SESSION 2: Information Aggregation and Distributional Concerns

The second session of the workshop explored issues related to information aggregation and distributional concerns. The presentations focused on different aspects of strategic behavior in collective decision making, by either party representatives or voters. Richard van Weelden (University of Chicago) opened this session on information aggregation, analyzing the incentives of electoral candidates to adjust their positions to match the preferences of voters; sometimes referred to as pandering. His paper develops a model that shows why politicians are more likely to pander when voters are more informed. Furthermore, the effect of the size of the minority is not monotonic on pandering. In fact, larger minority sizes may favor the preferences of the majority, because socially valuable information is being ignored by a candidate that panders. Moving from elections to intra-party politics, Francesco Squintani (Warwick University) followed with a theoretical paper about the formation and influence of factions within political parties. He argues that the leader of a faction should be the member who is most skilled at aggregating information. Within the party ideological spectrum the member who can do best at aggregating information is one who is close ideologically to many other members, because such members are more willing to share information with him. Dimitrios Xefteris (University of Cyprus) examined the limitations of voters as recipients of information. He showed that the classical results of electoral competition do not hold when it is assumed that voters can only observe the relative ideological positions of parties before elections. In fact, in this case the most likely outcome is complete extremism in the parties' ideological positions, in contrast with the moderate outcomes predicted by existing theories.

Leeat Yariv (Caltech) opened the second part of this session with a social choice analysis of collective time consistent decision making. She showed that none of the voting rules that in general could be thought of as desirable satisfies the requirement of time consistency. This impossibility result is parallel to the famous Arrow impossibility theorem. Since time consistency ensures that decisions will stand up over time without commitment, this issue is of great relevance for all kinds of decisions. Gilat Levy (London School of Economics) offered a new explanation for a well documented empirical fact: some individuals with incomes below the mean income level vote for redistribution and some individuals above the mean income level vote against redistribution. Her explanation is based on income complementarities: individuals are better off when they are surrounded by wealthier people. If income distribution is very unequal, it is not very costly to organize large selective clubs, and this induces people with low income levels to vote against redistribution.