“Distributional Guardrails?”, by Matt Adler
This is the fourteenth post in a series on Michael Livermore and Richard Revesz’s new book, Reviving Rationality: Saving Cost-Benefit Analysis for the Sake of the Environment and Our Health. For other posts in the series, click here.
In their terrific new book, Reviving Rationality, Mike Livermore and Ricky Revesz describe the emergence of a bipartisan consensus in favor of cost-benefit analysis (CBA) that developed over the years from the Reagan to the Obama administration, and that was shattered by Trump’s. Of course, this consensus was hardly universal; for example, many if certainly not all environmentalists continued to reject CBA even after it became clear that CBA could justify strongly pro-environmental measures (as argued by Mike and Ricky in their brilliant and influential first book, Retaking Rationality). But many others in both the Democratic and Republican parties came to support CBA over the period from 1981 (when CBA was first firmly put in place as part of the regulatory process, via Executive Order 12291) through 2016. The Trump’s administration use of CBA shattered the bipartisan “guardrails,” as Reviving Rationality details.
Now that Trump’s Presidency is over, could a bipartisan consensus in favor of CBA along with some basic norms (“guardrails”) reemerge? Writing as of July 2020, before the election, Mike and Ricky are hopeful if not unmitigatedly optimistic (p. 10). Fifteen months later, I’m in the same place: hopeful (what do we have if not hope?), but not unmitigatedly optimistic—indeed, not optimistic at all.
CBA needs to be revised to incorporate distributional consideration. (Mike and Ricky agree; pp. 205-09.) Along with climate change, inequality (along many dimensions: income, race, gender, and others) is the major ethical problem of our time. The progressive wing of the Democratic party, which cares deeply about inequality—and rightly so—will surely accept CBA only if thus revised. I don’t say that revising CBA to incorporate distributional considerations is a sufficient condition for progressive buy-in—that would be much too optimistic—but surely it is a necessary condition. Indeed, in a memorandum to the head of OMB that he issued on his first day in office, President Biden directed OMB to (among other things) “propose procedures that take into account the distributional consequences of regulation.”
Good news: the intellectual tools exist for rebuilding—improving—CBA, so that it reflects distributional impacts. An established scholarly literature in welfare economics documents how to adjust CBA with “distributional weights.” Distributional weights that are a function of individual income and perhaps other attributes are applied to individuals’ willingness-to-pay amounts, so as to give greater weight to the worse off. Distributional weights can be specified so as to mimic a utilitarian social welfare function (thereby reflecting the diminishing marginal utility of income) or a “prioritarian” social welfare function (which gives greater weight to the worse off beyond the diminishing marginal utility of income). A different possibility is to undertake CBA in tandem with inequality metrics—evaluating both how a regulatory measure fares with respect to CBA, and its impact on inequality. Such metrics could take the form of the Gini or Atkinson indices, quantifying the population distribution of income or other resources; but could also take the form of “social gradient” metrics such as the concentration curve, measuring skews between different social groups.
But what are the prospects for a bipartisan consensus in support of distributionally refined CBA? Consider what needed to happen to yield a consensus in support of using CBA to monetize the benefits as well as costs of lifesaving regulations (a major application of CBA in the pre-Trump period). We needed not merely the intellectual tools (namely, the “value of statistical life”: VSL), but also: (1) broad acceptance by applied economists of the VSL concept and intensive work on their part to estimate VSL values, undergirding the use of VSL by agencies; (2) support on the part of both Democratic and Republican administrations for monetizing lifesaving; and (3) eventual support by the federal courts, which (as Mike and Ricky describe; pp. 188-90) are now inclined to read statutes as permitting CBA, in particular CBA with lifesaving monetized (although courts may strike down particular agency applications of CBA as arbitrary).
By analogy, for a bipartisan consensus in support of distributionally refined CBA to emerge, we need not just the tools (distributional weights and/or inequality metrics) but also (1) substantial acceptance by applied economists of the use of those tools to evaluate regulation, so that the research community produces a body of empirical and modelling work that will help agencies apply them; (2) support by both Democratic and Republican administrations for distributional analysis; and (3) broad acceptance of distributionally refined CBA by federal judges (appointed by Presidents of both parties). You can see why I’m not optimistic.
Why should a Biden administration appropriately skeptical about the prospects for a consensus behind distributionally refined CBA undertake the effort to develop distributional methodologies? Why not just gut CBA? Two answers, the first tentative, the second less so. The first is that conservative judges may be somewhat less likely to strike down progressive regulation if supported by CBA of some sort, than if adopted without any CBA screen. The second is an R&D argument: distributionally weighted CBA is an ethical improvement over traditional CBA (or so I believe), and a distributional-analysis mandate issued by the Biden administration (even if overturned by a subsequent Republican President, just as Biden overturned Trump’s CBA executive orders) will build our knowledge base about these tools. At the very least, working at improving the methods of CBA now is an intellectual investment that can be hoped to inform the norms of CBA in a less polarized, more rational future.
Matthew D. Adler is the Richard A. Horvitz Professor of Law and Professor of Economics, Philosophy, and Public Policy, Duke University.