Will South Dakota's demographic shifts lead to changes in voting behavior? A county-level analysis, 1980--2008.

摘要

The concept of generational cycles offers distinctive promise for understanding and thinking strategically about political attitudes and the relative strength and dominance of America's political parties. As patterns of immigration, urban development, family structure, religion, and other demographic indicators change the landscape of postindustrial America, the merit of anticipating their impacts only increases. The presidential election of 2008 offered an early glimpse of how these transformations are shaping American politics in dynamic ways. Demographic groups like Millennials, nonwhites, college-educated professionals, and the religiously unaffiliated comprised an unprecedented share of the U.S. electorate, and current estimates project these subsets to continue to grow in proportion to other segments of voters. This study explores the utility of a cohort succession model for explaining differentials in voting patterns in South Dakota for each presidential election from 1980 through 2008, while accounting for changes in party registration, education, median age, family structure, race, and religious affiliation. Cohort succession is determined to be a statistically significant and meaningful factor in explaining changes in partisan voting. Shifts in social patterns involving marriage and children are also determined to be closely associated with voting behavior. By seeking to capture both structural and demographic transformations in American society over time and the unique contributions of cohorts as mechanisms of social change, this study may serve as an exemplar for similar analyses in other states---particularly those in the agricultural Midwest. In addition, because the changing social patterns occurring in South Dakota may prove to be more gradual and subtle than in other states, this work may lend valuable insight into the political implications of demographic shifts nationwide. Indeed, the results of this study hold promise for a better understanding of voting behavior in past elections---and for predicting electoral outcomes in the future.