I thank Jonathan H. Adler, Patrick Allitt, and William Dennis for their thoughtful and informative commentaries on my Liberty Forum essay on reforming U.S. environmental policy, in particular with respect to the National Environmental Policy Act (NEPA) and the Chevron doctrine of judicial deference to executive agency statutory interpretations in the face of legal ambiguities. I learned much from their observations, to each of which I offer brief responses here.

At the risk of some oversimplification, Adler’s argument can be summarized as follows:

• NEPA, “a quite limited statute,” is less important than I argued in my essay, as it requires only that federal agencies prepare an environmental impact statements (EIS) before approving major infrastructure projects and the like. Indeed, “there is nothing in NEPA that requires agencies to prioritize . . . policies or actions that minimize environmental impacts.” Agencies are required only to “consider such impacts.”
• There “is little reason to believe that getting rid of Chevron will do much to improve environmental policy, and many reasons to suspect that it could actually do more harm than good.”
• A better reform agenda would concentrate on enhancing agency attention to the costs of their actions and reform of their policies’ perverse incentives for private actors; a reduction of uncompensated regulatory takings; an increase in environmental federalism so as to increase experimentation and the degree to which policies can reflect regional differences in preferences with respect to the benefits and costs of environmental protection; an increase in the use of property rights as a tool for environmental stewardship; and a replacement of the regulatory approach with a system of taxes on private actions that inflict environmental harm.

Adler does agree that NEPA “can impose a significant procedural burden on federal agencies undertaking major projects,” and that EIS preparation “may take years to complete” with lawsuits by “environmentalist and NIMBY groups . . . challenging the adequacy of environmental assessments and impact statements.” But I think that he understates the underlying problem. Federal agencies must complete something on the order of 250 major EIS’s each year—each one takes about five years to complete—with many thousands of smaller assessments annually.

The budgetary cost of all this is not the central problem. Instead, the “completeness” requirement for the EIS is a source of massive litigation, with something on the order of 100 NEPA lawsuits filed each year, and it must be the case that such delays have resulted in the cancellation of hundreds or thousands of projects that otherwise would be justified in a benefit/cost context. (One summary estimate can be found here.)

Accordingly, Adler’s observation that “the federal government ultimately prevails in most NEPA suits” is rather beside the point. No EIS can consider every possible environmental impact of a proposed project, and efforts by courts to apply some sort of nebulous “common sense” criterion to the claims advanced in litigation would run counter to the diminished judicial role that Adler seems to favor in the context of environmental policy. Moreover, he does not address a central adverse impact of this litigation incentive: the bias in favor of the status quo (discussed in my essay), resulting in delays or cancellation of projects that would yield environmental improvement relative to the effects of the existing capital stock.

With respect to Adler’s argument that the Chevron doctrine is likely to yield beneficial net environmental impacts by limiting the interference of judges, in favor of executive branch actions undertaken by agencies controlled by Presidents constrained by public accountability (or political competition): Again, I believe that he is not addressing the central problem. Adler is a distinguished legal scholar—I am but a lowly economist—yet it seems to me that the constitutional issue is encompassed by his apparent approval of the “explicit premise of Chevron,” to wit, “that Congress has delegated to the implementing agency the authority to resolve ambiguities and fill statutory gaps . . . and that such power necessarily involves the policy judgments . . . better made by politically accountable agencies than unaccountable courts.”

Precisely where in the Constitution does Congress have the power to delegate such authority to the President and his executive branch agencies? If there is an ambiguity, why can Congress not fix it? If the congressional bargaining process cannot yield agreement on a clarification, then perhaps the law is unconstitutionally vague. True, not all conditions and questions can be discerned in advance, but that decidedly is not a convincing rationale for eroding the separation of powers. It would seem that the industrial “bubble” issue underlying the Chevron decision was something that Congress should have addressed.

With respect to Adler’s argument that Chevron “has been indispensable for those administrations seeking to defend deregulatory measures in federal court”: He notes, correctly, that “pro-regulatory administrations” (read: the Obama Environmental Protection Agency) “have utilized Chevron to justify regulatory expansions.” Put separation-of-powers questions aside; it is very, very far from obvious that this sort of dynamic tug of war between regulators and deregulators under Chevron, across administrations that sometimes have very different preferences, will yield better environmental policy results in a benefit/cost context. And let us not forget the uncertainty effects inflicted upon the private sector as a result of such political seesawing when one administration replaces another.

I heartily endorse all of the alternative policy reforms that Adler discusses, with the exception of the taxation alternative to regulation. (I note for the record that I did discuss the cost-shifting problem in the context of cost considerations and private incentives.) Whether those reforms would yield results more salutary than the NEPA and Chevron reforms that I discussed is an interesting question that, I believe, cannot be answered on the basis of first principles.

Adler’s very brief discussion of emissions taxes reflects a conventional wisdom that is deeply problematic. The revenues to be derived from emissions taxes would engender political incentives for Congress and for various executive branch agencies that have been curiously under-examined in the literature, but that would yield outcomes very far from obviously superior to those flowing from the regulatory approach. (See some of my discussions of carbon taxes here, here, here, and here.)

