Key points are not available for this paper at this time.
A model that reconciles the positive association between income and voting found in cross-sectional studies with the negative correlation between income and turnout over time is developed. Voting is modeled as a group rational activity. Voter turnout rates in the United States for the 1948, 1960, 1968, and 1980 elections are used to test the model. Voter turnout is found first to fall and then to rise as families move up the income distribution. Voter turnout also rises as real wages fall, as educational attainment increases, as elections become closer, and as poll taxes and literacy tests are removed.
Filer et al. (Mon,) studied this question.