Policy Representation in the American States II

Indiana University, Bloomington IN

Investigators

Abstract

This renewal project continues work on developing new data and a new approach to the study of representation. It uses the legislatures of the states and the US Congress as 101 representative systems through which we can study the linkages between constituency, party and roll call voting. This comparative perspective provides substantial advantages over previous work by taking the research designs beyond the limitations of a case study of a single institution and by providing enough variation in institutional rules and procedures, electoral settings and constituency preferences to achieve deeper understanding of the factors that enhance or retard governmental responsiveness to citizen preferences. The greatest obstacle to this approach in the past has been a simple lack of data. Roll call records are used to gauge legislators' policy positions within the legislatures, but, except for the U.S. Congress, these are not compiled in any easily available format for scientific analyses. This project has addressed that shortcoming with the development of a methodology of computer-assisted extraction of roll call data from the variety of sources and formats in which this information is found in the states. This approach to data generation is dramatically more efficient and less error-prone than traditional methods of hand coding and manual data entry. The methodology has been well-developed during the initial grant period and virtually a full set of roll calls and matching presidential vote totals for legislative districts have been compiled for the 1999-2000 sessions. Initial analyses point to the central importance of parties in the electorate and in the legislatures for policy representation and electoral accountability, and the researchers have found that parties take on a variety of roles, both in elections and in the legislatures. This project continues that work on four fronts. First, analysis of the completed data sets permits verification and expansion on the initial findings concerning the role of party and constituency in legislators' roll call voting. Second, the approximately fifty thousand roll call votes that have been compiled are coded to facilitate analyses by policy area or types of bills. Third, information on legislative electoral returns and the socio-political backgrounds of legislators is added to the data collection to increase substantially the analytic leverage of the roll call and constituency information. Fourth, a longitudinal component to the data collection is undertaken to permit analyses of change and stability in representation and legislative policy making. One aspect of this is the replication of the 1999-2000 data compilation with a full post-reapportionment roll call record of the 101 legislative chambers. This adds significantly to the investigators' ability to establish the temporal stability of legislative coalitions and the ideological positioning of legislators over time as well as to assess institutional changes like redistricting and term limits. The second aspect of this is the collection of the full time series of roll calls for the approximately half of the legislative chambers for which roll calls are web-accessible. This is possible by building on the expertise developed in the initial grant period, so that for a relatively modest investment, the project will produce for the research community a quantum increase in the data resources for comparative study of legislative policy making in the U.S. Besides greatly expanding the data resources for research on representation and policy making in the American legislatures, the project contributes to our understanding of how political contexts, institutional configurations and diversity influence the effectiveness of elections and political parties as mechanisms of citizen control of government.