Age Gap? The Influence of Age on Voting Behavior and Political Preferences in the American Electorate
Holland, Jenny Lynn
MetadataShow full item record
Much is known in political science about the influence of sociodemographic characteristics such as gender and race on political preferences, but much of this research has ignored age in spite of the fact that age was one of the strongest predictors of vote choice in the most recent presidential elections. In 2008 and 2012, younger people overwhelmingly voted for Democrat Barack Obama, while older people instead voted for Republicans John McCain and Mitt Romney by wide margins. Why might age have such an important influence on vote choice? Three theories are commonly offered. First, the age similarity hypothesis predicts that individuals will prefer candidates closer to them in age, and thus younger people should prefer younger candidates while older people should prefer older candidates. Second, generational replacement explanations assume that younger people who enter the electorate are more liberal than their older counterparts, and thus younger people should prefer more liberal candidates relative to older people. Third, the mainstream-insurgent hypothesis assumes that younger people prefer political candidates that are viewed as ideologically extreme and challengers to the traditional party establishment, whereas older people prefer more moderate and traditional mainstream candidates. Therefore, younger people should prefer insurgent candidates while older people should prefer mainstream candidates. I explore these hypotheses using nationally representative data sets for the period 1976 through 2008 (U.S. primary exit poll data, American National Election Study data, and General Social Survey data) and original survey data collected the week prior to the 2012 presidential election. I find the most support for the ideology and mainstream-insurgent hypotheses; younger people are usually more liberal than older people and younger people usually prefer insurgent candidates over mainstream candidates compared to older people. I find the least support for the age similarity hypothesis; although there were many cases in which younger people preferred younger candidates, there were additional cases in which younger people actually preferred the older candidate, contrary to expectations. In total, these results indicate that age is an important predictor variable above and beyond ideology and partisanship, though its effects are conditional upon electoral context.