Document Type


Publication Date



Conflicting views about presidential control of the administrative state have too long been characterized in terms of a debate over agency independence. But the term “independent” when used to describe administrative agencies carries with it the baggage of an unhelpful and unrealistic dichotomy: administrative agencies that are (or should be) subservient to presidential control versus those that are (or should be) entirely free from such influence. No agency fits into either category. This essay proposes reorienting the debate over presidential control around agency “autonomy,” which better conveys that the key issue is a matter of degree. Contrary to some proponents of the unitary executive theory, for example, all agencies — even departments headed by officials who can be removed by the president at will — should possess autonomy. They need autonomy to be able to make expert, evidence-based judgments and to avoid becoming weaponized purely for self-interested purposes or to gain partisan advantage. That said, in a democracy, agencies also ought never to operate with complete autonomy to the extent that they function with no constraints. The best notion of the administrative state borrows from anthropologist Sally Falk Moore’s notion of the semi-autonomy of law, leading to the conclusion that administrative agencies are semi-autonomous institutions.


administrative law, regulatory agencies, separation of powers, executive politics, bureaucracy, presidential oversight, accountability, Office of Information & Regulatory Affairs, OIRA, unfunded mandates

Publication Title

University of Dayton Law Review

Publication Citation

Based on remarks delivered on a plenary panel at the 2018 Federalist Society’s National Lawyers Convention; panel's full transcript at 44 U. Dayton L. Rev. 319 (2019)