Individual Heterogeneity in Economic Returns to College in Central Europe

Individual Heterogeneity in Economic Returns to College in Central Europe

In our desire to generalize about the social world, we often employ models that do not take into account individual heterogeneity. We assume the homogeneity of treatment effects across a population, and we ignore the effects of unobserved forces on our observations. Both are major sources of bias, and are very much at work in the mainstream approach to measuring the effect of education on earnings, pioneered by Jacob Mincer in the 1970s. Mincer rates of return are subject to selection bias in terms of the kinds of people who select into college –e.g. due to their family background, ability, and personality traits – that enable them to earn more in the labor market. That is, individual heterogeneity and self-selection give rise to a sorting gain, which biases the estimator for the causal effect of schooling with survey data. To measure the degree of this bias, this seminar utilizes a new approach to heterogeneous treatment effects and applies it to the measurement of earnings returns to college in Austria, Czech Republic, Germany, Poland, and Slovakia based on EU-SILC surveys with data on parental background. The results reveal positive self-selection across countries and for both sexes: the marginal returns to education for those who went to college are substantively larger than the estimated returns for those who did not go to college, as well as for estimates generated from a Mincer model using the same data. This suggests that these educational systems have mechanisms that strongly sort students by their economic prospects on the basis of unobservables. The results offer important lessons to both social policy and the study of inequality.