The media have been filled for many years with accounts of horrific future consequences of global warming, to little policy avail. At least where policy controversies are out in the public square rather than embedded in regulatory and judicial proceedings, to-discount-or-not-to-discount distant benefits is a distinctly subsidiary question.
G. People Only Value What They Use
The "existence value" of natural resources is the value people attach to the mere existence of wilderness areas and living species that they expect never to experience personally or to derive any economic benefit from. It is one of many examples of aesthetic, psychic, and moral benefits of environmental protection that are indubitably real but extremely difficult to measure for quantitative balancing in a cost-benefit assessment (because of the absence of immediate or derivative markets that reveal exchange value and willingness to pay). One would think that existence values could therefore be left unpriced in a cost-benefit analysis: like questions of distributive justice and the differential claims of younger, older, and future generations, they could be described in qualitative terms and their value left to the decisions of political officials. But the monetizers have been busy with existence values--conducting surveys of how much people say they would pay to protect a given wilderness area or other resource and introducing the results into regulatory cost-benefit analyses and private litigation. The purpose, clearly, is to put a price on the priceless and thereby give political officials a nudge.
Revesz and Livermore provide a good account of the shortcomings of stated-preference surveys and of various suggestions that have been made for improving them. They conclude only that the criticisms "are not sufficiently strong to justify abandoning this significant set of preferences" and that "[e]xistence-value studies should follow best practices" and "state-of-the-art techniques" (p. 129). In our view they give short shrift to the most fundamental difficulty with existence-value surveys, which is that the world includes an endless number of things, of which environmental things are only a subset, that people value and will put a price on if asked. That means that "agenda control"--which things get asked about and which do not--is paramount. Our authors acknowledge the problem but are happy with the current agenda--"we should focus our analytic energies on the class of existence values we know to be important, namely natural resources existence values" (p. 126).
In the meantime, we have yet another fallacy that has not blocked any environmental efforts--"challenges to existence value have thus far been mostly unsuccessful, and the technique is now used widely in cost-benefit analysis" (p. 125). The authors note that the Exxon-Valdez oil spill, which is their leading illustration of existence values, eventuated in damage and penalty payments far greater than the tangible economic damages of the spill-suggesting public recognition of the need for a "fuller reckoning from the company." (53) They also note that Exxon-Valdez prompted legislation providing for liability for destruction of natural resources, and that a contemporary court of appeals decision approved consideration of nonuse values under the Superfund program (p. 128). Their evidence of "some success" in "a campaign to erode support for the use of stated-preference studies of existence value" is a 2002 OMB/OIRA report that identified but did not monetize existence values for a forest preservation rule in a list of proposed regulations (p. 121). But the list was actually of final rules that OIRA had approved. (54)
H. Industry Cannot Adapt
Retaking Rationality's final cost-benefit chapter is a thorough discussion of the tendency for regulatory compliance costs to turn out to be less than estimated at the time of rulemaking. Realized costs are less than projected costs because of "learning by doing" and innovation in the course of complying with rules; the pursuit of efficient regulatory compliance is particularly strong where rules are based on market incentives rather than technological command and control. The tendency is clearly important and systematic, but it is not clear that it has operated to constrain the issuance of regulations (the authors suggest no instances) or that it amounts to a criticism of cost-benefit analysis, or any of its methodological particulars, at all. Cost-benefit analysis is an effort to estimate the consequences of a course of action foreseeable at the time the action is undertaken. Imperfect foresight is not limited to compliance costs--as we have seen, it is a critical component of the argument for discounting intergenerational benefits--and is hardly an argument for more aggressive regulation.
Although Revesz and Livermore are persuasive that lack of foresight will lead systematically to overestimation of regulatory costs, they say nothing about regulatory benefits, where overestimation is also systematic. Many (not all) health, safety, and environmental regulations aim to produce improvements in areas that are already improving and can be expected to continue improving independently of regulation. Efficiency in manufacturing and power generation (resulting in less wasted energy and less uncompensated byproduct in the form of pollution), workplace safety and health, highway safety, and the safety of many consumer products were improving before the establishment of the relevant regulatory programs in the 1970s--so there is no reason to credit regulation with all of the subsequent improvements. (55) The preregulation improvements reflected increased public demand for health, safety, and environmental goods and increased technological capacity to supply those demands--developments that were instrumental in generating effective political support for regulation. Of course, the enactment of the regulatory programs indicates that the rate of progress was considered inadequate, as it certainly was in the case of air and water pollution. But that does not gainsay the fact that regulatory improvements will be complemented by continued private improvements resulting from market demand and technological advances (augmented in many cases by tort liability). To say that assessments of regulatory effectiveness should take account of private innovations that reduce costs but ignore private innovations that increase benefits would be proregulation indeed--and completely arbitrary from the standpoint of social welfare.
A glimpse of the relationship between regulatory requirements and natural technological progress is provided by the George W. Bush Administration's unsuccessful effort to relax "New Source Review" requirements for old power plants needing engineering upgrades. Revesz and Livermore mention the episode elsewhere in their book, saying that "[i]t is inconceivable that weakening the New Source Review provision ... would result in a net benefit for the environment" (p. 154). (56) But it is not inconceivable--it was one of the central policy questions in the effort. (57) Older, "grandfathered" power plants are subject to less strict pollution standards than new plants, and that creates artificial incentives for utilities to keep older plants in operation. A longstanding EPA enforcement conundrum has been when to treat engineering improvements in old plants as triggering stricter "New Source" requirements. The Bush initiative was based on the hunch that permitting more substantial renovations without New Source Review would lead to a faster rate of pollution reduction (along with efficiencies in power generation) through more frequent maintenance and upgrades. It was only a hunch, and only a second-best expedient for coping with the differential in pollution requirements. (58) But it was an informed hunch, based on the experience of numerous enforcement cases that had resulted in lengthy delays in plant maintenance. And it risked little, because a concurrent cap-and-trade initiative would have provided added incentives for pollution-reducing engineering upgrades while eliminating the possibility that they would increase net pollution. (59)
Our ability to foresee benefit-increasing innovations is as limited as our ability to foresee cost-reducing innovations. Revesz and Livermore recommend the use of "prediction markets" and more diligent efforts to assess regulations ex post; these are excellent ideas, and although the authors recommend them only for costs we assume they would not object to applying them to benefits as well.
I. The Sum of All Nonfallacies
Retaking Rationality's tour of the cost-benefit horizon does not come close to showing that the authors' version of rationality has been lost and needs to be reclaimed nor that rationality per se is capable of resolving the open issues in regulatory policy. This is impressive. Our tour guides are exceptionally intelligent, knowledgeable, and committed: if cost-benefit analysis had blocked or rolled back important regulatory initiatives, they certainly would have told us. Instead we learn that many of the analytical refinements they dislike, such as health-wealth tradeoffs and life-years-saved valuations, have themselves been blocked by old-fashioned politics, often motivated by the book's primary audience of cost-benefit-hostile environmental activists. More thorough application of the authors' analytical preferences--low or zero discounting of future regulatory benefits, and stated-preference valuation of selected natural resources--would no doubt lead to a fuller agenda of regulatory proposals that look good on paper vying for political attention. And this might in turn lead to more and stricter rules and more social resources devoted to complying with (and working around) those rules. But the desirability of that result cannot be determined by abstract principle, and an empirical demonstration would require a richer understanding of the interactions between government rules and private decisionmaking than the authors provide.
Our own reading of the chapters is that the introduction of regulatory cost-benefit analysis has been a considerable achievement. The procedure has increased the transparency and accountability of agency rulemaking, flushed out some of the more vacuous special-interest arguments for and against regulation, and narrowed and clarified many worthy, important policy disagreements. There is reason to hope that continued cost-benefit debate can yield further improvements--better understanding of the nature of our political differences and better, more public-interested policies. But cost-benefit analysis is a technique for better politics, not politics itself. Investing it with too much capacity to settle political differences and extend government regulation, rather than to narrow political differences and clarify regulatory purposes and results, undermines its essential functions.
