The Elusive Search for Presidential Power

Abstract

Unilateral presidential actions, such as executive orders, are widely cited as a key strategic tool for presidential power. However, is unilateral action evidence of unilateralism or might it represent executive acquiescence? We answer this by (1) specifying three competing models, each with a different presidential discretion assumption and generating alternative hypotheses; (2) extending the canonical item‐response model to best measure executive‐order significance; and (3) comparing competing theoretical models to data for 1947–2002. Theoretically, we show that legislative preferences may impact unilateral actions differently than previously thought and indicate how parties may be influential. Empirically, a model where the president is responsive to the chamber's majority‐party median fits the data better than models assuming responsiveness to the chamber median or no presidential acquiescence. Unilateral action appears not tantamount to presidential power, as evidence implies that legislative parties, or the judicial actors enforcing their will, are key conditioning factors.