- 3 -
presidency” and on presidential appointments. Since the EOP’s creation in 1939, the president’s
personal and institutional staffs have expanded and, some argue, merged (Dickinson 1997).
And presidents have followed administrative strategies through which they attempt to
“implant their DNA throughout the government,” as one George W. Bush staffer put it (Allen
2004; more generally, see Nathan 1983). What is surprisingly lacking in the scholarly record to
date, however, is a systematic analysis of the relationship between the two strategies. The gap is
both theoretic and empirical. Yet understanding this aspect of presidential behavior is an
important concern since so much of executive policymaking now occurs through the
administrative apparatus of the presidency. How, and whether, presidents achieve the sort of
responsiveness they desire from the bureaucracy has important implications for the balance of
power between the branches, for the kinds of policy the government implements, and even for
assessing democratic governance.
In this paper we explore the relationship between centralization and politicization and
hypothesize about when presidents should prefer one strategy over another. Most basically,
are the two complements (as Moe implies) or substitutes (as Nathan [1983] suggests)? We
derive a theoretical grounding for the latter. New data on appointments across the various
departments, and research on the relative roles of the White House and departments in
presidential policy formation, are then used to analyze when presidents centralize policy
formulation. Importantly, we find that presidents centralize policy formulation less when the
relevant departments and bureaus are politicized. We conclude that the centralization and
politicization are substitute strategies rather than complements. The choice of strategies
depends on the broader political environment, presidential priorities, the capacity of the White
House and the departments, and the character of the proposals themselves.
Convention