Controversies Surrounding Regulatory Impact Analysis
The use of economic methods to evaluate the benefits and costs of new regulations in the areas of health and the environment has expanded dramatically over the past several decades in the United States and is now quite entrenched in the federal regulatory process. In 1981, one of the first actions of the Reagan administration was the issuance of an executive order (EO 12291) that required that “major” regulations—including those with an effect on the economy of $100 million or more—undergo a “regulatory impact analysis” (RIA). Thus, in most federal agencies, a major rule could not be proposed in the Federal Register until a cost–benefit analysis (CBA) had been prepared and submitted to the Office of Management and Budget (OMB) for review and approval. Executive orders by subsequent administrations, most notably President Clinton's EO 12866, put greater attention on ensuring an efficient OMB vetting process and a heightened focus on the nonquantitative consequences of major rules. Nevertheless, Clinton's executive order left in place the key components of regulatory benefit and cost estimation and OMB review. This order continues to govern regulatory review today.1
Most observers of the regulatory process expect that the outcome of the 2008 presidential election will not affect this general structure. Indeed, some time ago, law professor Cass Sunstein heralded the arrival of the “cost-benefit state” (Sunstein 2002). (Sunstein's views have become all the more important with his nomination to be head of the Office of Regulatory Affairs in the Obama administration.) More recently, New York University School of Law Dean Richard Revesz and attorney Michael Livermore declared that “[c]ost–benefit analysis is here to stay.”2
At the same time, a growing chorus of scholars and activists have decried what they consider to be an excessive focus on economic analysis and economic efficiency in federal rulemaking. They have argued, for example, that the cost–benefit approach inappropriately values impacts on “priceless” species, habitats, and other important, difficult-to-quantify resources; that the discounting of future regulatory consequences, including human mortality, treats lives unequally and trivializes the future; and that gains and losses to the rich should not be treated the same as those to the poor (Ackerman and Heinzerling 2004). In 2002, leading advocates of alternative approaches to regulatory assessment launched the Center for Progressive Reform, dedicated to the support of regulatory action to protect health, safety, and the environment while “rejecting the conservative view that government's only function is to increase the economic efficiency of private markets.”3
This report proceeds on two premises. The first is that the opponents of more stringent environmental regulation are not going to get everything they want, nor should they. Clearly, environmental regulation is here to stay, and the American electorate will not tolerate a return to the bad old days when, for example, polluters were free to dump whatever noxious substances they wanted into the air or water.
The second premise is that few want to see a world where every potential environmental risk, no matter how small or fanciful, leads to a new and potentially onerous restriction on product use or manufacture or on corporate or individual behavior.
In sum, both advocates and skeptics of more stringent regulation have been guilty of overreaching in the past, and both sides have paid a price for it. Of course, it is too simplistic to suggest that one could find a happy medium of environmental protection. What is possible from a legislative and regulatory perspective shifts constantly depending on precedent, environmental incidents, the state of technology, the philosophy of the party or individuals in power, and other factors.
The principal focus of this report is not on the lively, and ongoing, philosophical debate between proponents and opponents of the approach as to whether the analytical technique of cost–benefit analysis is necessary, rational, and environmentally protective—or unhelpful, indeterminate, and immoral. Rather, for the purposes of this volume, we embrace the pragmatic view of Sunstein and Revesz that CBA is here to stay.
This volume brings together, for the very first time, distinguished scholars with diverse views in an effort to improve the workings of the basic structure of regulatory impact analysis that we now have. We have asked proponents of CBA to approach fundamental features of current economic analysis with a fresh and skeptical eye. We have asked opponents of CBA, for present purposes at least, to set aside their general objections and to offer constructive possibilities for adjustments to the method. Thus, although a central premise of the volume is that some type of formal economic analysis will be used to support major federal regulations, the design of that analysis and its proper role in the regulatory process are very much at issue.
Rather than considering these matters in the abstract, the report considers the appropriate use of CBA by examining actual RIAS. Case studies of the RIAS for three U.S. Environmental Protection Agency (EPA) rules provide the fodder for the reforms we ultimately propose. The rules are the Clean Air Interstate Rule ),CAIR), the Clean Air Mercury Rule (CAMR), and the Cooling Water Intake Structure Rule (Phase II). The case studies help to clarify concrete differences between the sides in the cost–benefit debate and to suggest reforms to the current system for preparing and reviewing RIAS.
