*This post is part of a symposium on Modernizing Regulatory Review. For other posts in the series, click here.
This essay was originally published in Administrative & Regulatory Law News, the quarterly magazine of the American Bar Association’s Administrative Law and Regulatory Practice Section. Visit here to become a Section member.
The recently proposed revisions to Circular A-4—a guidance document issued by the Office of Management and Budget (OMB) to help federal agencies conduct regulatory impact analysis—are comprehensive and powerful. OMB, Proposed Circular A-4, Draft for Public Review (April 6, 2023). The revisions promote transparency and clear presentation of results within analyses, provide detailed guidance on longstanding but inconsistently implemented directives, highlight new considerations for measuring benefits and costs, and update estimation methods and parameters based on advances from the last twenty years. This essay will focus on just one aspect of the revisions: the expanded discussion of distributional analysis.
Pursuant to President Clinton’s Executive Order 12,866, federal agencies prepare an analysis that must “assess both the costs and the benefits of [an] intended regulation,” an analysis that is often referred to as a benefit-cost analysis. Exec. Order No. 12,866, 58 Fed. Reg. 51735, 51736 (Sept. 30, 1993). This analysis is important because, to the extent permitted by law, the order requires an agency to adopt a regulation “only upon a reasoned determination that the benefits of the intended regulation justify its costs.” And making sure agencies issue the right regulations is vital because regulations, whether “good” or “bad,” fundamentally affect the wellbeing of people.
Conventional benefit-cost analysis focuses on the aggregate effects of a regulation. The idea is that if the benefits (which accrue to some people) are greater than the costs (which accrue to some people), then the regulation roughly does more good than harm to society as a whole. But a focus on aggregate effects can obscure differences in a regulation’s effects on various subgroups within the population. For example, a proposed regulation might seek to address an environmental externality associated with the production of some good. Imagine that pollution causes groups living nearby to suffer some harm, such as increased mortality or lower productivity. A benefit-cost analysis might show that restricting the pollution would benefit the nearby population. But, because reducing the pollution has a cost associated with it, the price of the good might increase, resulting in changes in who consumes the good. In addition, some of those who are employed in the industry might receive lower wages or might lose their jobs entirely. These groups are the ones burdened by the regulation.
A conventional benefit-cost analysis would try to list, quantify, and then monetize the different kinds of effects that would fall on people (or use proxies for these effects, such as compliance costs) to help policymakers understand whether the regulation’s overall benefits are likely to be greater than its overall costs. But sometimes policymakers might want to know the distributional effects of a regulation. For example, in the above example, they might want to know whether the local population that would benefit from the regulation is wealthy or poor; whether those who would lose their jobs are older low-income workers with specialized skill sets; whether those who would lose their jobs live near the facility, too; whether those most affected by the good’s price increase are low-income individuals; and whether, across many regulations, the groups that benefit, on net, from regulations are largely the same and the groups that bear costs, on net, are largely the same.
Distributional effects are important for at least three reasons. One reason is that an accurate assessment of benefits and costs even from an aggregate welfare perspective might depend on who benefits and who is burdened by a regulation. Scholars have argued that there is diminishing marginal utility to additional income when your income is above a certain level. If so, this would mean that an additional dollar would have a greater effect on the welfare of a low-income individual than a high-income individual. For this reason, the HM Treasury’s Green Book, which performs a similar guidance role as Circular A-4 in the United Kingdom, encourages agencies to present both a conventional benefit-cost analysis and an analysis that weighs net benefits that go to low-income individuals more. The proposed revisions to Circular A-4 now do the same. OMB, Proposed Circular A-4, at 65-66.
The second reason is that the public might want to know the distributional consequences of agency regulations. Because all regulations have distributional consequences whether agencies analyze them or not, it is important from a transparency and accountability perspective to understand who benefits from and is burdened by the regulations that agencies issue. Imagine that, over the course of an administration, most regulations promulgated by a particular agency benefitted only one segment of society or burdened only one segment of society. Imagine further that the regulations typically benefited politically advantaged groups or typically burdened disadvantaged groups. Such concerns are supported by the public choice literature, theoretically and, when studied, empirically. For example, consider the Superfund Program, where the Environmental Protection Agency (EPA) uses federal funds to clean up contaminated sites. W. Kip Viscusi and James Hamilton found that, among sites with the lowest calculated actual risk levels, the EPA cleaned up contaminated sites located in more politically active areas. See W. Kip Viscusi & James T. Hamilton, Are Risk Regulators Rational? Evidence from Hazardous Waste Cleanup Decisions, 89 Am. Econ. Rev. 1010 (1999). The environmental justice literature also highlights examples of such concerns. Transparency about these impacts can be a good natural check on this behavior. It could also be a dimension on which presidential administrations could be held accountable by the public. Along these lines, the guidance within the proposed Circular A-4 focuses on ensuring that the analysis is presented in a way that makes it easy for the public to examine and understand distributional effects.
The third reason is that distributional effects could lead policymakers to make substantive changes to a proposed regulation. These changes could include taking steps to mitigate undesirable distributional effects (whether through changes to the regulation’s design or through other authorities) or to choose a different regulatory alternative (including no action). Executive Order 12,866 explicitly directs agencies, when “choosing among alternative regulatory approaches,” to consider “distributive impacts” and “equity” when permitted by law. Exec. Order No. 12,866, at 51735. In fact, the prior Circular A-4 encouraged agencies to conduct an analysis of distributional effects precisely “so that decision makers can properly consider them along with the effects on economic efficiency.” OMB, Circular A-4 (2003), at 14. The proposed revisions to Circular A-4 take this further, stating that distributional analysis could help agencies “to better identify alternative regulatory options or impacts that can be mitigated through other regulatory or non-regulatory decisions, whether by your agency or others.” OMB, Proposed Circular A-4, at 62.
