Policy Voting in the U. S. Senate: Who Is Represented?

This study develops new state-level measures of mass and party elite ideology to assess senatorial responsiveness to different constituencies. The ideological preferences of two groups-independent identifiers and senators' state party elites-prove important in accounting for senatorial Conservative Coalition scores (1981-84), while the preferences of state partisans have no direct effect on roll-call voting. The often noted ideological differences between same-state senators from different parties are explained by the differing values of their respective state party elites. Challengers, in contrast to incumbents, are unresponsive to independents; their campaign issue stances are consistent with the more extreme values of their state party elites. As a result Senate elections tend to be contests between relatively moderate incumbents and ideologically extreme challengers. Popular understanding and academic theory agree that a fundamental feature of modern democratic government is the representation of constituency interests by elected legislators. Studies of representation examine how constituency opinion is related to legislative attitudes and behavior and what conditions promote or inhibit this central relationship. In the theoretical and empirical work on legislative elections, a growing number of studies deals with the role of issues and policy in explaining Senate elections. This work assumes that elections are the chief mechanism for achieving accountability and for insuring legislative responsiveness to constituency preferences. Several features of the contemporary U.S. context shape the potential for legislative responsiveness to constituents. Our electoral rules-single-member district elections, the direct primary, and the Australian ballot-enhance legislative responsiveness to voters rather than to party or the president, as does the decline in partisanship among voters and within Congress. Split-ticket voting is at such levels that while many seats are safe for specific incumbents, fewer and fewer can be considered safe for the parties. Of course the institutional changes in Congress-decentralized power, increased perquisites of office with