III. RATIONALIZING REGULATORY REVIEW?
The third and final part of Retaking Rationality turns to the institutional environment in which rules proposed by the regulatory agencies are reviewed by OIRA. After criticizing the current review process but allowing that it is not beyond reform, the authors summarize their prescription as follows:
[A] new executive order to define a new role for [OIRA] is needed. This executive order would set out new guidelines for how OIRA is to approach cost-benefit analysis by changing the focus away from checking agencies and toward agenda-setting and the calibration of regulatory stringency. Further, the executive order will prioritize the non-cost-benefit analysis function of OIRA, including interagency coordination and harmonization, and distributional analysis. (p. 171)
Revesz and Livermore justify their new vision for OIRA review in two steps. First, they attempt to debunk the argument in favor of centralized regulatory review insofar as it is based upon the president's need to counter the influence of special interests in order to exert his own influence. Then they reconceptualize regulatory review as a way to counter the bias against regulation they claim inheres in cost-benefit analysis. In Part II we noted our disagreement with the authors' view that cost-benefit analysis is not merely a tool to help policymakers reach informed decisions about political questions but a device for social engineering that can subsume many of these political questions. We have the same objection to their vision of centralized regulatory review. In arguing against the presidential control model and in favor of their proposed reforms, they seek to recast centralized regulatory review as a purely rational, apolitical tool--social engineering turned upon the government itself. In this vision, the purpose of OIRA review is not to enable the president to exercise authority over the administrative state but rather to optimize the work of the regulatory agencies.
We have in the past argued, as Revesz and Livermore restate our point, "because the president is responsible to a national constituency, he (or she) will be less sensitive to the kinds of special-interest pressure that might dominate the agencies." (60) Being subject to direct presidential control, OIRA is similarly less sensitive to special-interest pressure. Revesz and Livermore posit two objections to this observation--that the president is no less subject to special-interest pressures than are the regulatory agencies, and that even if he is, he cannot keep OIRA from succumbing to those pressures any more effectively than he can keep a regulatory agency from doing so.
To our claim that the president is less susceptible to interest-group pressures than are regulatory agencies, the authors respond that "[e]xperience and theory ... show that this hope is not well founded" (p. 167). To support that assertion, they first point out that "the president will be attentive to those groups that can provide him with the resources, support, and votes to win elections for himself and his party" (p. 167), citing a point in our 1986 article at which we say "[a]lthough presidents and legislatures are themselves vulnerable to pressure from politically influential groups, the rulemaking process--operating in relative obscurity from public view but lavishly attended by interest groups--is even more vulnerable." (61) While we recognized that presidents are subject to interest-group pressures, it is surely incorrect to suggest, as Revesz and Livermore do, that they are no less likely than are specialized administrative agencies to succumb to such pressures. The only other support the authors provide for their attempted refutation is a pair of thin anecdotes about Republican vice presidents influencing regulatory policy and their own conclusory declaration, based upon one such anecdote, that "[e]xperience with direct presidential-level review of agency decisionmaking" in the Administration of George H.W. Bush was "atrocious" (p. 167). This is weak tea, indeed.
Revesz and Livermore not only fail to persuade with the evidence they do present, they fail to grapple with the evidence against their claim. In particular, they do not mention a certain empirical study of OIRA's review of several thousand rules which was designed specifically to test whether OIRA was influenced by special-interest pressures. Some of the study's findings directly support our view that presidential control of OIRA is relatively insulated against special-interest pressures and undermine Revesz and Livermore's anecdotal argument to the contrary:
The fact that the White House met only with broad-based interests concerning some rule under review, or only with narrow interests, or both, has no apparent bearing on the likelihood that the White House would require a change in the rule.... The fact that the economic significance of a rule is no predictor of change casts further doubt on the idea that the White House uses the review process as a cover to benefit politically powerful groups. (62)
This study is cited eight times in a 2006 article that Dean Revesz coauthored, (63) but Retaking Rationality does not mention it at all--even though the earlier article is listed in the Acknowledgements of prior work upon which the authors are drawing (p. 195).
After arguing the president and OIRA are as vulnerable to interest groups as the regulatory agencies, the authors challenge the point, which we advanced in 1986, that centralized review is necessary if the president is to put his stamp upon the regulations issued by his administration, for which he will be held accountable politically. Here the authors' claim is that the president can control the program agencies just as effectively as he can control OIRA. (64)
As a preliminary matter, the authors' claim completely misses a key point: regardless whether the president can control OIRA more or less effectively than he can control an agency, with centralized review he has to control only one rather than many agency heads. (65) To confront the book's argument on its own terms, however, we think the authors have an uninformed and quite mistaken view of the relationship between OIRA and the president. Although the administrator of OIRA and the heads of the program agencies are alike presidential appointees, they operate in very different environments and have different incentives. OIRA, which is a part of the OMB and therefore of the Executive Office of the President, is charged primarily with implementing the president's policies in a way that the heads of the program agencies cannot be counted upon to do. OIRA's success or failure reflects directly upon the president, whether his electoral platform called for more or less aggressive regulatory policies than those of the preceding administration. The OMB director, who typically represents OIRA in White House councils and before the president when an agency appeals an OIRA decision, invariably has a closer working relationship with the president than has the head of any regulatory agency--regardless whether it is (like OSHA) or is not (like the EPA) part of a department with a cabinet secretary to champion its interests before the president. (66)
The head of a regulatory agency such as the EPA or the FDA does not typically have a close working relationship with the president who appoints him or her. On the contrary, it is not uncommon that the two will never have met before the president offers the appointee the post. The appointee will have neither a tie of personal loyalty nor a regular relationship with the president from which a sense of common purpose might grow and be nurtured.
Of course, the president will seek to appoint someone whom he has reason to believe shares his policy perspective, at least at a fairly general level. It would be surprising, however, if the president had any particular idea about how the FDA should change its regulatory activity during his administration or how the EPA should administer some provision of the Clean Air Act, apart from possibly a sense that the agencies have been "too lax" or "too aggressive" during the preceding administration. Thus, Presidents Carter and Reagan campaigned on the premise that the regulatory agencies had been overly aggressive (67) and President Obama campaigned on the opposite plank. (68) The people they chose to head OIRA, in addition to having expert knowledge in the field of regulation itself (typically a background in economics, administrative law, risk assessment, or cost-benefit analysis) presumably shared that general view.
The Office of Presidential Personnel is entrusted with identifying people to head the regulatory agencies from among those candidates who are both qualified by specialized knowledge and generally agree with the president about the direction in which regulation should go. (69) Once confirmed, however, the agency head is unlikely to have regular contact with the president, apart perhaps from some ceremonial occasion, such as the dedication of a new facility or attendance at the White House Christmas party. It falls instead to the administrator of OIRA to keep up the day-to-day working relationship with the agency head on behalf of the president.
From the day they are nominated, agency heads spend most of their time with agency staff, members and staff of congressional oversight committees, and regulated parties and other "stakeholders" in their agencies' decisions. Inevitably, they become subject to pressures that may conflict not only with each other but with the policies of the president. Before confirmation these pressures are brought to bear by senators who request a "courtesy call" from the nominee that takes the form of a chat in the senator's office, during which the senator seeks from the nominee a commitment to some policy he wants to see pursued or some constituent interest he wants protected. These interests may be reiterated at the nominee's confirmation hearing in order to get his commitment on the public record. (70)
The conflicting pressures continue once the agency head is confirmed. The career staff of the agency has an agenda that may or may not align with the preferences of the agency head. He will have implicitly to negotiate with the staff to elicit the proposals and regulations he wants and will inevitably have to compromise along the way. The staff will have already done substantial work upon some regulatory initiatives, and may have taken one particular direction and forsaken all others; indeed, the senior staff or a prior political appointee may have made commitments to stakeholders and congressional oversight committees that cannot as a practical matter be abandoned.
The trade press, interest groups, and congressional overseers will also exert considerable pressure upon the agency head. In many cases, those pressures will conflict with one another, and some if not all will conflict with the agency head's own preferences (not to mention the philosophy of the president). Somehow the agency head must harness the energy of the staff, satisfy the relevant congressmen, defend himself against leaks and other attacks in the trade press relevant to his programs, and yet maintain his credibility with all concerned. (71) It will be no surprise if, left to swim in these treacherous waters, the agency head loses sight, sooner or later, of the shore from which he set off to implement the president's general policy preferences.
In sum, the president has much less control over agency heads than he does over OIRA: he has less contact with agency heads, and those heads are also subjected to considerable pressure from sources other than the president. In fact, the president is just one, and often not the most influential, of the agency's constituencies; OIRA review helps bolster his standing.