Overall, we seek to augment the often philosophical nature of the current debate with a quite pragmatic focus on actual regulatory analyses. We address a number of basic questions: Could particular changes to current practice improve the transparency of RIAS, enhance or modify their content, and increase the acceptability of the resulting studies? Even if the skeptics never fully embrace a cost–benefit framework, could it be made less objectionable?
This first chapter is designed to set the stage for the more detailed assessments that follow. We first present some essential background information on the role of CBA in the regulatory process and on the debates it has spawned. We briefly describe major contentions in this debate—not to answer them, but as part of our effort to improve the current process of regulatory analysis by understanding the perspectives of all sides. We then explain the process followed in engaging the multidisciplinary group of scholars involved in this study. Chapters 2–10 contain the detailed assessments of the three cases, including critiques of each RIA by proponents and opponents of CBA. The final chapter is an attempt by the editors to seek some common ground on the preferred means of conducting regulatory analysis, including recommendations for improving both the content of RIAS and the process by which they are developed and reviewed.
This report comes at a particularly timely moment, as President Obama, on January 30, 2009, issued a memorandum directed to the heads of executive departments and agencies, asking for their views on how to improve the process of regulatory review. The memorandum directs the head of OMB, “in consultation with representatives of regulatory agencies, to produce within 100 days a set of recommendations for a new Executive Order on regulatory review.” The memorandum invites particular attention to “the role of distributional considerations, fairness, and concern for the interests of future generations,” and to “the role of the behavioral sciences in formulating regulatory policy.”
How CBA Fits into the Regulatory Process
CBA can play several different roles in the regulatory process. First, one environmental law—the Safe Drinking Water Act Amendments of 1996—explicitly calls for formal CBA in deciding on the scope of regulation under the statute. Clearly, under this law, CBA conducted pursuant to the process of regulatory review aligns with the kind of analysis called for by the statute.
Second, quite a large number of health and environmental laws, such as the Toxic Substances Control Act and the Federal Insecticide, Fungicide, and Rodenticide Act, require agencies to conduct a generalized balancing of costs and benefits in coming to their decisions. Here, too, CBA arguably fits comfortably within the statutory framework.
Third, some statutes either dictate a precise regulatory result or forbid altogether the application of CBA in choosing a regulatory approach. In these cases, CBA is not in part of the statutory framework. All versions of the executive orders on regulatory review have provided that, where the law and CBA conflict, the law prevails. Thus, for example, where Congress dictates that a particular performance level must be achieved by regulated sources or rules out the use of CBA in regulatory decisionmaking, congressional directives—rather than the dictates of CB—prevail.
Even in the latter case, proponents of CBA argue that this approach can play a useful role in informing political leaders and the public about the consequences—good and bad—of regulatory decisions. This discussion demonstrates, however, that from the very outset, there may be quite different expectations regarding the analysis conducted pursuant to statutory directives and that undertaken pursuant to the executive orders on CBA. That mismatch can sometimes create legal conflict.
CBA: The Pros
To its proponents, the paramount advantages of a well-done CBA are twofold. First, it forces regulatory designers to think about quantification (that is, the physical effects of regulations they propose) on public health, environmental quality, ecosystem health, and a host of other potentially relevant outcomes. Second, it forces serious consideration of whether and how much those changes matter, and it does so in a particular way. CBA attempts to express the value of those physical changes using, as a metric, a monetary measure of the aggregate change in individual well-being resulting from a policy decision. Individual welfare is assumed to depend on the satisfaction of individual preferences, and monetary measures of welfare change are derived by observing how much individuals are willing to pay or give up in terms of other consumption opportunities. This approach can be applied to nonmarket “public goods” such as environmental quality or environmental risk reduction as well as to market goods and services, although the measurement of nonmarket values is more challenging. When measurement of such nonmarket values is impossible or in some way unacceptable, analysts may resort to cost-effectiveness analysis (CEA), a less ambitious approach in which a policy outcome (for example, a specified reduction in ambient pollution concentration) is taken as a given, and the analysis seeks to identify the least-cost means for achieving the goal, taking into account any ancillary benefits of alternative actions. Every CBA has at least one CEA buried inside.
