Today’s Ad Law Reading Room entry is “How Not to Design Expert Bureaucracy: Lessons from Administrative Law,” by Wendy E. Wagner, which is forthcoming in the North Carolina Law Review. Here is the abstract:

Can we trust our agency experts to provide reliable scientific knowledge to inform policy? This question has worried academics, policymakers, and the general public for decades. Now, in the wake of expert agency debacles during COVID, the advent of a new presidential administration, and a Supreme Court intent on reshaping the structure of administrative law, these concerns are escalating.

This article offers one answer to this question by examining the architecture of administrative law itself, and the findings are not comforting. Under the law as currently designed, political officials within U.S. agencies and the White House––regardless of the president in power––can exert unrestricted control over the scientific staff at all stages of their work while also protecting these political interventions from public disclosure as deliberative process. And, while administrative law assumes that vigorous engagement by affected stakeholders will ensure the resultant work is at least not “arbitrary” in health and environmental regulation, the notice-and comment processes are typically monopolized by the same corporate interests that enlisted the political officials in the first place. At the same time, the staff’s anticipation of the resultant one-sided litigation only serves to introduce more biasing pressures on the objectivity of the work. And, if that were not enough, the deployment of elaborate external peer review processes, which are viewed as providing the last word on the quality of agency science, are entrusted not to disinterested scientists but to political officials. These officials enjoy ultimate control over the selection of scientists as peer reviewers and implementation of the review process, again in ways that remain largely undisclosed and often undocumented. As a result of this overarching legal design, even the most committed scientific staff find themselves impeded and sometimes blocked from producing work that has integrity, both with regard to scientific factfinding and to the identification of residual uncertainties. Indeed, it is not hyperbole to suggest that if one wants to know how NOT to design an expert bureaucracy, they should look to U.S. administrative law.

In designing a legal process to govern agency expertise, we can do better. To that end, the article closes with a reform proposal that encourages agency experts to demonstrate why their work can be trusted, a step that is not only omitted from current institutional design but is generally precluded as a legal matter. Rather than impose this demonstration as a mandatory requirement, the proposal recommends offering incentives for agencies to voluntarily document the reliability of their scientific analyses. In return, agencies would receive increased judicial deference. Ideally, this framework would also include a safe harbor provision that grants complete deference to fact-finding that meets the highest standards of scientific integrity.

Wendy Wagner has long been one of the most trenchant critics of contemporary American administrative law. As its title indicates, “How Not to Design an Expert Bureaucracy” does not suggest a change of heart. Rather, the piece is a full-scale indictment of the law governing agencies’ scientific decisionmaking. That law, Wagner argues, not only fails to support the integrity of agency science, it actively subverts it.

Wagner’s diagnosis of the problem is richly textured and deeply thorough. And yet, despite painting a picture the author herself describes as “grim” and “bleak,” the article ends on an almost hopeful note, suggesting a transparency-oriented reform proposal that would “re-anchor agency science” in the norms shared by the broader scientific community.

The Ad Law Reading Room is a recurring feature that highlights recent scholarship in administrative law and related fields. You can find all posts in the series here.