Opportunity and Risk : Learning about Intergenerational Mobility from Conditional Variance

The standard measures of intergenerational mobility in economic literature do not account for risk. The estimates of relative and absolute mobility are usually based on the conditional mean function for children's life outcomes given parent's economic status. The existing literature largely ignores the information captured in conditional variance. We develop a broader approach where conditional variance of children's outcome is a summary statistic for life-time risks. Using data from China, India, and Indonesia, we estimate simultaneously conditional mean and conditional variance functions for children's schooling by a maximum likelihood estimator. Evidence suggests a strong negative association between father's schooling and conditional variance of children's schooling across all three countries. Being born to an educated father confers double advantages: a higher mean and a lower variance. We derive risk-adjusted measures of relative and absolute mobility by accounting for an estimate of the risk premium for the conditional variance faced by children at birth. The estimates of risk-adjusted measures suggest that the standard measure of relative educational mobility, intergenerational regression coefficient, substantially overstates relative mobility of the disadvantaged subgroups: children of low educated parents, daughters, and rural children. For the most disadvantaged children (fathers with no schooling), the overestimation is 35.63% in China, 46.58% in India, and 20.94% in Indonesia. In contrast, the overestimation is negligible for the children of college educated fathers. The methodology developed here is of wider interest as the risk-adjusted measures can be readily applied to intergenerational income mobility analysis