Abstract:
Natural resource abundance is generally considered to be a curse rather than a blessing for economic development. This dissertation examines selective mobility patterns and changes in educational investments in response to natural resource booms. The first study complements a theoretical model with an empirical investigation in order to relate resource abundance and selective international migration patterns. Theoretically, I combine a Dutch disease model with a Roy-Borjas setup in order to elaborate on the relationship between resource shocks and migrant selectivity. In this regard, I predict that skilled labor is relatively worse off in the course of a Dutch disease, incentivizing brain drain effects. Empirically, I provide evidence for the effect of resource shocks on migrant selectivity based on a simultaneous equation model in order to disentangle income inequality effects and migrant selectivity effects. The results show that resource shocks foster brain drain effects. The second study examines selective regional migration patterns arising as a consequence of resource booms. I show that natural resource booms lower the relative educational background of immigrants. This especially holds if resource abundant states embark on a policy of resource transfers. The theoretical predictions are consistent with interstate migration patterns in the US throughout the 20th century. The third study sheds light on the relationship between resource abundance and educational investments. In particular, I make use of a simple model of human capital formation, showing that resource windfalls might lower labor supply and returns to skills in the future. In light of lower returns to skills in the future, residents invest less in education at the present. I make use of an enormous oil boom in Prudhoe Bay, Alaska, in the 1960's in order to test the theoretical predictions. The results of a difference-in-differences model based on US census data indicate that the resource windfall led to a shortfall of educational investments compared to a control group composed of several US states.