UC San Diego

This dissertation is composed of three papers on distinct topics, each using a different method from applied microeconomics. In chapter 1, I study how candidates’ election results affect the future contribution behavior of their donors using a regression discontinuity design. I find that contributors to narrowly-winning House candidates are much more likely to contribute in the following cycle than contributors to narrow-losers. Much of this effect is driven by future giving to the same candidate, contrary to a “reinforcement learning” hypothesis. This has broader implications, as incumbents can rely more on contributions from past donors than can new challengers. I estimate that candidates from narrowly-winning parties receive $130K more in individual contributions than those from narrowly-losing parties in the following cycle, almost all of this coming from these repeat donors.

In chapter 2, we study the role of self- and social-image in social comparisons. We

propose and test a theory that casts peer effects as the result of a signal extraction. The theory posits that individuals receive signals of their own attributes through completion of costly actions. Signal extraction is improved through social comparisons with other’s actions. In experiments, we find that subjects choose to complete more real-effort tasks in exchange for charitable donations if they anticipate learning how their decisions compare to the choices of others. Further, differentiated responses to noisy or refined social information adhere to the dynamics of signal extraction.

In chapter 3, we characterize how municipal governments respond to economic fluctua- tions, using employment shocks as a proxy. We specifically study the role Tax and Expenditure Limitations play in this response. We find that, following a positive employment shock of one percent, limited municipalities persistently lag behind their unconstrained counterparts in capital- intensive spending, with little differential effect on public safety and administrative expenditures. Our findings illuminate an unintended consequence of fiscal responsibility measures in U.S. cities: limits designed to restrain the size of government may instead alter the government’s spending mix, inducing investment cuts that allow a government to maintain patterns of administrative and public safety spending.