Recent scholarship explores how agencies might mitigate certain distributional effects. Richard Revesz, the current Administrator of the Office of Information and Regulatory Affairs (OIRA), described in a 2018 law review article how concerns about coal mining job losses motivated some congressional and some regulatory efforts to offset costs. Richard L. Revesz, Regulation and Distribution, 93 N.Y.U. L. Rev. 1489, 1543-55 (2018). In a recent article, I also discuss how distributional analysis could help agencies identify opportunities to use their other authorities, especially authorities explicitly meant to help offset regulatory costs, in similar ways. Caroline Cecot, Efficiency and Equity in Regulation, 76 Vand. L. Rev. 361, 421-24 (2023); see also Caroline Cecot, An Equity Blindspot: The Incidence of Regulatory Costs, J. Benefit-Cost Analysis (2023).
Nevertheless, while distributional analysis is useful for the aforementioned reasons, the reality is that agencies have not done much distributional analysis at all, at least not the kind of distributional analysis that could help agencies approximate aggregate welfare gains better, apprise the public about these effects, or make informed decisions based on the analysis. This is despite the fact that the prior Circular A-4 encouraged agencies to do distributional analysis, too, albeit in only two paragraphs. The question then is whether the expanded discussion of distributional analysis in the proposed revisions to Circular A-4, now totaling almost six pages, will have a positive effect on this stagnant area of regulatory analysis. Only time will tell—but I offer some thoughts, nonetheless.
Based on our review of the poor state of distributional analysis to date, my coauthor Robert Hahn and I speculate that agencies have not felt much pressure to do this analysis and have lacked the requisite data and resources to do it well. The proposed revisions to Circular A-4, by setting expectations and providing significantly more guidance about distributional analysis, certainly suggest that there is now pressure on agencies to take this analysis seriously. The revisions also at least acknowledge data issues. For example, the revisions suggest that agencies might plan to collect data on distributional effects when current data are limited or unavailable. OMB, Proposed Circular A-4, at 64. In the section discussing non-monetized effects generally, the revisions encourage agencies “to outline the data collection or analysis that would enable quantification or monetization” as it “may encourage research that would allow for such effects to be monetized in future regulations.” Id. at 44. These efforts hold promise.
Nonetheless, the revisions could do more to recognize how little attention is placed on cost incidence. Even when agencies have made some efforts to analyze distributional effects, they largely focused on only the distribution of benefits among groups. Analysis of the distribution of only benefits will always present an incomplete and potentially misleading picture of distributional effects. If the costs of a regulation are borne by, say, low-income individuals, then there is a real question about whether the regulation is net beneficial for them, even if such individuals are expected to disproportionately benefit from the regulation. That’s because costs might have a greater (negative) effect on the wellbeing of low-income individuals. These potential effects should be discussed openly and honestly, especially because highlighting these concerns might motivate efforts to offset these effects.
A related concern is squaring the revisions encouraging presentation of an income-weighted benefit-cost analysis with standard agency practice to use constant (or population-average) estimates for the willingness to pay for relevant risk reductions. The revisions, correctly, tell agencies not to apply income weights when benefits and costs reflect the use of constant estimates. OMB, Proposed Circular A-4, at 66. But this clarification then raises two issues. First, because constant estimates are the norm, it suggests that, in practice, the weighted benefit-cost analysis would rarely occur. Second, it misses an opportunity to address the underlying concern with using constant estimates in some contexts: that using constant estimates might suggest that benefits justify costs, but population-specific estimates would show that benefits would not justify costs. An agency’s decision to move forward in this context should come with an honest discussion of the need to alleviate costs for this group, whether through the regulation or through another action (as discussed above), so that the recipients of regulatory benefits can truly benefit.
On his first day in office, President Biden issued a memorandum on modernizing regulatory review, promising to ensure that regulations appropriately benefit and do not inappropriately burden disadvantaged groups. See Memorandum on Modernizing Regulatory Review (Jan. 20, 2021). More than two years later, under the leadership of OIRA Administrator Richard Revesz, the Administration has finally revealed its plan to accomplish this task. The revisions to Circular A-4 certainly raise the profile of distributional analysis and push agencies to increase their efforts in this area. I am hopeful that the revisions, if adopted, will stimulate the kind of informative and useful distributional analysis that could improve regulatory decisionmaking.
Caroline Cecot is an Assistant Professor of Law at the Antonin Scalia Law School at George Mason University.
 See, e.g., Lisa A. Robinson, James K. Hammitt, & Richard J. Zeckhauser, Attention to Distribution in U.S. Regulatory Analyses, 10 Rev. Env’t Econ. & Pol’y 308, 323 (2016) (examining 24 regulations issued during the Obama administration and finding that CBAs “rarely quantify” the distribution of risk reductions by population subgroups); Caroline Cecot & Robert W. Hahn, Incorporating Equity and Justice Concerns in Regulation, Regul. & Governance (2022) (examining 189 regulations issued between 2003 and 2021 and finding few examples of quantitative distributional analysis); Richard L. Revesz & Samantha P. Yi, Distributional Consequences and Regulatory Analysis, 52 Env’t L. 53 (2022) (analyzing in detail the sparse discussion of distribution in a few prominent regulations).