Exercises for seminar 3 - Feb. 28

A. Redistribution: Theoretical predictions

1. Explain what the median voter model predicts on the relationship between income inequality and the magnitude of redistribution.
2. What would a model of probabilistic voting say on the same relationship?
   To answer the question you may either have a continuous distribution of income and hence an infinite number of groups, or focus on a given number of groups. You may (at least initially) assume that the number of swing voters is the same in all income groups.
3. Can probabilistic voting models give the opposite conclusion than the median voter model?

B. Redistribution: Empirical findings

1. What are the data required to test such for the presence of a positive relationship between income inequality and redistribution? Pay particular attention to how inequality should be measured. Which regressions should we run? What would you expect to find?
2. There may be reasons to believe that there are variables affecting both the income distribution and political preferences. Give some examples of why this may be so. What are the consequences for findings from OLS regressions as explored in the question above?
3. Barth and Moene (2013) suggest using wage bargaining regime as an instrument for inequality to study the impact of inequality on redistribution. Explain what an instrument is and what requirements have to be satisfied for an instrument to be valid. Discuss whether Barth and Moene's instrument is valid.