>
SESSION 3: Dynamic Models and Models of Turnout

The presentations of the third session were on dynamic models and voter turnout, respectively. First, Marco Battaglini (Princeton University) explored the incentives to free ride in a dynamic framework considering that decisions may be reversible or irreversible. The main point of the paper was to characterize and compare the Markov perfect equilibria under the two polar assumptions about reversibility. The main finding is that equilibria are less inefficient if investment is irreversible. He also showed that with large depreciation rates it is more difficult to reach the optimum, and that there are non-monotonic equilibria where the level of public good converges with oscillations and limit cycles. Then, we moved on to a specific case of free-riding problems that refers to the decisions made by cabinet members. Torun Dewan (London School of Economics) showed that in this set up the solution to free-riding problems creates hold up problems, which can be alleviated with elections with optimal retentions. John Duggan (University of Rochester) presented a paper that develops a general model of dynamic elections. In his model, the preferred policy of the median voter evolves over time endogenously as a function of the policy choices, and he is able to identify general conditions under which a dynamic version of the median voter theorem holds. Marina Agranov (Caltech) discussed the results obtained from laboratory voting experiments that explored the effect of polls on voter turnout. The surprising result is that polls do not necessarily increase turnout, except for the majority party. This contrasts with the theoretical prediction that turnout should increase overall with polls, and the strongest effect is to increase turnout by the minority party. In the last paper, Santiago Oliveros (University of California) presented a study that offers one explanation for why voters tend to demand moderately biased information from the media. That is, leftist voters will tend to obtain news that is biased toward a leftist position, while rightist voters do the opposite.