Essays on Individual Decisions in Large Collectives

Abstract

This dissertation consists of three essays in microeconomic theory on the topic of how individual decisions shape a large collective and how in return the large collective affects the individual behavior. Chapter 1, Voting with Endogenous Timing, studies the role of timing in common-value elections. It contributes to the branch of strategic voting literature by exploring the effects of an endogenous timing decision on voting procedures. In welfare-optimal equilibria, agents use their timing to communicate the strength of their private information. This communication allows for better information aggregation in the collective than simultaneous voting or voting with an exogenously fixed timing. Chapter 2, On-the-Match Search and Match-Specific Productivity Growth, which is joint work with Sophie Kreutzkamp and Axel Niemeyer, investigates a frictional search-and-matching model with heterogeneous agents that continue searching when matched. The novelty of our model is the combination of productivity growth inside a match with the option for both partners to rematch. We show that even minuscule productivity growth is sufficient to eliminate unreasonable equilibria that typically occur in models with on-the-match search. Therefore, vanishing growth rates can serve as a criterion for equilibrium selection in such models. Productivity growth can accelerate sorting in the market. In particular, the assertiveness of the market stems from the rematching behavior of both the less productive agents and the more productive agents. Chapter 3, Partnership Dissolution in a Search Market with On-the-Match Learning, analyzes the steady-state equilibria of a search-and-matching market with ex-ante homogeneous agents and on-the-match learning. As the continuation value of a match does not only depend on the profitability for oneself but also on the partner’s rematching behavior, agents have an endogenous interest in their partners finding the match profitable. Comparative statics show that a faster learning rate is beneficial for the agents if the profitability in a match has a strong positive correlation between partners, and, in contrast, that with a strong negative correlation, a faster learning rate reduces the ex-ante expected payoff of the agents.