Education and Inequality in Rural Bangladesh: A Longitudinal Study

Abstract

Education develops human skills, raises human productivity and, consequently, enables them with higher monetary incentives and better jobs. But the realisation of benefits may differ across income groups due to various limiting factors to achieve it. This article estimates the impacts of education on income and consumption of rural households in Bangladesh, using mean differential approach and unconditional quantile regression approach. It utilises Bangladesh Integrated Household Survey (BIHS) data for the years 2012 and 2015 to estimate the impact of education on the income and consumption of rural households. To address the potential endogeneity problem in impact estimation, ‘total distance from school’ is used as an instrumental variable (IV) in the case of the fixed-effect regression model applied here. Though education affects mean differentials of income and consumption positively, the fixed-effect regression coefficients are surprisingly insignificant. However, quantile regression results suggest that education contributes to income and consumption of lower quantile households more than that of uppermost quantile households. Consequently, these indicate a decline in inequality in rural areas of Bangladesh. Interestingly, education has diminishing positive returns for lower quantiles, implying a declined inequality with an increase in education, but at a diminishing rate, confirming that the impact is non-linear in nature.