He supports a “revenue-neutral carbon tax,” a policy preference that I believe is virtually certain to be defeated as a matter of equilibrium outcomes of interest-group political competition. Notice that a 2016 ballot initiative in Washington state, Initiative 732—a revenue-neutral carbon tax—was defeated by 59 to 41 percent; and Initiative 1631, a carbon tax explicitly not revenue-neutral, with payoffs to innumerable interest groups, lost on Tuesday, 56 to 44 percent. That carbon taxes cannot come close to winning popular approval in a blue state suggests that such proposed policies represent a political dead end. For a discussion, see this.

On to Patrick Allitt, who seems to agree with much of my essay but makes several additional arguments:

• Polluting industries could have policed themselves, thus avoiding the economic and political downsides of government regulation. At the same time, those industries have come to embrace environmental regulation, one reason for which is that doing so “reassured the public.”
• Two other reasons for industrial endorsement of environmental regulation are the opportunities it creates for rent-seeking (the use of regulation to hinder competition) and regulatory capture (long-term employment in the regulated industries for experts in the regulatory agencies).
• “Cap-and-trade” reform of environmental regulation is a clear improvement over command-and-control regulation.
• Some degree of Chevron-type judicial deference is inevitable given the complexities of environmental policy.
• Environmental policy since 1970 largely has been a success story in terms of actual improvement in U.S. environmental conditions.

I am dubious about the then-prospective benefits of self-regulation. Agreement among the major players would have been very difficult to achieve, and the antitrust implications would have been likely to sink the whole enterprise, as self-regulation would have had to deal with costs, input combinations, and product prices. Moreover, the negotiating parties would have had powerful incentives to use such self-regulation to make entry by new competitors more difficult; were that not the case, new entrants would have undercut the costs and prices characterizing operations by the negotiating parties, thus making ineffective the self-regulation itself.

Allitt surely is correct that environmental regulation creates large opportunities for rent-seeking and regulatory capture; but it is hard to see how those problems can be avoided under any system of dealing with environmental problems, as rent-seeking and regulatory capture are inherent in the very nature of government. An approach emphasizing greater federalism—more authority for state and local policymakers—would help to reduce those effects, but we should recognize that the federalization of environmental policy must create wealth transfers among regions and economic sectors. An example: The Obama climate-action plan would have transferred wealth from red states to blue ones, unambiguously.

It is far from clear that such market-simulating approaches as cap-and-trade represent efficiency improvements. Just as regulators must determine the ambient pollution levels that will satisfy the relevant statutes, someone must determine the allowable number of emission permits, and there is little reason to predict that the latter systematically will be more efficient than the former. If permits simply are given to the affected industries, the initial allocation must be determined, with all of the rent-seeking and other effects that are certain to ensue. If they are sold, with the government keeping the revenues, there is likely to be an incentive to drive the number of permits to inefficiently low levels. If the revenues are to be distributed, the identity of the recipients has to be determined, and the choices made will affect political incentives on the number of permits. Et cetera.

Perhaps some level of judicial deference to the agencies is, as Allitt says, inevitable; but that does not solve the separation-of-powers problem noted above, and such questions as the “bubble” issue are sufficiently central that congressional abdication cannot be acceptable. I agree with Allitt that environmental conditions in the United States have improved massively since the 1970s, but that obvious reality does not reduce the need for reform.

William Dennis also seems to agree with much of my essay, making the following main arguments:

• The traditional common law is well-suited to deal with the problem of environmental degradation by private parties, and the substitution of government regulation in place of that common law creates a drift “away from the constitutional order of a free society.”
• Because of “NEPA in conjunction with other acts” and policies, “American public lands [use] is probably ‘planned’ more than any other natural resource in the world.”
• The “caveman environment” parable (in my essay, the tradeoffs confronting a homo sapiens baby born in a cave some tens of thousands of years ago) is useful in terms of illustrating the complexity of the “environment” concept and the tradeoffs among them.

I am dubious as well about the usefulness of the common law to address environmental problems created by multiple-party befouling of such commons as air and water. Whom to sue in the absence of well-defined property rights? Who is responsible for what? If those suffering the adverse effects of environmental degradation can avoid them at little cost, who should be using the common law to effect a particular outcome? (Ronald Coase had much to say about that question.) And the “constitutional order of a free society” is less clear than Dennis seems to assume: Suppose at a constitutional level that free citizens, making decisions from behind a Rawlsian veil of ignorance, agree jointly to subject themselves to environmental regulation implemented outside the contours of the common law?

Dennis certainly is correct that the use of American lands owned as a legal matter by government is driven by a form of central planning; but the assertion that such use “is probably ‘planned’ more than any other natural resource in the world” is a bit of hyperbole. More than, say, agricultural land in Cuba or the Democratic People’s Republic of Korea? I appreciate Dennis’ endorsement of my “caveman environment” parable, as I believe it to be a useful way to think about the costs of environmental absolutism.

Again let me thank Adler, Allitt, and Dennis for their thoughtful and informative comments.