Every year, thousands of law students take a course in administrative law. It’s a great course, and we wish even more students took it. But there’s a risk that students may come away with a vision of the regulatory process that is increasingly disconnected with reality. Worse, the leading judicial opinions on the subject suggest that judges may from a similar disconnect.
The Administrative Procedure Act is based on the premise that Congress delegates the power to address a problem to an agency, which then applies the statute to formulate a regulation. Policy is driven by the statute along with the views of the agency head, who is appointed by the President and confirmed by the Senate. But the realities are often different. Policy is often driven, not so much by Congress, as by Presidential orders requiring the use of cost-benefit analysis. The final decision about whether to regulate, and even the details of the regulation, may be decided by a White House office called OIRA. The head of the agency is frequently a temporary appointee, generally a lower level agency official who may not have much clout within the executive branch. The regulatory system as it actually operates is much different from the world envisioned by administrative law.
In a recent paper, Anne Joseph O’Connell and I document this disconnect and discuss its consequences. We think it likely that something like the current system will persist. Administrative law aims to make the regulatory process more open and transparent, more faithful to statutory mandates, and more attentive to scientific expertise — all while respecting the primary role of the executive branch in issuing regulations. To further these goals given current realities, OIRA process must become much more transparent and accountable. Transparency will help ensure that an agency’s statutory mission and its scientific expertise don’t get submerged by OIRA staff who care only about their own policy goals and lack deep expertise. Regardless of whether you share OIRA’s passion for cost-benefit analysis or revile it, we should all be able to agree on the need for improving the process.