Policy Representation in the American States

Indiana University, Bloomington IN

Investigators

Abstract

Representation lies at the very heart of democratic processes. Although there are many definitions of democracy, almost all include the need for some level of responsiveness of governors to the will of the governed. This project seeks to extend studies of representation to the American states. It follows the tradition in the literature of examining representation in terms of the congruence between the policy preferences of citizens and the legislators who represent them. The innovations of this project are (1) the extension of theories of representation into the comparative institutional and electoral contexts of the American states; (2) the development of new data resources on legislative policy behavior and constituency preferences. Most of the empirical work on representation focuses on the United States House of Representatives and, consequently, the lack of variation in electoral and institutional conditions means that there is little theoretical development of the conditions which enhance or retard effective representation. The major reason for this is the paucity of data in contexts other than the Congress to address empirically questions of representation. This study includes all state legislatures of the United States. The comparative analysis examines how the varying electoral, constitutional and district features of the states influence the process of policy representation. The project also examines and attempts to account for variations in the intensity of partisan conflict among state legislatures, drawing substantially from the recent literature on the impact of party with the Congress. A major part of this project is the collection of data for the analysis of representation in the states. The biggest task is the collection of a comprehensive set of roll call votes in the 50 states. Two technological advances makes this effort possible: most state journals, the source of roll call data, are not available over the Internet and text-processing technology allows the conversion of the text of legislative journals into roll call data sets. The period covered is the 1999-2000 terms (varying them where appropriate for states with odd-year elections). These data are supplemented with a comprehensive collection of interest group ratings of state legislators, and surveys of state legislative candidates. The surveys serve as validation of the roll call scales and to establish a common metric for roll call-based measures which enhances their value in comparative analysis across chambers and states. On the voter side, the project uses an extensive collection of demographic characteristics available for the state legislative districts. The project also processes newly available presidential voting measures, which serve as an indicator of voter preferences. The project promises to make important progress in our understanding of how electoral and institutional differences among the states shape the ability of citizens to achieve effective policy representation. It also produces important new data for use by the state, electoral and legislative politics communities. The approach and the lessons learned have implications for the study of representation well beyond the context of the American states.