Public Opinion in the American States: New Perspectives Using National Survey Data

Abstract

General measures of ideology and partisanship derived from national survey data concatenated to the state level have been extremely important in understanding policy and political processes in the states. However, due to the lack of uniform survey data covering a broad array of survey questions, we know little about how specific state-level opinion relates to specific policies or processes. Using the General Social Survey (GSS) disaggregated to the state level, we develop and rigorously test specific measures of state-level opinion on tolerance, racial integration, abortion, religiosity, homosexuality, feminism, capital punishment, welfare, and the environment. To illustrate the utility of these measures, we compare the explanatory power of each to that of a general ideology measure. We use a simulation to clarify conditions under which a national sample frame can produce representative state samples. We offer these measures to advance the study of the role public opinion plays in state politics and policy. The public opinion-policy linkage is a crucial topic for democratic theorists and has preoccupied students of state government and politics for years. Without survey data at the state level, pioneering studies employed surrogates derived from demographic variables or simu? lations to judge the responsiveness of state policymaking to public prefer? ences (Plotnick and Winters 1985; Weber and Shaffer 1972). Some ingenious studies also explore the causes and consequences of public opinion using national survey data disaggregated to subnational units (Gibson 1989,1992,1995; Miller and Stokes 1963; Norrander 2000). Wright, Erikson, and Mclver's research (1985) significantly advanced our understanding of the state public opinion and policy linkage by pooling 1976 through 1988 CBS/New York Times polls and disaggregating them to the state level to create reliable, stable, and valid measures of state ideology and partisanship (Erikson, Wright, and Mclver 1993). A host of influential studies employ these measures (e.g., Hill and Hinton-Anderson 1995) to illustrate fundamental linkages between general mass political at? titudes and the general choices of state policy makers. Yet, they represent only a first step in gauging the effects of opinion on state policy. The gen? eral nature of the ideology measure developed by Erikson, Wright, and Mclver leaves open many remaining questions about how specific attitudes may influence specific political outcomes and processes in the states.