Economic salience and differential abstention in presidential elections

Previous studies of American voting behavior have assumed that variations in turnout across election years are the result of fluctuations in the determinants of the voting decision. In this article I suggest that levels of turnout may also vary from year to year because of shifting relationships between economic variables and the voting decision. I use pooled cross-sectional data from the 1980, 1984, 1988, and 1992 American National Election Studies to test the hypothesis that Black, working-class, and unemployed individuals who stayed home from the polls in 1980 and 1988 became mobilized during the 1984 and 1992 presidential elections because of the nature of the economy and the party of the incumbent president. I extend this analysis to include attitudinal variables that measure perceptions about the state of the national economy and the effect of national policies on one's personal financial well-being. The results of the probit analyses suggest that (a) working-class individuals were less likely to vote in the 1980 and 1988 elections, but this variation vanished in 1984 and 1992, and (b) both ''pocketbook'' and ''sociotropic'' concern regarding unemployment led to increased participation in the 1984 and 1992 elections, in contrast to the 1980 and 1988 elections.