Risk, sorting and inequality in Malaysia

This study estimates the effect of risk on choice of agricultural self-employment. Modern theory of contracts suggests both moral hazard and adverse selection cause the risk and monitoring costs of developing country agriculture to favor self-employment of farmers. These considerations provide a basis for a role for wealth in determining contract status. In this study, the effect of rainfall variation is incorporated in a model of comparative advantage in choice of sector and contract status in Malaysia. Following the approach of Heckman and Sedlacek (1985), the model corrects for selection effects in estimating earnings equations by category, in this case using a multinomial logit specification for choice. This model of the Malaysian labor force is used to examine the effect of sorting on earnings and inequality, and to simulate the effects of capital redistribution, risk reduction, and productivity changes. Important findings include a negative effect of weather risk on likelihood of choosing agricultural self-employment and a large reduction in inequality due to individuals sorting themselves by sector. Simulations cause substantial change in choices of categories, but only small changes in earnings, because sorting by comparative advantage implies that shifts are made by marginal individuals. This study improves on previous estimation of self-employment in low income agriculture, by incorporating risk, by correcting for sector choice, by using predicted education to avoid confounding evidence on ability, and by using only exogenous inherited wealth. Additional findings include: capital increasing the likelihood of agricultural self-employment, but the interaction between capital and risk failing to show the expected positive sign; and a positive selection correction coefficient in farming. Statistical tests, though somewhat ambiguous, tend to support intuition and other evidence questioning the specification of the choice equation. Simulations support the idea, associated with the Kuznets hypothesis, that productivity growth outside agriculture will raise income at the expense of equality, with the opposite effect when agricultural productivity increases.