Accounting for disparity: Essays on agricultural development and racial earnings inequality

This thesis consists of three essays which focus on accounting for the sources of economic inequality in various contexts. The first two essays focus on disparity of agricultural labour productivity across the developing countries, while the third analyzes racial earnings inequality. The main goal of these essays is to shed light on some of the mechanisms which have generated these types of inequality, in order to better design policies for ameliorating them. The first essay is empirical, while the second and third essays construct general equilibrium models so as to quantitatively assess the importance of proposed sources of inequality.;The second essay analyzes the factors which have contributed to agricultural stagnation in Sub-Saharan Africa, despite productivity improvements in agriculture in other developing regions. I construct a quantitative model which can match average Sub-Saharan African trends of agricultural labour productivity, crop yields and input use from 1965 to 2000. The model points to key factors which have constrained agricultural productivity growth over this period, and to the need for diverse yet concerted policies to arrest this stagnation.;The third essay presents a quantitative model which sheds light on racial earnings inequality in the U.S., South Africa and Brazil. This model indicates that a large proportion of the racial wage gap in these three countries can be attributed to differential human capital accumulation by race. Most notably, distortions created by the explicit, racially-biased education system which existed in South Africa during Apartheid can explain roughly three quarters of the racial wage gap in South Africa in the early 1990's.;The first essay finds that a large proportion of the variation of the level and growth of agricultural labour productivity across a sample of developing countries can be explained by variation in input use across countries. I demonstrate that our understanding of disparity of labour productivity in developing country agriculture can be significantly improved by accounting for variation in the adoption of high-yielding seed varieties and for correlation between input use and technological change across countries.