In addition, the advantages of CBA (and CEA) include the following:
transparency and the resulting potential for engendering accountability;
the provision of a framework for consistent data collection and the identification of gaps and uncertainty in knowledge;
the development of metrics for both the beneficial and adverse consequences of alternative regulatory approaches, allowing those alternatives to be compared to one another; and
with the use of a monetary metric, the ability to aggregate dissimilar effects (such as those on health, visibility, and crops) into one measure of net benefits.
Most economists would acknowledge that CBA does not incorporate all factors that can and should influence judgments on the social worth of a policy and that individual preference satisfaction is not the only criterion. Nevertheless, most would also argue that rigorous CBA can elucidate for a broader audience how various regulatory choices are supposed to work and who is likely to be affected. At a minimum, then, it can play a useful informational role in the decisionmaking process. CBA also makes a moral argument that private economic activity, as well as regulation, can generate value, and hence that good public policymaking is a balancing process.
From an economist's perspective, the usefulness of CBA is primarily limited by the ability to quantify the effects of regulations and to measure people's willingness to pay for those different health and environmental outcomes. Fortunately, the state of the science of measuring such economic values is quite active. Estimates of the willingness to pay for reductions in mortality and morbidity risks, for avoiding environmental damage to recreation opportunities, and for avoiding visibility degradation are the subjects of much ongoing research. Issues of a higher order stalk the estimation of nonuse values, and a variety of mostly empirical concerns have left material damage poorly understood. Often, estimation of the costs of reducing environmental effects, generally thought to be relatively straightforward, can be as challenging as estimation of the benefits.
The RIAS in which the CBAS are embedded have also undergone considerable changes because so many have been subjected to critical scrutiny, including internal agency and OMB reviews prior to publication of regulatory proposals, commentary from stakeholders, and, in many cases, review by the courts. Nearly 30 years of experience has led to an informal list of best practices—characteristics and components that, according to various commentators, belong in most if not all RIAS. These include the following:
the use of clear and consistent baseline assumptions;
the evaluation of an appropriately broad range of policy options, including alternatives to new regulation;
transparency in the use of assumptions, data and models, the comparison of alternatives, and the reporting of results;
appropriate treatment of discounting future benefits and costs and accounting for the cost of risk-bearing;
the use of probabilistic analyses and other methods to explore the robustness of conclusions;
the identification of nonmonetizable or nonquantifiable aspects of a policy and the potential incidence of all effects; and
the use of benefit and cost measures that are grounded in economic theory (measures of willingness to pay and opportunity cost).
Not surprisingly, in practice many RIAS do not contain all these elements. The reasons for the omissions vary: sometimes general resource limitations may be to blame, in other cases the omissions may be more strategic. Overall, the sophistication of RIAS appears to have increased over time. In part, this tracks with the growth in the field of environmental economics. Yet one unfortunate result of this growing sophistication is that RIAS have become much longer and more technically oriented documents, leading perhaps to a certain sacrifice in transparency.
CBA: The Cons
To its critics, CBA is a flawed technique that, among other things, excessively emphasizes the quantification and monetization of risks, trivializes the future through discounting, fails to meaningfully assess the value of avoiding nonmarginal consequences (including environmental catastrophe), and ignores distributional concerns.
In many cases, it is currently impossible to quantify all of the important benefits of an environmental regulation. Indeed, it is often impossible to quantify even a substantial portion of them. When the quantified benefits of a rule include only cancer cases averted, yet the rule will also prevent many other illnesses as well as adverse effects on ecosystems, a CBA of that rule will be woefully incomplete (see Heinzerling 1998).
Even cost–benefit proponents concede that there are no good estimates of the monetary value of many of the benefits of environmental regulation, including the avoidance of many kinds of illnesses and other adverse health conditions and the prevention of harms to species and ecosystems. Where no good estimates of value exist, the benefits will not count for very much in CBA. Even where cost–benefit analysts believe that the estimates are pretty good, cost–benefit critics often disagree.
The value of preventing death is a prime example. It has become standard to measure the health risk reductions associated with proposed regulations as statistical lives. If the risk of dying from a particular cause is reduced by one in one million for one million people, it is said that the regulation would save one statistical life. To measure the value of this statistical life, cost–benefit analysts ask how much individuals are willing to pay to avoid—or how much they are willing to accept to take on—the extra risk of one in one million. If, say, everyone is willing to pay $1 to avoid this risk, then the analysts would say that the value of the statistical life in this case is $1,000,000.
