LABOUR REALLOCATION OUT OF AGRICULTURE. EVIDENCE FROM DEVELOPING COUNTRIES

Abstract

This dissertation studies the process of reallocation of labour out of agriculture in developing countries. I develop a two-sector general equilibrium model where land is a quasi-fixed factor, and population growth constrains the reallocation of labour from agriculture. Productivity growth in agriculture can alleviate this constraint. The quantitative analysis shows that during the 1970-2010 period, population growth accounted for most of the changes in employment in agriculture in developing countries, while the attenuating contribution of productivity growth was negligible. In the absence of population growth, compared to data, agricultural employment would have declined 1.5\% more, while agricultural labour productivity would have increased 0.5\% more per year. I, then, study the effect of Green Revolution in agriculture on the speed of labour reallocation out of agriculture. Green Revolution is a form of land-augmenting technical change which increases labour-intensity and hence, slows down the reallocation of labor out of agriculture. I present a model with land-augmenting technical change. Using cross-national data on the adoption of high yielding varieties technology, I find that the Green Revolution can explain 28 percent of the increase in agricultural employment in the adopting countries over the period 1965 to 2000. Finally, I study the relation between agricultural land productivity and share tenancy. Using district-level data from rural Bangladesh, this study finds no evidence of a negative relation between farm output per unit of land and share tenancy. The reason is that while, in Bangladesh, share tenancy is highly prevalent, the land is scarce, about ten percent of the farm households are landless, contracts are short term, and non-farm jobs are limited.