CLEM BROOKS, Indiana University

DAVID BRADY, Indiana University

This study analyzes the effects of income on voting behavior and the outcome of presidential elections in the U.S. since 1952. The political effects of income have not been examined using appropriate measures and systematic analysis of individual-level data. Our analyses build from political behavior research on economic evaluations and sociological and economic research on trends in income distribution. We first consider the theoretical grounds for expecting income to have shaped voting behavior in the postwar era. Our analysis then shows that while income has had a significant and generally stable impact on voting behavior, cumulative changes in mean household income have had a very large effect on the outcome of presidential elections since the 1950s, with Republican candidates benefiting disproportionately. We find that household income shapes voting behavior indirectly, affecting policy evaluations, which in turn influence vote choice. The mediating roles played by evaluations of social welfare and government size help to explain why past increases in mean income have consistently benefited the Republican party. We discuss the contributions of this study to understanding income as an important but underappreciated source of political change in the U.S.

Over the past two decades, analysts of political behavior have increasingly focused their attention on the economic sources of both vote choice and change in the outcome of elections (see Fiorina 1981; Haynes & Jacobs 1993; Hibbs 1987b;

-1339-