A Test of the Learning-by-Doing Explanation for the Urban Wage Premium

Abstract

We test if the measured large and stable urban wage premium is related to acquired skill differentials between urban and non-urban workers. We newly document (1) a large urban premium is not fully explained by the differences in the cost of living and (2) urban workers do not experience higher wage growth than non-urban workers. These empirical findings contradict the learning-by-doing explanation for the urban wage premium and they can be fully explained by a gradual ability sorting hypothesis in which more able workers gravitate to urban areas.