Returns to education in self-employment in India: A comparison across different selection models

This study focuses on estimating the returns to education in non-farm self-employed businesses in the Indian context, using nationwide individual- and household-level data provided by the India Human Development Survey for the year 2011/12. Given that different studies have used different types of regression models for estimating returns to education in self-employment for different economies, the present paper further examines the sensitivity of the estimation of returns to education to the choice of different types of selection model-namely, ordinary least squares, instrumental variable technique, Heckman selection model, and double-selection models, based on the familiar Mincerian earnings equation. The results indicate that, although the trend in the movement of the rate of returns across different educational levels is similar for the different selection models, the estimated value of the rate of returns with respect to the different levels of education is very sensitive to the choice of model. This points to the need to exercise a fair degree of care in the choice of an appropriate model(s) for estimating the rate of returns to education and for drawing policy implications from it.