As Richard Nathan pointed out in The Administrative Presidency 25 years ago, much contemporary policy making occurs through the execution of the laws and the management process. Thus, the conditions under which administrative power can be effectively exercised are important components of future research. In short, we must revisit an old question: Who controls the bureaucracy? This article sets out a research agenda for doing so by examining a set of related strategies that presidents have utilized in their efforts to assert control over the bureaucracy-namely, centralization, politicization, and reorganization-in the course of linking two strands of literature with significant overlap but little conversation between them: the largely quantitative bureaucratic control literature, and the largely qualitative "politicized presidency" literature focused on presidential structures. Each strand, I suggest, can inform and enrich the other; both would benefit from better measures of inputs, outputs, and, crucially, the processes that connect the two and influence the success of policy implementation.A quarter century ago, Richard Nathan (1983) called renewed attention to the phenomenon of the "administrative presidency." In describing the strategies of the inner circles of Richard M. Nixon and Ronald Reagan as they sought to enhance the responsiveness of their respective executive establishments, Nathan's basic query did, of course, far predate his own study-for it boiled down to whether the president could control the bureaucracy and, if not, who did or could. And as Francis Rourke observed, that question of bureaucratic control "has haunted the relationship between the president and Congress from the very beginning of their history together" (1993, 687).Still, whether the question is 25 years old or 225, it is increasingly important that we turn our attention to answering it, as well as its corollaries. How do presidents seek to control the bureaucracy? How effectively do they do so? We might even ask, how