This analysis has many problems, according to cost–benefit critics. Here we cite three. First, in inventing the statistical life, cost–benefit analysts have not escaped the fundamental moral conundrum of valuing life itself; they have merely glossed over it. When a person dies as a result of environmental contamination, she really dies: a real life, not a statistical one, is lost. CBA ignores this fundamental fact.
Second, in assuming that willingness to pay is the measure of value, CBA takes as a given that decisions made in private economic markets will be the same as decisions made by individuals acting as public citizens. But, for many reasons, these settings might produce different decisions, not least among them the fact that in environmental matters, the problem of public goods will press individuals acting alone to devote few or no resources to cleaning up a problem.
Third, valuing life—or health—according to how much people are willing to pay for it invites inequality. Some cost–benefit proponents have advocated, for example, that the rich should be valued more highly than the poor—a position at which many among us would flinch but which is perfectly consistent with the underlying theory of willingness to pay. And, indeed, we see glimmers of this approach in recent EPA analyses. In a preliminary assessment of the social costs of carbon, EPA relied on an estimate that embedded wildly differential values for the lives of people in rich and poor countries. Equally troubling are EPA’s fitful efforts to reduce the value of the elderly compared to that of younger people.
Much of environmental law aims to protect the future—to protect people living now from illnesses that might befall them in the future, to protect future people from such events, and to preserve ecosystems so that future generations might use and enjoy them much as we do now. Although discounting does account for the costs to present generations of providing these protections, opponents of CBA believe that discounting is not consistent with environmental law's forward-looking premise because the standard technique of constant exponential discounting can have a potentially large adverse effect on the perceived benefits of policies—such as policies to address climate change or policies to protect against long-latency diseases like cancer—that aim to prevent future harms.
Although the studies increasingly emphasize the incorporation of uncertainty, CBA often assumes stable problems with stable solutions. It works at the margins, but not when the margin is a cliff's edge. Many environmental problems—perhaps the most important example is climate change— involve great uncertainties and potential irreversibility, features ill-suited to the cost–benefit framework.
As implied in the above discussion of the value of life, opponents of CBA believe that the willingness to pay criterion has an inherent inequity. There is no corrective to this problem in CBA as it is currently conducted, and indeed CBA is largely blind to the distributional implications of environmental degradation. The kinds of concerns stated above have led numerous cost–benefit critics to call for abandonment of the method—ending it rather than mending it. However, this report assumes that CBA is here to stay. Thus, we focus on the reform rather than the rejection of the approach. Specifically, we aim to improve the cost–benefit method, as currently applied by EPA, without fully engaging the larger debate over whether it is fundamentally flawed. Yet, as we shall see, some of the more fundamental criticisms of the method turn out to be helpful in recommending ways in which current practices might be improved.
Analysis versus Decisionmaking
Throughout this volume, it is useful to distinguish between the analytical and decisionmaking components of rulemaking. Although the two are closely related, they are not one and the same. In fact, some of the differences between the two components were clarified in EO 12866. Specifically, EO 12866 replaced the stipulation contained in EO 12291 that benefits “outweigh” costs with a requirement for “a reasoned determination that the benefits...justify the costs.” Further, agencies were mandated to “include both quantifiable measures...and qualitative measures of costs and benefits” and “to select those approaches that maximize net benefits (including potential economic, environmental, public health and safety, and other advantages; distributive impacts; and equity) unless a statute requires another regulatory approach.”4 In effect, EO 12866 embraces social welfare considerations that may not be easily quantified, such as public health and distributional impacts, and rejects the idea that quantified CBA provides a rigid rule for decisionmaking.
Thus, EO 12866 is consistent with the views of most economists, who see CBA as a tool rather than a strict rule for decisionmaking. As Nobel Laureate Kenneth Arrow and others have written:
...[In] many cases, benefit–cost analysis cannot be used to prove that the economic benefits of a decision will exceed or fall short of the costs....[But it] can provide illuminating evidence for a decision, even if precision cannot be achieved because of limitations on time, resources, or the availability of information. (Arrow et al. 1996, 5)
Arrow et al. (1996) also note that agencies may want to consider other factors in their decisions, such as equity within and across generations, or they may want to place greater weight on particular characteristics of a decision, such as irreversible consequences. They recommend that when the expected costs of regulations far exceed expected benefits, agency heads should be required to present a clear explanation justifying the reasons for their decisions.