The authors also argue that cost-benefit analysis, as practiced by OIRA, introduces a systematic bias toward under-regulation--quite apart from the cost-benefit "fallacies" discussed previously--because it is used only to determine whether the benefits of a particular regulation outweigh its costs and is not applied, or not frequently enough applied, to determine whether more stringent regulation would yield a still greater margin of benefits in excess of costs. (72) OIRA no doubt keeps unduly expensive regulations from being promulgated more often than it prompts an agency to increase the stringency of a proposed regulation; in this simple sense it reduces the burden imposed upon the economy by regulation, which by the authors' lights makes it "antiregulation."
But to say that OIRA review causes systematic under-regulation conflates the methods used to set the agendas of the regulatory agencies--that is, to determine ex ante what they will consider regulating and when--with the method OIRA uses ex post to determine whether a given regulation is worth implementing. So understood, to say that OIRA cost-benefit analysis is a "one-way ratchet" that causes systematic under-regulation puts the whole weight of efficient choice framing (or agenda setting) upon the method used for making a choice once framed (p. 153). At the same time, it shifts to OIRA a responsibility for which the regulatory agency has the comparative advantage by reason of its specialized knowledge of the field within its charge. Revesz and Livermore's argument is analogous to a claim that, because a produce inspector only tosses out the bad apples and does not simultaneously encourage additional apple growing, he is causing too few good apples to reach the market.
The authors accordingly suggest that OIRA review be reoriented "away from checking agencies and toward agenda-setting and the calibration of regulatory stringency" (p. 171). They do not, however, develop the implications of such a change. It would be virtually impossible for OIRA to reach back to the agency's agenda-setting stage and revise or create new rules. A major rule typically has a long gestation period. One study of major rules promulgated during the Carter and first Reagan Administrations showed that on average about eight years--and in some cases as many as twelve years--had elapsed between the initiation and the completion of the regulatory process, broadly conceived to encompass the early surveys, studies, and analyses undertaken before the agency committed itself to promulgate a regulation. (73) Those early steps are rarely irreversible but they generate momentum. By the time the agency is ready to issue an "Advance Notice of Proposed Rulemaking" the die may well have been cast. To be sure, the details of the ensuing regulation have not yet been decided, but the de facto commitment to regulate the subject matter has been made. When a draft proposed rule is eventually submitted to OIRA, both the program agency and the OIRA staff may have very limited options; all the more so when the program agency later sends the draft final rule to OIRA. (74)
Although we doubt (for reasons detailed below) the agencies' current methods for determining which regulations to consider leads to underregulation, we agree that a role for OIRA in setting agencies' regulatory agendas likely would improve the quality of their overall work product. For all the reasons that presidential review of final agency decisions is desirable, enhanced presidential control over agency agenda setting should be as well. But the best mechanism for agenda setting is a complicated question, one the authors do not explore in any depth. Centralized review of agency agenda setting has been tried. President Reagan established the strongest version in Executive Order No. 12,498, (75) which required each agency to submit a "Regulatory Program" to the director of the OMB, who could review each agency's draft program and recommend new regulatory or deregulatory activities he thought necessary to achieve consistency with the administration's policies and priorities. Any disagreement between the agency and the director was to be resolved by the president if the agency head wanted to pursue an appeal that far. President Clinton's Executive Order No. 12,866 (76) instituted a similar system, which is still in place but is less centralized in that the OMB no longer reviews and recommends changes to the agencies' published Regulatory Plans. (77) Instead it simply circulates the plans to any affected agencies, the CEA, and the vice president, who may raise any concerns with the issuing agency. In practice, most agencies supplement the required annual Regulatory Plan with a more comprehensive biannual regulatory agenda. (78)
Even if centralized regulatory review were biased against regulation, that bias is justified if, as we and most observers believe, regulatory agencies are themselves biased in favor of regulation. But the authors also reject the possibility that any bias toward under-regulation created by cost-benefit analysis or its implementation might be a useful check upon the correlative bias of the regulatory agencies. In their view, there is simply no reason to be concerned that the program agencies will be systematically inclined toward overzealous regulation. Indeed, they worry that "[i]f anything ... agencies tend towards inertia and inaction" (p. 165). This concern betrays a serious misunderstanding of the incentives facing the program agency.
Contemporary regulation is concerned almost exclusively with the mitigation of risks--risks to human health and the environment; risks to the solvency of financial institutions; and risks from hazardous products, automobiles, workplace conditions, and so on. During the early years of centralized regulatory review in the Carter and Reagan Administrations, there was also a great deal of economic regulation of business, much of which was harmful to consumers and socially wasteful. (79) It was economic regulation that gave rise to the widespread call for deregulation. (80) In contrast, neither President--indeed, no sensible person--ever proposed wholesale deregulation of risks in general or of environmental risks in particular. On the contrary, it was universally agreed both that serious environmental risks required regulation because polluters do not otherwise bear the costs they impose upon other users of the air and water, and that such regulation should be accomplished in the most cost-effective manner possible.
The authors' assertions to the contrary notwithstanding, there is good reason to believe agencies charged with risk regulation are systematically biased towards over-regulation because of a similar externality: risk regulators are judged by their output of rules, and punished when they fail to avert a risk that becomes manifest, but the costs of their rules are born by regulated firms and the general public. Delaying approval of a drug, for example, may impose substantial costs not only on the manufacturer, which typically will have hundreds of millions of dollars invested in it, but also on the public, in the form of avoidable morbidity and mortality. (81) If so, however, it is unlikely the agency will be criticized because those harmed ordinarily do not know they could have been helped by a drug not yet on the market; hence, the costs the agency will have imposed will not be attributed to the agency's perhaps excessive caution. If, on the other hand, the agency approves a drug that turns out to have significant untoward side effects, then the agency will be roundly criticized for its apparent imprudence. (82) Because of this asymmetry, risk-regulating agencies--the only significant regulators left standing--have a bias in favor of risk avoidance, regardless whether that is the welfare maximizing outcome for the public.
One should not be surprised, therefore, if the EPA attempts to suppress pollutants that cause respiratory ailments beyond the point where the benefits exceed the cost. If the EPA sensibly will not attempt to suppress "the last molecule," asthmatics are understandably going to complain to their congressmen that the EPA is not doing its job, regardless whether the agency has struck the right balance between the costs and benefits of further pollution controls. (83) On the other hand, if the EPA tightens the regulation to a point where costs exceed benefits, there will be fewer such complaints from asthmatics and more strenuous objections from polluters. The agency will often prefer to be criticized by a small number of polluters (and their congressmen) than by a large number of angry individuals (and their more numerous congressmen).
Neither are the authors correct in arguing that a disparity in interest-group pressure will tend toward under-regulation. The authors dismiss the notion that large numbers of aggrieved citizens can organize to exert pressure on regulatory agencies, whether directly or through their congressmen, to the same degree that the smaller number of polluters can do (pp. 163-64). By the authors' account there is no reason to believe that "special interest groups--such as environmental organizations and consumer protection groups--will have greater power in the administrative process than industry groups" (p. 163). They state:
[C]ollective-action theory predicts just the opposite: Environmental and consumer groups will have the most difficulty organizing themselves into effective lobbying blocks. Industry, with fewer players and more at stake for each firm, will have the advantage. That prediction is vindicated in our actual process in which environmental and consumer groups are dwarfed by industry trade associations, individual corporate lobbying, and the general probusiness and antitax lobbies. (pp. 163-64)
The authors cite no authority, empirical or otherwise, for their proposition about "our actual process," and with good reason. They reach their conclusion by sleight of hand, obscuring the inconvenient fact that environmentalists and consumers--greater difficulty notwithstanding--have managed to organize themselves into highly effective lobbying groups. One way or another, mainstream environmental organizations with millions of members have emerged in the last several decades and have succeeded in exerting substantial influence upon the legislative process. (84) Indeed, had they not done so, there is no way Congress would have passed many of our most protective environmental statutes. The key fact is that mainstream environmental organizations represent large numbers of people and those people vote. Trade associations representing "fewer players [with] more at stake for each firm," contrary to the authors, do not "have the advantage" (p. 164). Not only do they deliver fewer votes but their second-best substitute, political contributions, have been substantially limited by campaign finance legislation. In any dispute between environmental groups and oil refiners, forest product companies, or the like--whether in the Congress or before its agents such as the EPA--observation of "our actual process" teaches that both sides will be well represented and frequently influential. (85)
We do not contend that the regulatory agencies are dominated by risk-averse bureaucrats, fire-breathing zealots, and political appointees intent on mollifying staff and stakeholders at the expense of the president's policies. But we think it is naive not to recognize that the American administrative state has become a vast and complex enterprise, incorporating the shortcomings as well as advantages of specialization, and profuse with conflicting purposes and opportunities for the pursuit of individual and group self-interest--in sum, a reflection of the political system that created it. (86) In this environment, no politician who has mastered the system sufficiently to have gained the White House is going to be content to act as a neutral optimizer and coordinator of the regulations of dozens of agencies administering hundreds of statutes. Central oversight of agency rulemaking is but a modern extension of central oversight of agency budgets, inaugurated in the 1920s, and modern presidents are going to pursue the policy goals that got them elected through regulation as well as through budgeting. We are not surprised, therefore, by the first-hand report of Elena Kagan that President Clinton was even more assertive in putting his personal stamp on regulatory policy than President Reagan had been. (87) Readers of Retaking Rationality will have no doubt that a President Revesz and a Vice President Livermore would overcome their qualms about White House vulnerabilities and the "shaky foundation" of regulatory review and be more assertive still. They would combine cool rationality with hot rhetoric and use a capacious form of cost-benefit analysis to pursue an aggressively "proregulation" agenda.
