Essays in Applied Microeconomics

This dissertation explores topics in the field of applied microeconomics: labour supply, education, and health. The unifying theme across all three papers is the application of a modern econometric quasi-experimental methodological lens to rich datasets in order to estimate causal effects.

In the second chapter, I employ the twin birth instrument and census data to estimate the causal effects of fertility on female labour supply in more than 30 countries. I match the estimated motherhood employment penalties with gender wage gaps in each country and document a negative relationship between the motherhood employment penalty and gender wage gap, indicating that the effects of childbirth on employment are more negative in countries with larger gender wage gaps. The association is especially strong in within-country analysis and remains nearly the same when controlling for macroeconomic factors related to the labour market.

In the third chapter, I apply fixed effects models to a rich panel dataset of U.S. high schools merged with schools’ annual varsity team football win-loss records. Motivated by a recent literature documenting negative effects of sports success on the wider student body, I leverage the within-school variation in team success to estimate the causal effects of football team’s success on a variety of outcomes including measures of university aspirations, academic performance, and antisocial behaviour. I find that greater athletic success increases SAT/ACT test participation and AP course enrolment, implying positive effects on university aspirations. However, there are no effects of a team’s success on actual academic performance or antisocial behaviour. The findings are particularly important given previous work that has documented negative impacts of sports on college campuses.

In the final chapter, I use synthetic control methods with macroeconomic and remote-sensing nightlight data to estimate the causal effects of the 2003 SARS outbreak on China's provincial economies. I apply the algorithm to create a synthetic control for each of the most affected Chinese regions: Beijing, Guangdong, Hebei, and Shanxi. By comparing outcomes in a treated unit with its synthetic control, I find a long-term negative impact on night-time lights per capita in the affected Chinese region, Beijing. This suggests identifiable adverse impacts on economic activity from pandemic outbreaks prior to COVID-19.