Beyond BLS briefly summarizes articles, reports, working papers, and other works published outside BLS on broad topics of interest to MLR readers.
In “Employment, wages, and voter turnout” (American Economic Journal: Applied Economics, October 2013), Kerwin Kofi Charles and Melvin Stephens, Jr. investigate the relationship between labor market activity and voter turnout. Specifically, they address this question: Are citizens more or less likely to vote when wages are high and jobs plentiful?
Charles and Stephens build several models to address their question. They use county-level data covering three decades (1969–2000) for five offices: President, Governor, U.S. Senate, U.S. House of Representatives, and state legislator. The county-level data are compiled from various sources, including the Bureau of Economic Analysis, the Regional Economic Information System, and the Bureau of Labor Statistics. In the authors’ models, they account for state differences, year differences and any state-specific factors that vary across election years. They also create models to account for variables that have interrelated effects on each other. For instance, the county in which voters live may affect wages and voter turnout, although wages may affect voter turnout.
Charles and Stephens find an inverse relationship between voter turnout and both local wages and employment. They find that higher local wages and employment are associated with lower turnout in each type of election they examine except presidential elections. The authors believe that leisure time and obtaining more political information are complementary; hence, increased employment lowers the exposure to media and thereby decreases political knowledge. The authors find that more time spent at work relates to less time available to become informed about elections.
They also find that the better the local labor conditions, the more people will vote in just the more prominent of two elections on the same ballot, even though there is no additional logistical cost for voting in both. A presidential election, for instance, is easier for a worker to become informed about than a less publicized election.
Moreover, the authors’ analysis of county-level data shows that voters are not more likely to show up to vote if they’re pleased with the economy’s performance under an incumbent. Further work on the relationships among turnout decisions, employment level, and economic performance could help analysts understand why one candidate gets elected instead of another.