In the end, Revesz and Livermore have combined two different books in crafting Retaking Rationality. One is a spirited political polemic directed to environmental and other proregulation activists, aiming to persuade them that they can advance their causes by joining rather than spurning the cost-benefit debates. The other is an intelligent if uneven survey of the difficult questions in cost-benefit analysis. By combining the two, however, Revesz and Livermore have created an amalgam that is less than the sum of its parts.
In their discussion of cost-benefit analysis, the authors time and time again put forward a thorough analysis of some tough academic issue, only to revert to political polemic and leave the reader wondering why there is a debate if the answers are so easy. The fact is that the answers are not as easy as Revesz and Livermore make them out to be, down to the last "fallacy."
By the time they turn to "Rethinking OIRA" and the purpose of centralized regulatory review, Revesz and Livermore have all but abandoned their scholarly mantle in favor of personal policy prescriptions. Their case against OIRA is essentially that it enables the president, through the device of cost-benefit analysis, to exert an antiregulatory influence upon the rulemaking output of his administration. Instead, they would recast OIRA not as a servant of the president but as a machine to calibrate the optimal stringency of each regulation and "prioritize the non-cost-benefit analysis function of OIRA, including interagency coordination and harmonization." (p. 171). We remain doubtful that should or will happen for the reason we gave more than 20 years ago: "[E]very president has a program, and no program can be implemented in the modern regulatory state without regulatory planning and regulatory review by the Executive Office of the President." (88)
Most recently Cass R. Sunstein, President Obama's OIRA administrator, made the point this way in his Senate confirmation hearings:
The foundation of regulatory review--the first question to be asked by the Office of Information and Regulatory Affairs is what are Congress's instructions? That is the starting point for any mechanism for regulatory review. The second task is to ensure that within the boundaries set by Congress, things done are consistent with the President's own priorities and principles. The third task is to kind of institutionalize the notion of looking before you leap so that when the government is starting with a regulation, whether it involves homeland security, education, energy, or anything else, there's some sense of what the consequences are likely to be. That promotes accountability. It helps ensure that citizens and government can know what the likely effects of government action are. (89)
Thus centralized regulatory review aims to ensure that regulations are consistent with law and, to the extent permitted by law, consistent with the policies of the president. White House review and cost-benefit analysis can inform, but they cannot displace, the political judgments essential to regulatory policymaking.
(1.) Revesz is faculty director and Livermore is executive director of the New York University School of Law's Institute for Policy Integrity, which they founded to advocate the teachings of Retaking Rationality. The Institute's website is http://www.policyintegrity.org.
(2.) The authors of this Review were administrators of the Office of Information and Regulatory Affairs ("OIRA") in the United States Office of Management and Budget ("OMB"), which oversees agency regulations under statutory and Executive Order authorities, during the Reagan Administration when an explicit cost-benefit standard was first being applied (DeMuth in 1981-84 and Ginsburg in 1984-85). We later wrote an article explaining and defending the OIRA review process, Christopher C. DeMuth & Douglas H. Ginsburg, White House Review of Agency Rulemaking, 99 HARV. L. REV. 1075 (1986). Our tenures and our article both come in for criticism in Retaking Rationality. Pp. 23-29, 165-68.
(3.) In the authors' words, cost-benefit analysis is "necessary to improve the quality of decisions that regulators make," p. 3, but "[b]ecause current cost-benefit analysis is biased against regulation, efficient and worthwhile regulations do not pass the test .... The result? Less safety, worse health, and a dirtier environment." P. 11. The argument recurs throughout the book and is presented in stronger terms in the book's publicity release. See infra note 21.
(4.) Party affiliations are never mentioned, but the pattern is obvious. President Ronald Reagan's cost-benefit Executive Order and OMB review process "led to fears that industries would be able to kill regulations contrary to their interests under cover of night," and "[t]hese fears were largely vindicated," p. 27; President George H. W. Bush's White House Council on Competitiveness "'was sharply critical of any regulation and deeply solicitous of business interests," p. 30; when President Bill Clinton established a new Executive Order and modified the OMB review process, "[t]he message from the Clinton White House was that centralized review and cost-benefit analysis could serve as a neutral tool," p. 32; but then the dark times returned--"[u]nder President George W. Bush, the link between the deregulatory agenda and cost-benefit analysis has become nearly complete." P. 42. The contrasts are elaborated throughout the book. Retaking Rationality was published in 2008, before the presidential election of that year, but the authors anticipate the return of rationality: "With a new administration entering office in January 2009, [progressive proregulation] groups may again have an opportunity to participate," p. 12. And there is hope that these groups might do so--"[H]ope is always at a zenith during a presidential election cycle, especially if no incumbent is running. It is during these times that many things seem possible, and the transformative power of democratic politics is at its maximum.'" P. 193.
(5.) As in candidate Barack Obama's charge, during the 2008 presidential campaign, that the unfolding financial crisis was the result of "Bush deregulation"; but even that turned out not to be particularly partisan, because his opponent, John McCain, seemed to agree with him. See Adam Nagourney, Economic Woes Set Tone for Rivals in 2nd Debate, N.Y. TIMES, Oct. 8, 2008, at Al (noting that while Obama placed blame for the crisis squarely on the Bush Administration, McCain left these charges unanswered and characterized his own senatorial record as being "sometimes at odds" with Bush).
(6.) Justices Stephen G. Breyer and Antonin Scalia, whose differences on many jurisprudential issues are well known, provide a good example of the liberal-conservative congruence on regulatory policy. Justice Breyer played an instrumental role in the abolition of the Civil Aeronautics Board and wrote approvingly of the results. Stephen G. Breyer, Antitrust, Deregulation, and the Newly Liberated Marketplace, 75 CAL. L. REV. 1005, 1010-11 (1987) (arguing that as a result of airline deregulation, "[c]ompetition has forced a better match between supply and demand, leading to greater efficiency and increased satisfaction of consumer preferences"); see also Richard B. Stewart, Tribute to Justice Stephen G. Breyer, 64 N.Y.U. AYN. SURV. AM. L. 19, 21 (2008) (discussing the leadership of Senator Kennedy, Justice Breyer, and Professor Kahn in airline deregulation). He has also been a proponent of reforming health, safety, and environmental regulation. See STEPHEN BREYER, BREAKING THE VICIOUS CIRCLE: TOWARD EFFECTIVE RISK REGULATION (1993); STEPHEN BREYER, REGULATION AND ITS REFORM 131-55, 261-84 (1982). Justice Scalia was general counsel of the United States Office of Telecommunications Policy (which promoted telecommunications deregulation) from 1972 to 1974, and from 1977 to 1981 was coeditor-in-chief (along with Murray Weidenbaum, described by Revesz and Livermore as a particularly zealous deregulator) of the American Enterprise Institute's magazine, REGULATION, which consistently advocated abolition of economic regulation in competitive markets and improvement-through-reform of health, safety, and environmental regulation. See Archive of REGULATION issues, www.cato.org/pubs/regulation/ regultn_past_issues.html (last visited Nov. 14, 2009). For additional examples, see the archives of the American Enterprise Institute-Brookings Institution Joint Center for Regulatory Studies (1996-2007), http://www.reg-markets.org/publications/index.php?menuid=3, featuring hundreds of papers on economic deregulation and health-safety-environmental regulatory reform by scholars from across the philosophical spectrum.
