Abstract:

This dissertation has three chapters on the economics of development and health. The first chapter investigates the effect of natural resources on growth in a two-sector model where government taxes the non-resource sector and chooses institutional quality, which determines both productivity in the non-resource sector and the extent of resource rent expropriation by the government. The approach departs from the literature in that it abstracts from ex-ante institutional differences and endogenizes institutional quality, modeling the effect of resources on institutions. Resource booms compromise institutions and growth, and the effect on growth is exacerbated when the non-resource sector is relatively small and mitigated when it is relatively large. The intuition behind the result is that when resource rents are substantial, the potential for appropriating more of the rents makes the government willing to incur the cost of corrupting institutions, despite the ensuing reduction in tax revenues from the non-resource sector.

The second chapter uses a cross-country panel dataset to examine the effect of natural resources on growth and institutional quality, and finds that, as the model in chapter 1 predicts, resources have a negative effect on institutions and growth; the latter can be overcome when the non-resource sector is large (top 35th percentile of manufacturing exports per-capita). The results are persistent when different measures of resource and non-resource wealth are used, and robust to changes in the specification and econometric approach.

The third chapter considers the relationship between mortality, income and relative deprivation by using South African data to test the individual-level foundations of the Wilkinson Hypothesis. Relative deprivation gives a better account of mortality than does income, but for some population groups the fit is improved in specifications which include both income and relative deprivation. While no effect of either income or inequality is observed for the most privileged race-gender groups, for many of the less advantaged groups, income is both protective (particularly for the poorest) and a shield against the nefarious effects of inequality. The patterns are robust to other measures of health but difficult to generalize without looking further into the mechanisms through which inequality affects health.