*This is the thirteenth 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.

Does the executive branch have an obligation to abide by past practices when it comes to cost-benefit analysis (CBA)? If so, what is the source of that obligation and does it matter whether past decisions have been bipartisan in nature? These questions arise after reading Michael Livermore and Ricky Revesz’s important retrospective on the Trump Administration’s collision with CBA. One of their central critiques is the extent to which agencies failed to follow a bipartisan set of cost-benefit approaches. For example, they excoriate the Trump Environmental Protection Agency (EPA) for ignoring indirect or ancillary benefits, despite “long-standing consensus, supported by economic theory, logic, and a consistent administrative practice by administrations of both parties over decades.” (p. 151).

One could argue, however, that it is unclear—at least at first glance—why this is a problem; after all, presidents are empowered by electoral mandates to pursue changes, sometimes radical ones. Shouldn’t they be able to do so with CBA? But Livermore and Revesz are right to draw on a shared sense that CBA demands some kind of consistency, both across and within administrations. In this post, I briefly consider some potential explanations for this sense and conclude that there are conventions of CBA, internally-compelled patterns of administrative behavior.

First, take the argument that Trump flouted a kind of CBA stare decisis within the executive branch. In this view—richly developed by Nestor Davidson and Ethan Lieb—there is a kind of “common law” CBA that has developed over time that demands respect and adherence, much in the same way that judicial precedents do. A critical premise for this position is that CBA consists of a number of “legal” judgments, perhaps as interpretations of underlying executive orders. Suffice to say, however, that many features of the centralized review process lack “law”-like qualities. For example, determinations by the Office of Information and Regulatory Affairs (OIRA) are arguably not “binding” in the same way that, say, Office of Legal Counsel decisions are. Some have characterized OIRA, in fact, as more of a “convener” than a decider and argued that the agencies have independent reasons to want to abide by the process. OIRA’s decisions are also often non-transparent, both to the public and to the agencies themselves. OIRA has also been accused of procedural unfairness, that is, of being “ad hoc and chaotic.” Finally, the outcomes of the process do not always require internal or external justification; to the contrary, they are sometimes “simply mysterious.”

If the notion of legal stare decisis does not explain the pull of CBA consistency, then perhaps the concept of an emerging scientific consensus does. The concept is imprecise, but generally refers to a collaborative set of judgments or opinions of scientists in a field of study. These scientists achieve consensus through scholarly debate and publication, replication, and consensus conferences. In this view, perhaps CBA should be understood as a scientific, technical endeavor performed by economists. The executive branch, in turn, either channels or reproduces this emerging external consensus over time, which explains why departures from this consensus should be condemned. The problem with this argument, however, is that CBA is inherently value-laden. It requires resolutely political, even philosophical, choices: tradeoffs, for example, between past and future generations, decisions about whose benefits and costs should count. This is not to say that CBA is political all the way down; only that CBA as practiced by the executive branch cannot be understood as wholly neutral and technical in nature.

Finally, consider the idea, then, that the practice of CBA has instead given rise to a set of conventions. Under one definition, conventions are unwritten “regular patterns of political behavior . . . followed from a sense of obligation.” In other words, political actors exhibit empirically verifiable patterns that flow from a sense of internal compulsion, rather than mere habit. In this sense, conventions occupy a middle space between “law” and “politics.” While they are often thought of as “unwritten,” they can, in fact be written down, but they do not draw force solely from being documented. Moreover, conventions themselves are not directly enforceable in courts, but can be “indirectly recognize[d] . . . in the course of performing their usual duty of interpreting laws or rules of common law.”

Applying these ideas here, Livermore and Revesz meticulously document regular patterns of behavior with respect to CBA across administrations. To return to the opening example, the EPA since at least the Reagan Administration has incorporated indirect benefits in guidance and rulemakings. The harder question is whether presidents have done so over time out of a sense of obligation, a question to which we will return. In the meantime, note that the importance of indirect benefits has been memorialized in important guidance documents such as Circular A-4, which does not in itself give the requirement force. Livermore and Revesz also show that courts have sometimes required the consideration of co-benefits. But judges have nominally done so in the course of interpreting underlying statutes or engaging in arbitrariness review under the Administrative Procedure Act. Some of these legal justifications have arguably been strained, however, perhaps speaking to the “recognition” of cost-benefit conventions in court.

Turn now to the question of whether this pattern of administrative practice has arisen out of a sense of obligation and, if so, what kind. Analyzing these issues raises a number of methodological conundrums about how to know the reasons behind an individual’s behavior, none of which I can engage here. Perhaps one data point on this dimension, however, can be found in the writing of John Graham, a former OIRA administrator. As Livermore and Revesz point out, Graham co-authored a book advocating for the consideration of indirect benefits. His stated motivation for doing so was to “ensure that efforts to reduce health, safety, and environmental dangers are more fully thought through and more fully effective.” In other words, Graham professed a desire to want to reduce harms in a way that he thought required an accounting of indirect benefits—perhaps a thicker moral sense of obligation.

Much more must be done to write a fully convincing account of cost-benefit conventions (perhaps for future work!). Sources of obligation, for instance, may differ for different administrative actors. Career civil servants may abide by conventions given ideals of professionalism that may not be strictly legal or scientific in nature. The bipartisanship adoption of some conventions may reflect a thinner sense of obligation with game-theoretic premises embraced by political appointees. There is also the intriguing possibility that the morality of administrative law has been internalized within agencies. Though disparate, these motivations may nevertheless converge on particular cost-benefit conventions. These are just some of the many rich questions spurred by Livermore and Revesz’s Reviving Rationality.

Jennifer Nou is Professor of Law at the University of Chicago.