(7.) A prominent exception to the academic consensus on regulatory reform is FRANK ACKERMAN & LISA HEINZERLING, PRICELESS: ON KNOWING THE PRICE OF EVERYTHING AND THE VALUE OF NOTHING (2004). Although Ackerman and Heinzerling, like Revesz and Livermore, inveigh against particular techniques of cost-benefit analysis, their essential position is rejectionist: they do not want to regulate within resource constraints but rather to fight for more resources for the things they favor. For responses to PRICELESS, see ROBERT W. HAHN, IN DEFENSE OF THE ECONOMIC ANALYSIS OF REGULATION (2005); Cass R. Sunstein, Your Money or Your Life, NEW REPUBLIC, Mar. 15, 2004, at 27 (book review), available at http://www.tnr.com/article/books-andarts/your-money-or-your-life.
(8.) Cf. Christopher DeMuth, Regulator. Policy in the Reagan Administration, in AMERICAN ECONOMIC POLICY IN THE 1980s 504, 508-10 (Martin Feldstein ed.) (1994) ("[i]t is an interesting puzzle that economists have been much more successful in persuading others of the evils of price fixing than of the evils of quality fixing [product standardization]"). Compare SAM PELTZMAN, REGULATION AND THE NATURAL PROGRESS OF OPULENCE (2005), available at http://www. aei.org/book/821?action=add&id=821, with RICHARD H. THALER & CASS R. SUNSTEIN, NUDGE: IMPROVING DECISIONS ABOUT HEALTH, WEALTH, AND HAPPINESS (2008).
(9.) Price and entry controls are not covered in the book, although the authors occasionally confuse the distinction between deregulation and regulatory reform by using the term "economic regulation" when they seem to mean environmental regulation. Pp. 9, 10.
(10.) See R. SHEP MELNICK, REGULATION AND THE COURTS; THE CASE OF THE CLEAN AIR ACT 149-51 (1983).
(11.) Pub. L. No. 96-51 l, 94 Stat. 2812 (1980) (codified as amended at 44 U.S.C. [section] [section] 3501-3521 (2006)).
(12.) See Christopher C. DeMuth, Constraining Regulatory Costs--Part I: The White House Review Programs, REG., Jan./Feb. 1980, at 13, 18-20; Richard M. Neustadt, Taming the Paperwork Tiger: An Experiment in Regulatory Management, REG., Jan./Feb. 1981, at 28, 32; Janet A. Weiss, The powers of problem definition: The case of government paperwork, 22 POL'Y SCI. 97, 106-07 (1989).
(13.) Revesz and Livermore say that "[e]ven Ronald Reagan did not stop the phaseout of leaded gasoline." E 18. In fact, the Reagan Administration significantly tightened the lead phasedown rules, with the wholehearted support of OIRA and one of the authors of this review, and added a marketable-permits policy that further accelerated the removal of lead from the U.S. gasoline supply. See DeMuth, supra note 8, at 508; Richard G. Newell & Kristian Rogers, The Market-Based Lead Phasedown, in MOVING TO MARKETS IN ENVIRONMENTAL REGULATION: LESSONS FROM TWENTY YEARS OF EXPERIENCE 171, 171-78 (Jody Freeman & Charles D. Kolstad eds., 2007). On the enforcement point, see Philip Shabecoff, E.P.A. Record Set in Pollution Cases, N.Y. TIMES, Dec. 9, 1988, at A24.
(14.) Revesz and Livermore's discussion at pp. 136-37 of the sulfur dioxide trading program offers no credit to Bush. On the provenance of the program, see John M. Broder, From a Theory to a Consensus on Emissions, N.Y. TIMES, May 17, 2009, at A1.
(15.) See, e.g., Randall Lutter & Richard B. Belzer, EPA Pats Itself on the Back, REG., Fall 2000, at 23, 27; John H. Cushman, Jr., E.P.A. Advocating Higher Standards to Clean the Air, N.Y. TIMES, Nov. 25, 1996, at Al; John H. Cushman, Jr., Top E.P.A Official Not Backing Down on Air Slandards, N.Y. TIMES, June 1, 1997, at A1.
(16.) Energy Independence and Security Act of 2007, Pub. L. No. 110-140, 121 Star. 1492, available at http://frwebgate.access.gpo.gov/cgi-bin/ getdoc.cgi?dbname=ll0_cong_public_laws& docid=f:publl40.110.pdf; Press Release, The White House, Fact Sheet: Twenty in Ten: Strengthening Energy Security and Addressing Climate Change (May 14, 2007), available at http:// georgewbush-whitehouse.archives.gov/news/releases/2007/ 05/20070514-2.html.
(17.) See Control of Emissions of Air Pollution From Nonroad Diesel Engines and Fuel; Final Rule, 69 Fed. Reg. 38957 (June 29, 2004) (codified at 40 C.F.R. pts. 9, 69, 80, 86, 89, 94, 1039, 1048, 1051, 1065, 1068); Standards of Performance for New and Existing Stationary Sources: Electric Utility Steam Generating Units; Final Rule, 70 Fed. Reg. 28605 (May 18, 2005) (codified at 40 C.F.R. pts. 60, 72, 75); National Primary Drinking Water Regulations for Lead and Copper: Short-Term Regulatory Revisions and Clarifications; Final Rule, 72 Fed. Reg. 57781 (Oct. 10, 2007). John Graham, administrator of OIRA for President Bush from 2001 through 2006, provides an illuminating account of regulatory policymaking in the Bush administration in JOHN D. GRAHAM, BUSH ON THE HOME FRONT: DOMESTIC POLICY TRIUMPHS AND SETBACKS chs. 7, 8, and 10 (forthcoming 2010). The EPA mercury rule was less stringent than a Clinton Administration "midnight regulation" would have been and was opposed by environmental groups (it was also, like many George W. Bush Administration rules discussed in this Review, a legal botch--vacated in New Jersey v. EPA, 517 F.3d 574 (D.C. Cir. 2008)). But, as Graham explains, the controversy was dominated by regional conflicts, not partisan conflicts. See GRAHAM, supra at 210-14.
(18.) The one discernable pattern is that, within administrations of both parties, rulemaking, and especially the issuance of very costly new rules, tends to increase as the administration ages and crescendos in its final year. See U.S. Office of Mgmt. and Budget, 2008 Report to Congress on the Benefits and Costs of Federal Regulations and Unfunded Mandates on State, Local, and Tribal Entities 45-47 figs.2-1 & 2-2 (2008), available at www.whitehouse.gov/omb/assets/information and regulatory_affairs/2008 cb final.pdf; cf. Clyde Wayne Crews, Jr., Competitive Enterprise Inst., Ten Thousand Commandments: An Annual Snapshot of the Federal Regulatory State (2009), available at http://cei.org/cei_files/fm/active/0/Wayne%20Crews%20-%2010, 000%20Commandments% 202009.pdf: Veronique de Rugy & Melinda Warren, Mercatus Ctr. at George Mason Univ. & Murray Weidenbaum Ctr. on the Econ., Gov't, and Pub. Pol'y at Washington Univ. in St. Louis, Expansion of Regulatory Budgets und Staffing Continues in the New Administration: An Analysis of the U.S. Budget for Fiscal Years 2009 and 2010 (2009), available at http://regbudget.mercatus.org/wp-content/uploads/2009/l0/ Regulators-Budget-Report-Final-VersionOctober-29.pdf; Robert W. Hahn, AEI Ctr. for Reg. & Mar. Stud., An Analysis of the 2008 Government Report on the Costs and Benefits of Federal Regulations (2008), available at http://regmarkets.org/admin/authorpdfs/redirect-safely.php? fname=../pdffiles/phpnZ.pdf.
(19.) See the articles by John D. Graham and Paul R. Roe, Cary Coglianese, Jerry L. Mashaw, and Roger G. Noll in Symposium, Reflections on Executive Order 13, 422, 25 YALE J. ON REC. 77 (2008), and Stuart Shapiro, Presidents" and Process: A Comparison of the Regulatory Process Under the Clinton and Bush (43) Administrations, 23 J.L. & POL. 393, 417 (2007) ("One trend is consistent through the various data: despite profound differences in ideology and widespread perceptions of difference in the substance of their regulatory efforts, the regulatory process looks very similar between the Bush and Clinton Administrations.").
(20.) See HAHN, supra note 18.
