Implementing Executive Orders in the Administrative State

ABSTRACT Scholarship once counted the raw number of executive orders (EOs) issued as a kind of proxy for presidential power; now it grapples with whether such directives are actually implemented, a challenging puzzle since agency action is difficult for scholars to systematically observe. This article addresses that difficulty by examining bureaucratic promulgation of a final rule responding to a given EO. To factors previously hypothesized to matter for the likelihood of implementation it adds new ones, notably (1) centralization, or the extent to which an agency itself was involved in the formulation of an EO directing it to act, and (2) an agency's structural insulation from presidential control. In a set of 103 executive orders issued between 1988 and 2004, we find that centralization has little independent effect but has important interaction effects with agency insulation. More insulated agencies are less responsive than their more permeable peers to EOs created via centralized processes that exclude their input. Other factors predicting the likelihood of responsiveness are the specificity of the presidential directive and whether the agency has the substantive capacity to comply—as well as the bureaucratic influence not to.