UC Irvine

This dissertation consists of three essays in the political economy of development. It uses a combination of empirical methods and microeconomic theory, utilizing both historical and contemporary data.

The first essay examines the long-run effects of forced migration on the origin economy, using Czechoslovakia's expulsion of 3 million Germans after WWII. For identification, I use the discontinuity at the border of the "Sudetenland" region where Germans lived, as made formal with the Munich Agreement in 1938. Since Germans had similar socioeconomic characteristics to Czechs, this bypasses factors that might drive effects elsewhere, such as differences in human capital and geography. The expulsion produced differences in population density, sectoral structure, and education between neighboring municipalities, which persist 70 years later. I trace effects to a selective resettlement of affected areas, generating de-urbanization and human capital decline. Empirical and historical evidence suggest agglomeration economies and extractive institutions as two forces driving this response.

The second essay examines how formal institutions influence local recovery to population shocks, using a model with multiple regions and increasing returns to economic activity within regions. Extractive institutions crowd out productive activity, making its spatial coordination more difficult in the aftermath of large, negative shocks. Given this, I show that when one region experiences such a shock, extractive institutions can hinder recovery, ensuring a redistribution of productive activity away from that region over the long-run.

The third essay considers the conditions under which nondemocratic regimes invest in public education. Nondemocratic regimes face a tradeoff when investing in public education. Education promotes human capital acquisition, expanding the tax base. Yet it also enhances political sophistication and participation, at a cost to nondemocratic regimes. To relax this tradeoff, a regime can disseminate propaganda through its education system. I show that even Bayesian citizens can be influenced by propaganda. By deterring political opposition, propaganda can induce nondemocracies to invest in education when they otherwise would not, improving social welfare. When propaganda is too strong, however, it can generate a backlash. Using cross-country and survey data, I find evidence consistent with the predictions.