(21.) We note with dismay that Oxford University Press's publicity release for Retaking Rationality proclaims:
That America's natural environment has been degraded and despoiled over the past 25 years is beyond dispute. Nor has there been any shortage of reasons why--short-sighted politicians ... and the dramatic weakening of environmental regulations. In Retaking Rationality, Richard Revesz and Michael Livermore argue convincingly that one of the least understood--and most important--causes of our failure to protect the environment has been a misguided rejection of reason.
Oxford University Press, Description of Retaking Rationality, www.oup.com/us/catalog/general/ subject/Law/EnvironmentalLaw/?view=usa&ci=9780195368574 (last visited Nov. 14, 2009). In fact, environmental quality has been improving throughout the period since the establishment of the EPA and the first modern pollution control statutes in the early 1970s. See, e.g., JOEL M. SCHWARTZ & STEVEN E HAYWARD, AIR QUALITY IN AMERICA (2007); PACIFIC RESEARCH INST., INDEX OF LEADING ENVIRONMENTAL INDICATORS (1996--2009), available at http://environment.pacific research.org/index-of-leading-environmental-indicators.
(22.) Our characterization is itself overbroad: industry groups are typically much less united than environmental groups, because firms may favor stricter regulation in order to gain private competitive advantages. See Jonathan H. Adler, Rent Seeking Behind the Green Curtain, REG., Fall 1996, at 26; Kimberly A. Strassel, Op-Ed., Rent-Seekers Inc., WALL ST. J., Oct. 2, 2009, at A17.
(23.) P. 4. See also p. 51.
(24.) Retaking Rationality begins and ends with thoughtful admonitions that cost-benefit analysis is not "a master decisionmaking procedure" p. 15, or a "panacea" and "must also be augmented with discussions of our values, political debate, and analysis of non-cost-benefit factors," p. 190. But, as we shall see, the intervening discussions seek to introduce a great many political and values questions into cost-benefit calculations.
(25.) The rationalist conviction is plumbed and criticized in MICHAEL OAKESHOTT, Rationalism in Politics, in RATIONALISM IN POLITICS AND OTHER ESSAYS 1 (Liberty Press 1991) (1962).
(26.) According to the authors, one study found that carbon monoxide controls reduced deaths by accident and suicide by an average of 25,000 per year, as compared to 212 to 551 deaths (from cancer) resulting from lower air pollution. The 25,000 figure is impossibly high--total U.S. suicides have been in the range of 30,000 per year. See Centers for Disease Control and Prevention, Suicide and Self-Inflicted Injury, http://www.cdc.gov/nchs/fastats/suicide.htm. The only study cited found much lower ancillary benefits (although still much higher than direct benefits): annual reductions of 2000-3000 accidental and suicidal deaths. And the averted suicides were from exhaust inhalation; some people who would commit suicide by CO inhalation when the technique is available will find other means of taking their lives, so the net suicides averted by CO controls would be less than the reduction in exhaust suicides. This is an example of a risk tradeoff; the study mentioned the tradeoff but did not attempt to estimate it. See p. 59; M. Shelef, Unanticipated benefits of automotive emission control: reduction in fatalities by motor vehicle exhaust gas, 146/147 Sci. TOTAL ENV'T 93, 98-100 (1994).
(27.) See Christopher DeMuth, AM. ENTER. INST. FOR PUB. POL'Y RESEARCH, Unintended Consequences and Intended Non-Consequences (2009), available at http://www.aei.org/docLib/ 20090608-Bradley-June.pdf.
(28.) Pp. 67-68. But the authors note darkly that OIRA may have continued to be influenced by health-wealth thinking behind the scenes and "[w]e do know that health-wealth tradeoff analysis has not disappeared from the public debate."
(29.) The cases are cited and two of them quoted at pp. 70-71.
(30.) For a careful empirical study finding that job displacement, especially when accompanied by substantial loss of income, results in substantial short- and long-run increases in mortality, see Daniel Sullivan & Till von Wachter, Job Displacement and Mortality: An Analysis Using Administrative Data, 124 Q.J. ECON. 1265 (2009).
(31.) Am. Dental Ass'n v. Martin, 984 F.2d 823, 826 (7th Cir. 1993).
(32.) But Judge Posner does indeed subscribe to the health-wealth tradeoff. See RICHARD A. POSNER, ECONOMIC ANALYSIS OF LAW 394-95 (7th ed. 2007).
(33.) Revesz and Livermore assert that the 1990s studies relied on by proponents of a health-wealth tradeoff attribute all of the correlation to wealth causation. Pp. 69, 71-72. They are mistaken: those studies, and the rich literature on the subject that has developed since, are concerned not only with measuring the correlation but with understanding how much of it is caused by differences in wealth, health, education, and other factors.
(34.) A debatable proposition cannot be pseudo logic and a sham devoid of meaningful content. This is one of many instances in Retaking Rationality, of overbroad assertions later undermined to an unexplained degree by an inconsistent concession.
(35.) Robert Hahn & Scott Wallsten, Whose Life Is Worth More? (And Why Is It Horrible to Ask?), WASH. POST, June 1, 2003, at B3. See generally John Tierney, Life: The Cost-Benefit Analysis, N.Y. TIMES, May 18, 2003, [section] 4, at 14.
(36.) Cass R. Sunstein, Lives, Life-Years. and Willingness to Pay, 104 COLUM. L. REV. 205, 214-16 (2004).
(37.) See, e.g., id. at 252 (listing several case examples in the appendix); cf. INST. OF MEDICINE OF THE NAT'L ACADS., VALUING HEALTH FOR REGULATORY COST-EFFECTIVENESS ANALYSIS (Wilhelmine Miller et al. eds., 2006) (providing recommendations regarding the measurement of health and safety improvements using cost-effective analysis). Forbidding life-years-saved estimates is particularly odd where regulations have both morbidity and mortality benefits: morbidity benefits are perforce measured by extended years of life, so the best way to characterize aggregate mortality and morbidity benefits is by measuring mortality benefits in the same manner. See. e.g., U.S. ENVTL. PROTECTION AGENCY, EPA 815 R-05-010: APPENDICES TO THE ECONOMIC ANALYSIS FOR THE FINAL STAGE 2 DISINFECTANTS AND DISINFECTION BYPRODUCTS RULE, VOLUME III (I-N) (2005), available at http://www.epa.gov/safewater/disinfection/ stage2/pdfs/analysis_stage2_ economic_appendix3.pdf: U.S. ENVTL. PROTECTION AGENCY, EPA 815-R-06-001: APPENDICES TO THE ECONOMIC ANALYSIS FOR THE FINAL LONG TERM 2 ENHANCED SURFACE WATER TREATMENT RULE, VOLUME II (H-U) (2005), available at www.epa.gov/safewater/ disinfection/lt2/pdfs/ anaylis_lt2_economic_appendix2.pdf.
(38.) Food Labeling; Trans Fatty Acids in Nutrition Labeling, Nutrient Content Claims, and Health Claims; Final Rule, 68 Fed. Reg. 41, 434, 41, 487-90 (July 11, 2003); Prevention of Salmonella Enteritidis in Shell Eggs During Production, Storage, and Transportation; Final Rule, 74 Fed. Reg. 33,029, 33,054-56 (July 9, 2009).
(39.) Regulations Restricting the Sale and Distribution of Cigarettes and Smokeless Tobacco to Protect Children and Adolescents; Final Rule, 61 Fed. Reg. 44,395, 44,568-69 (Aug. 28, 1996) (codified at 21 C.F.R. pts. 801,803, 804, 807, 820, 897).
(40.) FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120 (2000).
(41.) See Family Smoking Prevention and Tobacco Control Act of 2009, Pub. L. No. 111-31, 123 Stat. 1776, 1786-87, available at http://acscan.org/pdf/ protectkids/Family-Smoking-Preventionand-Tobacco-Control- Act.pdf.
(42.) We are, however; happy to see continued refinement of willingness-to-pay estimates for reductions in risks to different age groups.
(43.) We know that government quality-of-life rulings are unpopular from Oregon's experience with rationing medical care according to patients' health status (discussed in RETAKING RATIONALITY at pp. 85-87) and from the strong public reaction, in the 2009 debates over the Obama Administration's health-care proposals, to the prospect of similar rationing at the national level. See Scott Gottlieb & Elizabeth DuPre, The Living Truth about "Death Panels", HEALTH POL'Y OUTLOOK (Am. Enter. Inst., Wash., D.C.), Oct. 7, 2009, available at www.aei.org/outlook/100073.
