Subsistence Agriculture and Labour Reallocation in Developing Countries

Abstract

This dissertation examines the economic determinants of labour reallocation in Sub-Saharan African (SSA). I first document the historical trends in labour reallocation in 11 countries in SSA and use a general equilibrium model to quantitatively assess the extent to which standard theories can account for labour reallocation in SSA. The analysis shows that structural change has been slow in SSA and this was due to a combination of low productivity level in agriculture and slow productivity growth in non-agriculture. Moreover, standard theories can explain 50% of the decline in the share of employment in agriculture in only 40% of the sampled countries. Next, I study farm-level productivity in five countries in SSA. Using an empirical production function approach, I test if there are total factor productivity (TFP) differences across the subsistence and commercial-farm types and quantitatively assess the implications of farm commercialization for labour reallocation. The analysis shows no substantial differences in TFP across farm types. In the most ``optimistic'' case of Ghana, TFP is about 15% higher in commercial farms than subsistence farms. A counter-factual analysis of a 15% increase in agricultural productivity could lead to at most a ten percentage points reduction in the share of employment in agriculture in Ethiopia, Malawi, and Tanzania. Finally, I examine farm-level factors that stimulate agricultural commercialization in SSA. In particular, I estimate the likelihood of being a commercial versus a subsistence farmer and the likelihood of transitioning from one farm type to another based on observable characteristics. The analysis shows that although a substantial proportion of farms has no market participation in a given year, there are rich transition dynamics over time. The results from the probit regression show that resource endowments (land, labour, chemical use) and farm characteristics such as multiple-cropping, irrigation, crop type (fruits, vegetables, cash crops), and farm machinery use are positively correlated with market participation and the transitioning of subsistence farms into the market economy.