Three Essays on the Political Economy of Development

Abstract

In Chapter 1: Research has shown that government spending can affect GDP growth rates, yet there is no comprehensive study that looks at how a country's choice of political institutions affects government spending. Using a panel of 92 democracies, this paper focuses on how the choice of regime type (presidential, parliamentary or mixed), legislative chamber structure (bicameral or unicameral), legislative chamber size, and electoral rules affect the level of government spending. The results show that the relationship between legislative chamber size and government spending is linear in unicameral countries but non-linear in bicameral countries, plurality electoral rule is always associated with less spending than any other type of electoral rule, and unicameral and bicameral countries should not be modeled together.

In Chapter 2: In a panel of 18 Latin American countries from 1900-2007, we test the degree to which institutions and geography affect country income. Using a new instrument, we find strong evidence that both institutions and geography are important determinants of country income. However, the penalty for economically unfavorable geography is much smaller than the potential benefits from good institutions. The coefficient estimates do not vary significantly when there are changes in the number of countries included in the analysis; the results for institutions are robust to the inclusion of country fixed effects.

In Chapter 3: Latin America's turbulent political history and marked dependence on commodity exports naturally raises questions about their interdependence. Constructing a new panel dataset from 1919-2000, I examine how international prices for Latin America's principal exports influence national political stability. The data show a significant effect of commodity prices on political instability. This significance holds when disaggregating each country's commodity price index into point source commodities (e.g. natural resources and plantation crops) and diffuse commodities (e.g. small farm production and livestock) and these results are consistent when using a count model. Further disaggregation reveals that the point source effect is primarily driven by increases in mineral prices.