(44.) But discounting issues are ubiquitous in regulatory policy. Energy efficiency regulations, for example, trade higher up-front product prices for lower energy expenditures over the life of the products. See Gloria Helfand & Ann Wolverton, Evaluating the Consumer Response to Fuel Economy: A Review of the Literature (U.S. Envtl. Prot. Agency, Nat'l Ctr. for Envtl. Econ., Working Paper No. 09-94, 2009).
(45.) 947 F.2d 1201 (5th Cir. 1991).
(46.) Id. at 1218-19.
(47.) The court's decision was based on the EPA's failure to follow statutory requirements and judicial precedent (under the Toxic Substances Control Act) and substantial irregularities in the rulemaking proceedings and record, all unrelated to the issue of discounting future benefits. The criticism of the EPA's failure to discount latency-period benefits was one of several "concerns" expressed by the court "[i]n order to aid the EPA's reconsideration of this and other cases." ld.
(48.) Circular A-4 from U.S. Office of Mgmt. and Budget to the Heads of Executive Agencies and Establishments 32 (Sept. 17, 2003), available at http://georgewbush-whitehouse. archives.gov/omb/circulars/a004/a-4.pdf.
(49.) This however merely recasts the dilemma of projecting our current knowledge over long periods into the future.
(50.) The authors immediately repeat the point: "We cannot punt on this issue, and must face up to the fundamentally moral question of what we owe future generations." R 146.
(51.) The authors dismiss as simply "wrong" the suggestion of Cass Sunstein and Arden Rowell that intergenerational benefits be discounted and ethical and political obligations to future generations be considered separately. P. 110. But the Revesz-Livermore and Sunstein-Rowell approaches to ethical and distributive issues are similar, so the difference between right and wrong is whether benefits estimates are discounted at all. Compare pp. 116-17, 146, with Cass R. Sunstein & Arden Rowell, On Discounting Regulatory. Benefits: Risk, Money, and Intergenerational Equity, 74 U. CHI. L. REV. 171 (2007) (part of a superb symposium volume on intergenerational equity and discounting) and p. 110.
(52.) Pp. 107-09. It is worth noting, however, that the climate-change debates concerning benefit discounting have been vigorous and sophisticated and have not obscured, but rather have illuminated, the issues of ethical obligations to future generations, uncertainties over future states, and the opportunity costs of immediate regulation to populations that would be particularly affected by climate change. See, e.g., WILLIAM NORDHAUS, A QUESTION OF BALANCE: WEIGHING THE OPTIONS ON GLOBAL WARMING POLICIES (2008); THOMAS C. SCHELLING, COSTS AND BENEFITS OF GREENHOUSE GAS REDUCTION (1998); NICHOLAS STERN, THE ECONOMICS OF CLIMATE CHANGE: The Stern Review (2007), available at http://www.hm-treasury.gov.uk/ independent-reviews/stern_ review_economics_climate_change/sternreview_index.cfm; Thomas C. Schelling, What Makes Greenhouse Sense?, FOREIGN AFF., May/June 2002, at 2 (2002); David Weisbach & Cass R. Sunstein, Climate Change and Discounting the Future: A Guide for the Perplexed, 27 YALE L. & POL'Y REV. 433 (2009); William D. Nordhaus, A Review of the Stern Review on the Economics of Climate Change, 45 J. ECON. LIT. 686 (2007) (book review).
(53.) P. 121. The Supreme Court subsequently reduced one major assessment arising from the Exxon-Valdez oil spill punitive damages in a class-action suit under maritime law. See Exxon Shipping Co. v. Baker, 128 S. Ct. 2605, 2634 (2008). But Revesz and Livermore did not even consider the stakes in that case in comparing tangible damages with total payments to date; so the case's conclusion reinforces their point, but by less than it might have.
(54.) P. 121, n. 358. See U.S. OFFICE OF MGMT. BUDGET, STIMULATING SMARTER REGULATION: 2002 REPORT TO CONGRESS ON THE COSTS AND BENEFITS OF REGULATIONS AND UNFUNDED MANDATES ON STATE, LOCAL AND TRIBAL ENTITIES 47-49, 50 tbl.9 (2002).
(55.) See INDUR GOKLANY, CLEARING THE AIR: THE REAL STORY OF THE WAR ON AIR POLLUTION (1999); PELTZMAN, supra note 8; DeMuth, supra note 27.
(56.) The effort to relax New Source Review requirements for engineering renovations would be an accurate example of a regulatory rollback (by the authors' lights). But it is offered not as an example of a cost-benefit fallacy but rather of the inconsistency of OIRA review (because no cost-benefit analysis was performed for this "deregulatory" initiative). In any event, the New Source requirements were never relaxed: the rules were vacated in New York v. EPA, 443 F.3d 880 (D.C. Cir. 2006) and New York v. EPA, 413 F.3d 3 (D.C. Cir. 2005).
(57.) See GRAHAM, supra note 17. at 201-04; Randall Lutter, Rationalizing Air Pollution, REG., Spring 2002, at 13.
(58.) A key dilemma was that improved efficiency in power generation could lead to increased output and therefore increased pollution even when pollution per unit of output was reduced.
(59.) See Juliet Eilperin, Bush Pollution Curbs Are Rated Equal to Clinton's, WASH. POST, July 22, 2006, at A5. This effort, too, came to naught. See North Carolina v. EPA, 531 F.3d 896 (D.C. Cir 2008). Revesz and Livermore do not mention this critical aspect of the New Source Re view initiative.
(60.) P. 166 (citing DeMuth & Ginsburg, supra note 2, at 1082).
(61.) DeMuth & Ginsburg, supra note 2, at 1081.
(62.) Steven Croley, White House Review of Agency Rulemaking: An Empirical Investigation, 70 U. CHI. L. REV. 821, 860, 875 (2003). Croley found some conflicting evidence concerning OIRA's sensitivity to interest group pressure, but found that "[m]ost of the data [in the study] go farther to assuage fears about presidential involvement in rulemaking than they do to accentuate those fears." Id. at 879.
(63.) Nicholas Bagley & Richard L. Revesz, Centralized Oversight of the Regulatory State, 106 COLUM. L. REV. 1260, 1268 n.38, 1289 n.182, 1298 n.231, 1305 n.264, 1307 nn.275, 277 & 279, 1310 n.294 (2006).
(64.) "It is nearly as difficult for the President to personally keep track of the goings-on at OIRA as it would be to keep track of the internal workings of a regulatory agency." P. 167.
(65.) OIRA's annual reports typically report on the regulations of about thirty executive branch agencies and another nine so-called "independent" agencies. See, e.g., U.S. OFFICE OF MGMT. AND BUDGET, DRAFT 2009 REPORT TO CONGRESS ON THE BENEFITS AND COSTS OF FEDERAL REGULATIONS AND UNFUNDED MANDATES ON STATE, LOCAL, AND TRIBAL ENTITIES (Sept. 21, 2009), available at www.whitehouse.gov/omb/inforeg_regpol_reports_congress.
(66.) The current OIRA administrator, Cass R. Sunstein, is himself a friend and former academic colleague of President Obama. In our experience, the OIRA administrator is closely involved in the selection of regulatory agency heads when vacancies occur.
(67.) See, e.g., ROWLAND EVANS, JR. & ROBERT NOVAK, THE REAGAN REVOLUTION 142 (1981) (noting that, while campaigning, Reagan "blamed empty steel mills and long lines of unemployed on EPA zeal"); James F. Blumstein, Regulatory Review by the Executive Office of the President: An Overview and Policy Analysis of Current Issues, 51 DUKE L.J. 851, 859 (2001) (finding that deregulation was a "centerpiece" of Reagan's presidential campaign). President Carter's campaign bad a similar theme. See, e.g., Andrew Downer Crain, Ford, Cartel; and Deregulation in the 1970s. 5 J. ON TELECOMM. & HIGH TECH. L. 413, 431 (2007) ("During his presidential campaign. Carter promised to pursue deregulation."); see also Mark A. Goldberg, Theory and Practice, 25 YALE J. ON REG. 326, 326 (2008) ("President Carter had promised, during his successful campaign in 1976, to reorganize the federal bureaucracy, to increase its effectiveness and efficiency, and to streamline regulation.").