Critics of CBA are concerned that even this attenuated process places too much emphasis on CBA and, more generally, on economic efficiency in the decisionmaking process. Despite the language of EO 12866, they also fear that, in its review of new rules, OMB continues to apply the more rigid criteria of EO 12291, namely that benefits outweigh costs.
The true influence of RIAS on regulatory outcomes is not well understood. Indeed, it may be that some of the regulatory “successes” of CBA would have reached a similar outcome even if no CBA had ever been prepared.
Interestingly, the direct effect of a CBA on the regulatory outcome is not the only, and may not be its most important, influence on the regulatory process. Twenty years ago, an EPA report (EPA 1987) listed four specific areas—besides supporting regulatory decisions—where the RIA influenced the development of regulations:
guiding the development of the regulation;
adding new alternatives;
eliminating noncost-effective alternatives; and
adjusting alternatives to account for differences among industries or industry segments.
The RIA requirement also has been credited with making upper management at EPA and other regulatory agencies more aware of the implications of their decisions.
For each of the three RIAS examined in this volume, the first task was to produce a coherent and readable description of the analysis that was conducted by EPA. We assigned this task to two of the volume editors (Harrington and Morgenstern), and to a former Resources for the Future (RFF) colleague who has joined the staff of EPA’s Policy Office (David Evans). The assignment for these three authors was to faithfully report the content of the RIA, including the stated justification for the rule; to digest and summarize its technical complexities; and, where appropriate, to highlight potentially controversial issues. However, the authors of these chapters were asked not to take a stand on any of the controversies. Rather, their role is largely reportorial in nature.
The second task, for three scholars on each side of the cost–benefit debate, was to review the RIAS, critique them, and explore what complementary or substitute analyses could have been included. The three authors skeptical of CBA (all law professors) are: Douglas Kysar, Yale Law School; Catherine O'Neill, Seattle University Law School; and Wendy Wagner, University of Texas Law School. The three authors favoring CBA (all economists) are: Nathaniel Keohane, originally Yale University, now Environmental Defense Fund; Alan Krupnick, RFF; and Scott Farrow, University of Maryland. Purposely, we did not specify a precise format for the critiques. The resulting papers cover a wide-ranging set of issues, with the economists generally focusing more on the techniques used in the individual RIAS, and the lawyers often including broader philosophical critiques of CBA.
A small authors workshop was convened in June 2008 at RFF in Washington, DC, to review the papers and consider possible reforms of both the analytical facets of CBA and the process for doing and using this type of analysis. Workshop participants included the paper authors, designated peer reviewers (Frank Ackerman, Tufts University; James Hammitt, Harvard University; and William Pedersen, attorney at law), and selected additional experts from EPA’s air and policy offices and from OMB: Alexander Cristofaro, Arthur Fraas, Bryan Hubbell, Albert McGartland, and Sam Napolitano.
Following the workshop, the authors exchanged drafts and had the opportunity to revise their papers on the basis of workshop comments, with the overall goal of ensuring relatively parallel coverage of topics.
Case Study Selection
We make no claim that the RIAS studied in this volume are representative of the 100-plus such analyses prepared by EPA since 1981. However, the regulations selected for inclusion in this project involve issues that are clearly relevant to examining CBA in operation; for example, the valuation of statistical lives, the monetization of ecological and other types of difficult-to-measure benefits, the discounting of future consequences, and the distribution of costs and benefits. All three cases are based on relatively recent rules, thereby ensuring that they reflect current practices in RIA development at EPA. Two of the three rules involve air pollution and one addresses water bodies. The fact that all focus on the electric utility industry is somewhat accidental and certainly not part of any grand design. Arguably, however, an RIA focused on this industry, which has been subject to environmental controls since even before the days of national regulation, is probably among the most sophisticated, thus revealing some of the “best” practices for RIAS. The fact that all three rules have been invalidated by the courts was clearly not part of our design. A description of the three rules follows:
1. The CAIR, issued by EPA in March 2005, aimed at considerable reductions in emissions of sulfur dioxide (SO2) and nitrogen oxides (NOx) in the eastern United States. EPA expected that, by 2015, SO2 emissions would decline by 70 percent from 2003 levels, whereas NOx emissions would fall by 60 percent. Although the RIA demonstrated that the monetized benefits of the rule—largely in the form of reduced human mortality—exceeded the costs by a considerable margin, thus providing solid justification for the rule, critics contend that the agency's own analysis would have supported even larger emissions reductions with earlier compliance deadlines. Key issues involve both the interpretation and the use of the RIA. The U.S. Court of Appeals in Washington, DC, rejected the CAIR in its entirety, concluding, among other things, that the regional trading program created by the rule paid inadequate attention to the state-specific focus of the relevant statutory provisions.
