January 2021

IZA DP No. 14088: Over-Education and Immigrant Earnings: A Penalized Quantile Panel Regression Analysis

Despite evidence that immigrants experience a higher incidence of over-education, relatively few studies have considered the labour market outcomes of over-education for immigrants. Using longitudinal data and penalized quantile panel regression, we inspect the earnings effects of job mismatches for immigrants in Australia. This first application of the method to this question addresses both individual and distributional heterogeneity. Results confirm divergent effects across the earnings distribution. Immigrants from non-English speaking countries experience lower earnings returns and a substantial earnings penalty of up to 25 percent from educational mismatched employment. However, in contrast to conventional findings of a penalty based on regression at the mean, at the highest earnings quantile, both Australian-born and immigrants from non-English speaking countries with a host country qualification receive earnings premiums of 2.7 and 11 percent, respectively, from years of over-education. We discuss the policy implications of this new result.