Individual Submission Summary

Direct link:

Empirical Implications of Presidential Centralization with Politicization

Sat, August 31, 12:00 to 1:30pm, Marriott, Harding


Both centralization and politicization serve as key presidential tools for influencing the creation of policy. However, each involves a distinct set of trade-offs. Centralization requires using extremely limited bureaucratic capacity within the Executive Office of the President, resulting in policy near the president’s ideal point but with high opportunity costs. In contrast, delegating to a politicized agency lowers the cost of policy creation, but any attempts to increase ideological congruence through politicization may lower agency capacity. When both strategies are modeled together, this changes the president’s use of each strategy compared to when they are examined in isolation, as most existing research has done. Furthermore, consideration of electoral accountability may further constrain the use of these strategies. In this paper, I draw from a formalization of these concepts and use an original dataset to test a number of new predictions about centralization and politicization, including: (1) there is a non-monotonic relationship between politicization and agency ideology; that is, presidents are initially more likely to politicize as ideological distance from the agency increases, but as the distance becomes greater, centralization eventually becomes the preferred choice; (2) given constant ideological distance from the agency, more extreme presidents are more likely to politicize; and (3) presidents are less likely to centralize in their first term. Finally, I discuss implications with respect to the current administration, which has seemingly abandoned some aspects of both centralization and politicization.