Policy Balancing and Preferences for Party Control of Government

Abstract

Divided party control of government is one of the defining features of contemporary American politics. Of the competing theories offered to account for this phenomenon, the notion of policy balancing offered by Fiorina (1988, 1992, 1996) and Alesina and Rosenthal (1995) has garnered the most attention, though it has received limited empirical support. The lack of support stems in part from: (1) the failure to separate voter preferences for party control of government from the actual act of casting a straight or split ticket, (2) using suspect measures of preference for divided government, and (3) ignoring voters who support unified control of government regardless of which party has that control. In this article, we provide a direct test of the policy balancing thesis that addresses these issues. Specifically, we examine how citizen preferences for partisan, divided, or unconditional unified control of government depend upon their own ideological locations and their perceptions of the locations of the two parties. Using original data collected from a statewide survey in Illinois in 2000, we uncover clear support for the policy balancing argument.