Is the Clean Air Act unconstitutional?

• Author: Sunstein, Cass R.
• Position: Statistical Data Included

The American people are suffering from what can be called "a syndrome of paranoia and neglect" about potential dangers to their health, safety, and the environment. This leads to a paradox that is becoming increasingly recognized. Large amount's of resources are devoted to slight or speculative dangers while substantial and well-documented dangers remain unaddressed. --John Graham(1) The last third of the twentieth century is not too late a time for turning the rule of law and the non-delegation doctrine into effective and useful legal tools for minimizing injustice from improper discretionary power. --Kenneth Culp Davis(2) I. INTRODUCTION: ENVIRONMENTAL POLICY AND ADMINISTRATIVE LAW

This Article deals with two linked questions. The first involves the future of the Clean Air Act.(3) The particular concern is how the Environmental Protection Agency ("EPA") might be encouraged, with help from reviewing courts, to issue better ambient air quality standards, and in the process to shift from some of the anachronisms of 1970s environmentalism to a more fruitful approach to environmental protection. The second question involves the role of the nondelegation doctrine in American public law, a doctrine that shows unmistakable signs of revival. I will suggest that improved performance by EPA and agencies in general, operating in tandem with a new approach to judicial review, would make it altogether unnecessary to revive the nondelegation doctrine. Indeed, the nondelegation doctrine would emerge as a crude and unhelpful response to existing problems in modern regulation, even a form of judicial usurpation. Far better responses are available.

1. Basic Proposals and First Principles

   In issuing and revising a national ambient air quality regulation under the Clean Air Act, EPA should provide a detailed "benefits analysis." To this end, it should undertake two tasks. First, it should specify the range of benefits that it believes are likely to result from the regulation, along with a specification of the range of benefits that it believes would result from at least two reasonable alternative approaches, one stricter and one more lenient. In the process EPA should identify the residual risk left under the competing regulatory regimes; it should also acknowledge scientific uncertainty, to the extent that uncertainty exists and requires guesswork. This proposal is an effort to strengthen the role of sound science in environmental protection.

   Second, EPA should explain why it believes that the chosen rule is preferable to the less and more stringent alternatives -- why the set of benefits to be received from the selected rule justify that rule, whereas the set of benefits to be received from the less and more stringent rules do not. In the process it should explain why the residual risk left by the selected rule is acceptable, while the residual risk left by the less stringent rule is not. This proposal is an effort to strengthen the role of democratic forces in environmental protection.(4)

   If necessary, reviewing courts should require EPA to perform these tasks. Taken together, the two proposals should increase the level of consistency across regulations, reduce the power of well-organized private groups, and diminish the risks associated with both insufficient and excessive environmental regulation. If EPA has undertaken the two tasks, and carried them out in a reasonable way, judicial review is at an end; courts should uphold EPA's decision.

   Ideas of this kind have potentially broad implications, extending well beyond the Clean Air Act and even EPA, to the work of the Occupational Safety and Health Administration ("OSHA"), the Consumer Product Safety Commission, and the National Highway Traffic and Safety Administration as well. They would mark a key moment in the movement toward a system of environmental protection that is at once more democratic and better informed. At the same time, they would accelerate the continuing shift from 1970s environmentalism and indeed 1970s regulation in general, away from recognizing the existence of problems of safety and health and toward assessing their magnitude, in such a way as to reduce both regulatory paranoia and regulatory neglect, and to put a premium on the acquisition of information.(5)

   If agencies undertook tasks of this kind, there would be little point to the recent resurgence of interest in the nondelegation doctrine. The sensible impulses that underlie that interest -- impulses that involve accountability, deliberation, and sound policymaking -- can be handled through other means, a point that casts a more general light on the proper role of the nondelegation doctrine in American public law. I suggest that the proper role of that doctrine consists in statutory construction that imposes floors and ceilings on agency action, and in a set of "nondelegation canons" that prevent agencies from acting without clear congressional authorization. Taken together, judicial requirements of this kind would constitute a form of "democracy-promoting minimalism"(6) in the distinctive context of administrative law -- and a partial step toward a better and more sophisticated approach to environmental protection. These are the basic claims that I will attempt to defend in this Article.