(68.) See, e.g., BLUEPRINT FDR CHANGE: OBAMA AND BIDEN'S PLAN FOR AMERICA (2008), available at http://www.barackobama.com/pdf/ObamaBlueprintForChange.pdf (proposing in general terms more regulation of agriculture, antitrust, finance, consumer credit, education, the environment, energy, health care, housing, and the workplace).
(69.) JAMES P. PFIFFNER, BROOKINGS INST., RECRUITING EXECUTIVE BRANCH LEADERS: THE OFFICE OF PRESIDENTIAL PERSONNEL (2001), http://www.brookings.edu/ articles/2001/spring_ governance_pfiffner.aspx ("The obligations of the OPP are threefold--to serve the nation by recruiting executive branch leaders, to serve the president by finding qualified loyalists, and to shepherd nominees through the sometimes treacherous appointment process.").
(70.) Illustrative is the experience of President Obama's nominee to be administrator of the EPA. At her confirmation hearing, Senator Ben Cardin of Maryland began his statement as follows:
I've had the chance to talk with ... you in my office so you know the first issue I'm going to bring up, and that is the Chesapeake Bay.... We don't want to see press releases, we want to see results in the cleaning up of the bay.... I thank you for your willingness to make that a priority of your agenda.
Nomination of Lisa Jackson to be Administrator of the Environmental Protection Agency: Hearing Before the S. Env't and Pub. Works Comm., 111th Cong., at *14 (2009) (Statement of Sen. Ben Cardin) (LEXIS, Federal News Service File).
(71.) This problem of the "iron triangle" was famously discussed in THEODORE J. LOWI, THE END OF LIBERALISM 33-35, 58 (2d ed. 1979).
(72.) "OIRA mostly seeks to ensure that the agency regulation is not too stringent, and does not impose higher economic costs than are justified.... Because there is no comparable formalized procedure geared to increasing regulatory stringency, regulations tend to be less stringent than would be economically efficient." P. 153.
(73.) These data, from an unpublished study done during the Reagan Administration, are not out of line with subsequent published reports. See, e.g., Shapiro, supra note 19, at 415-16 (finding that the time between inclusion of a rule in the annual Unified Agenda, which comprises regulations the agency believes may be proposed that year, and final rulemaking averaged 932 days during the Clinton Administration and 1,073 days during the W. Bush Administration, not including the years it may take to study, survey, and analyze an issue before deciding to go forward with rulemaking).
(74.) On a related point, Shapiro suggests that during the last two administrations the procedures through which centralized regulatory review was conducted had very little to do with the substance, stringency, or number of regulations developed, a finding at odds with Revesz and Livermore's argument that the "institutions of regulatory review" are "biased against regulation." P. 171. Shapiro shows that in spite of using similar review procedures, Presidents Clinton and Bush presided over very different regulatory states, both in terms of the amount and the nature of the regulations issued. See generally Shapiro, supra note 19.
(75.) 3 C.F.R. 323 (1985) (repealed 1993), reprinted in 5 U.S.C. [section] 601 (1988).
(76.) 3 C.F.R. 638 (1993), reprinted in 5 U.S.C. [section] 601 (1994).
(77.) For a thorough overview, see Peter M. Shane, Political Accountability in a System of Checks and Balances: The Case of Presidential Review of Rulemaking, 48 ARK. L. REV. 161, 179-81 (1995).
(78.) See RegInfo.gov, Unified Agenda and Regulatory Plan, http://www.reginfo.gov/public (last visited Nov. 15, 2009).
(79.) See supra notes 8, 67 and accompanying text.
(80.) E.g., ROBERT E. LITAN & WILLIAM D. NORDHAUS, REFORMING FEDERAL REGULATION (1983) (proposing a series of regulatory reforms).
(81.) There is extensive literature elaborating upon and attempting to estimate the cost of delay in the drug approval process. For an overview, see Patricia M. Danzon, The Pharmaceutical Industry, in 3 ENCYCLOPEDIA OF LAW AND ECONOMICS 1055, 1058-61 (Boudewijn Bouckaert & Gerrit De Geest eds., 2000), available at http://encyclo.findlaw.com/ 5880book.pdf.
(82.) The favorable press that the FDA and Dr. Frances O. Kelsey received for the decision to delay approval of Thalidomide is one prominent example of the attention paid to the side effects of approved drugs as opposed to the costs imposed upon others by the unavailability of delayed drugs. Thalidomide had been approved as a treatment for morning sickness during pregnancy in many countries throughout Europe but was later determined to cause severe birth defects. For example, in West Germany in 1962 alone between 3500 and 5000 babies were born with phocomelia, a deformity caused by Thalidomide. Ashley Ochs, Comment, A Study in Futility: Abigail Alliance for Better Access to Experimental Drugs Will Not Expand Access to Experimental Drugs for the Terminally Ill, 39 SETON HALL L. REV. 559, 565 (2009). Dr. Kelsey, the FDA reviewer assigned to the drug, delayed its approval in the United States, thereby averting a certain tragedy, for which she was awarded the Presidential Award of Distinguished Federal Civilian Service by President John E Kennedy. The episode made national headlines, and the story remains a "powerful  memory" influencing the American public's demand for safe drugs decades later. Elizabeth M. Rutherford, The FDA and "Privatization"--The Drug Approval Process, 50 FOOD & DRUG L.J. 203, 212 (1995). We are aware of no decision to expedite review of a life-saving drug receiving the same laudatory treatment.
(83.) They are also going to take the agency to court, of course. E.g., Am. Lung Ass'n v. EPA, 134 F.3d 388 (D.C. Cir. 1998); Am. Lung Ass'n v. EPA, Nos. 04-1275, 07-1021, 04-1378, 05-1409, 2008 WL 3198886 (D.C. Cir. July 29, 2008) (unpublished order); Am. Lung Ass'n v. EPA, Nos. 96-1251, 96-1252 96-1255, 1996 WL 734086 (D.C. Cir. Nov. 15, 1996) (unpublished opinion); Am. Lung Ass'n v. EPA, No. 93-1305, 1994 WL 329342 (D.C. Cir. June 27, 1994) (unpublished disposition).
(84.) For example, the Sierra Club has 1.3 million members, the Natural Resources Defense Council bas 1.2 million members, and the Defenders of Wildlife has 576,000 members. Defenders of Wildlife, About Us, http://www.defenders.org/about_us (last visited Nov. 15, 2009); Natural Resources Defense Council, About NRDC: Who We Are, http://nrdc.org/about/who_we_are.asp (last visited Nov. 15, 2009): Sierra Club Homepage, http://www.sierraclub.org/welcome (last visited Nov. 15, 2009).
(85.) Moreover, business firms may lobby for regulations in order to gain competitive advantages. See supra note 22.
(86.) An excellent, nuanced analysis of administrative politics is the "Political Oversight" chapter of MATTHEW D. ADLER & ERIC A. POSNER, NEW FOUNDATIONS OF COST-BENEFIT ANALYSIS 101-23 (2006).
(87.) According to then-professor Elena Kagan, who had served in a senior policy post in the Clinton White House:
In light of this criticism [of the rulemaking interventions of Presidents Reagan and Bush], observers might have predicted that when a Democratic President assumed office in 1993, a radical curtailment of presidential supervision of administrative action would follow. Instead, the very opposite occurred.... President Clinton treated the sphere of regulation as his own ... in a way no other modern President had done.
Elena Kagan, Presidential Administration, 114 HARV. L. REV. 2245, 2281 (2001).
(88.) DeMuth & Ginsburg, supra note 2, at 1088.
(89.) Nomination of Cass R. Sunstein to be Administrator of the Office of Information and Regulatory Affairs of the Office of Management and Budget: Hearing of the S. Homeland Security and Governmental Affairs Comm., 111th Cong., at *6 (2009) (Statement of Cass R. Sunstein) (LEXIS, Federal News Service File).
Christopher C. DeMuth, D.C. Searle Senior Fellow, American Enterprise Institute for Public Policy Research.
Douglas H. Ginsburg, Judge, United States Court of Appeals for the District of Columbia Circuit.
|Printer friendly Cite/link Email Feedback|
|Author:||DeMuth, Christopher C.; Ginsburg, Douglas H.|
|Publication:||Michigan Law Review|
|Article Type:||Book review|
|Date:||Apr 1, 2010|
|Previous Article:||How Judges Think.|
|Next Article:||Nudge: Improving Decisions About Health, Wealth, and Happiness.|