EPA's ANPRM on Increasing Consistency and Transparency in Benefit Cost Analysis

Among notable environmental policy news in the past few weeks, the Environmental Protection Agency (EPA) has been considering some even broader changes to how it does business that merit a closer look. Back in June, the agency requested comment on “Increasing Consistency and Transparency in Considering Costs and Benefits in the Rulemaking Process.” In short, EPA is considering changing how it does benefit-cost analysis that’s central to much of its regulatory action. As we detail in a recent comment to EPA, some of the agency’s suggestions are good ones. But others seem pointless, and a few are just bad ideas.

We should be clear at the outset that we agree with the overall view presented in the Advanced Notice of Proposed Rulemaking (ANPRM) that benefit-cost analysis is an important tool for EPA decisionmaking. The government-wide policy of every president since Ronald Reagan has required it (at least when consistent with statute). We further applaud efforts to improve the quality, consistency, and transparency of that analysis. However, some suggested policy changes in the ANPRM are, in our view, contrary to best economic practice, to existing law, or both, and therefore will not further the ANPRM’s stated objectives.

Currently, use of benefit-cost analysis at EPA is shaped by underlying statutes and by interagency guidance documents, such as executive orders and OMB Circular A-4. To the extent the ANPRM identifies areas where use of benefit-cost analysis can be improved at EPA, the same issues are likely present in other agencies. We see no reason why this ANPRM should apply only to EPA. These documents should continue to guide EPA’s cost-benefit analysis, and to the extent either need to be improved or updated, a formal, interagency review process should be used.

We also see no benefit to making any changes to cost-benefit analysis via the complex and lengthy rulemaking process. Less formal (though still public and peer-reviewed) guidance documents currently guide this and similar internal agency deliberations, and we see no reason for that to change. If the desire is to create a more durable set of guidelines capable of surviving changes in administrations (perhaps a laudable goal), even a rulemaking cannot achieve that goal since a future EPA could revise it. Only congressional action can create a durable change in policy or procedure.

EPA also raises questions about the proper scope of cost-benefit analysis, on both the benefits and costs side. We agree that greater transparency and consistency in this area would be a welcome development. But a core principle must be that cost-benefit analysis requires a full presentation of the benefits and costs for it to be useful. This means considering a broad range of alternatives, taking uncertainty seriously, and including effects that can’t be quantified.

EPA suggests in its proposal that the definition of benefits should be narrowed, and that of costs should be expanded. This would have predictable results on regulatory outcomes. In both cases, we think EPA has it wrong. On the benefits side, EPA questions whether ancillary benefits—environmental benefits other than those explicitly targeted by a regulation—should count in cost-benefit analysis. For us, the answer is easy: both common sense and best economic practice require them to be considered (where possible quantified) just like any other costs or benefits. On the other side of the equation, EPA suggests that “costs” are inadequately defined in current EPA guidance. While it is true that statutory commands to EPA may be complex, current EPA and OMB guidance is adequate here, and the agency has the ability to consider a wide range of possible regulatory costs.

Finally, EPA suggests the use of retrospective analysis for evidence-based insights into future rulemaking and the revision, modification, and even repeal of existing rules. We agree that current efforts to carry out retrospective review of regulations is sorely lacking. But, we also believe such review should be bidirectional, i.e., not biased toward less regulation. Identifying opportunities for achieving benefits at low cost should be as important as identifying wasteful or poorly designed regulations.

EPA’s early proposal is therefore a mixed bag. In broad strokes, we agree with its aims: more clarity and consistency in cost-benefit analysis, and meaningful retrospective review. But we see serious flaws in the details, both procedurally and substantively. A single-agency rulemaking is the wrong regulatory vehicle for the job. And biasing cost-benefit analysis by narrowing the scope of benefits and broadening the scope of costs is transparently unwise.