2. The Clean Air Act

   The Clean Air Act may well be the most important of all environmental statutes. Its effects include a wide range of beneficial consequences for human health and well-being and high costs on the private sector. The Environmental Protection Agency estimates overall compliance costs at $0.5 trillion.(7) The Act's claim to success rests on enormous improvements in ambient air quality and corresponding health benefits. EPA estimates that the Act prevents at least 45,000 deaths annually and that it also prevents a minimum of 13,000 heart attacks and 7,000 annual strokes.(8) On a standard (though not undisputed) view, the benefits of the Act, ranging between $5.6 trillion and $49.4 trillion, far exceed its costs.(9)

   The Act has nonetheless been subject to telling criticism. The foundation of clear air regulation consists of EPA's issuance of nationally uniform ambient air quality standards;(10) but in light of the extraordinary diversity of the fifty states, it is not clear that the idea of national standards can be rationally defended.(11) In addition, both lower courts and EPA seem to think that the standard-setting process does not and cannot involve consideration of costs.(12) But does it make sense, or is it even feasible, to say that national standards will be founded on an assessment of benefits alone, conducted in a cost vacuum?(13) If an improvement in ambient air quality would produce health benefits that are small but not trivial, isn't it clear that the improvement is justified if compliance costs are trivial, but not if the costs are extremely high? There is reason to think that at least in some cases, an understanding of costs has affected EPA's decision about appropriate standards -- but that the cost-benefit balancing has been left implicit and free from public scrutiny and review.(14)

   Perhaps the largest question involves the criteria by which EPA decides whether one or another level of regulation is (in the statutory phrase) "requisite to protect the public health."(15) For most pollutants, air quality at various levels is not either "safe" or "not safe"; there are diminishing degrees of risk associated with diminishing degrees of exposure. On what basis is a particular level of residual risk said to be the appropriate one? What judgments do, or should, enter into that conclusion? EPA has been criticized for sometimes suggesting, in an unhelpful and conclusory fashion, that it chooses the "safe" level, as if this were solely a technocratic judgment and as if "safety" were an on-off switch,(16) when its decision about permissible levels rests instead on a series of political, scientific, and economic judgments and compromises.

   There are two problems with this state of affairs. The first involves democratic deliberation.(17) If EPA does not give a clear sense of the range of adverse effects, and if it does not say why one set of such effects calls for regulation and another does not, the public and its representatives are not informed of the nature of the underlying questions, and they are unable to evaluate the choices actually made. Under EPA's articulated position, a purely technical issue (would a certain level be safe?) is sometimes substituted, at least publicly, for the real and more complicated ones (what level of safety is appropriate in light of all the relevant factors? why should one level of regulation be preferred to another?).

   The second problem involves sound regulatory policy. Any proposed national standard could be loosened or tightened, and the question is whether the agency has chosen the optimal, or at least a reasonable, regulatory "point." Without a clear and (to the extent possible) quantified presentation of the expected environmental benefits of the various alternatives,(18) there can be no assurance that the agency has chosen that point, rather than one that is too strict or too lenient.

3. A Remarkable Decision and a New Doctrine

   In its extraordinary decision in American Trucking Ass'ns v. EPA,(19) the United States Court of Appeals for the District of Columbia Circuit responded to this last concern in the strongest possible terms. It held that as interpreted by EPA, the key provisions of the Clean Air Act -- those that give EPA authority to issue national air quality standards -- represent an unconstitutional delegation of legislative power.(20) The decision announces the birth of a new nondelegation doctrine, one with potentially large implications for regulatory policy. Under the new doctrine, open-ended statutory terms will be invalidated unless agencies are able to specify the governing legal criteria -- to discipline their own authority...