UC San Diego

This dissertation is composed of three chapters on distinct topics within applied microeconomics. Each chapter uses econometric methods to analyze data from administrative sources, aiming to establish causal relationships between variables of interest.

In chapter 1, I study whether social interactions with residential neighbors in very close geographical proximity influence students' educational outcomes. In a cross-sectional analysis, I show that close neighbors have more similar levels of English achievement than pairs of students living in the same neighborhood but farther away. In a longitudinal analysis, I show that in the case of both English and behavior, close neighbors’ outcomes are positively associated with own outcomes. I also show that pairs of close neighbors are more likely to attend the same school, compared to those living in the same neighborhood but farther away.

In chapter 2, we study how extreme weather and natural disasters affect political outcomes such as campaign contributions and elections. We suggest that these weather events increase the salience of climate change and also lead some individuals to update their beliefs about it. In a short-run analysis, we find that the number of contributions to the Democratic Party increases in response to weekly average temperature, and the effect is stronger in counties with more anti-environment incumbent politicians. In a medium-run analysis, we find suggestive evidence that, following a natural disaster, the level of competition of elections rises, as measured by the probability of a challenger entering the race, the number of donors, and total funds raised.

In chapter 3, I study the effects of increased unemployment insurance benefit duration on individuals' short and long term labor market outcomes, using a regression discontinuity design. In the short term, I show that for a subset of workers, receiving 60 additional days of UI benefits induces them to spend 21 more days on UI and 14 more days in nonemployment. In the long term, a similar sample of workers spends 7 additional days on UI as a result of extended benefits, compared to the short term. Furthermore, I find suggestive evidence that the effect on days spent in nonemployment in the long term is lower than in the short term, which is consistent with increased post-unemployment job stability.