2. The CAMR, also issued by EPA in March 2005, was designed to reduce mercury emissions from power plants. Although interim emissions reductions of 20 percent from 2003 levels were expected from the installation of controls required under the CAIR, the CAMR itself was expected to result in further reductions in mercury amounting to almost 70 percent reductions by the year 2018. Critics contend that the agency's own analysis supported earlier and deeper cuts of these toxic emissions. Further, they expressed concern that the provision for emissions trading could lead to unacceptably high exposure levels for individuals in certain areas. The U.S. Court of Appeals in Washington, DC, also invalidated this rule, reasoning that EPA had failed to follow proper procedures in taking mercury off the list of pollutants to be regulated under a provision of the Clean Air Act requiring strict technology-based controls for regulated sources.
3. The Cooling Water Intake Structures Rule (Phase II), issued by EPA in June 2006, was designed to minimize the harmful impacts on aquatic life caused by cooling water intake structures at existing power plants. The rule set performance standards (rather than technology requirements) for these plants and also allowed plants to avoid these requirements either through restoration measures or through a site-specific CBA indicating that the costs of meeting the standards were not worth the benefits at a specific plant. EPA’s stated reason for the content of its rule was that the CBA did not support a more stringent approach. Critics contend that the agency's RIA failed to adequately account for certain ecological damages and so understated the net benefits to be achieved from further controls. The U.S. Court of Appeals in New York invalidated the rule, finding that the relevant provision of the Clean Water Act did not permit CBA. The U.S. Supreme Court agreed to hear the case in its 2008–2009 term.
The case studies of these three rules form the backbone of our work, which is to recommend changes, both substantive and procedural, to improve the process of regulatory review.
1. In March 2007 President Bush issued EO 13422, which expanded OMB’s jurisdiction to include the review of guidance documents issued by federal agencies. It also required that regulatory agencies provide a written rationale for new regulations and estimates of aggregate annual costs and benefits of all regulatory activities in the agencies’ plans. President Obama rescinded EO 13422 on February 4, 2009.
2. Revesz and Livermore 2006, page 11.
3. Center for Progressive Reform (n.d.)
4. EO 12866 1(a), 3 CFR at 638-39 (1995).
Ackerman, F., and L. Heinzerling. 2004. Priceless: On Knowing the Price of Everything and the Value of Nothing. New York: The New Press.
Arrow, Kenneth, M. Cropper, G. Eads, R. Hahn, L. Lave, R. Noll, P. Portney, M. Russell, R. Schmalensee, K. Smith and R. Stavins. 1996. “Is There a Role for Benefit-Cost Analysis in Environmental, Health, and Safety Regulation?” Science, 272:5259 (April 12), pp 221–222.
Center for Progressive Reform (n.d.), www.progressiveregulation.org.
U.S. Environmental Protection Agency (EPA). 1987. EPA's Use of Benefit–Cost Analysis, 1981–1986. USEPA-230-05-87-028. Washington, DC: U.S. EPA.
Heinzerling, L. 1998. Regulatory Costs of Mythic Proportions. Yale Law Journal 107: 1981–2070.
Revesz, R.L., and M.A. Livermore. 2008. Retaking Rationality: How Cost–Benefit Analysis Can Better Protect the Environment and Our Health. New York: Oxford University Press.
Sunstein, C.R. 2002. The Cost–Benefit State: The Future of Regulatory Protection, The American Bar Association (Section of Administrative Law and Regulatory Practice), Chicago.