Educational and gender heterogeneity of the rural-urban earnings premium: New evidence from Norway

Abstract

We explore urban earnings premiums for young, native, rural-to-urban movers in Norway. Using an augmented difference-in-differences estimator (DiD-TR) on microdata we challenge previous claims about urban earnings premium's size and sources. Conventional econometric estimators understate the static premium and overstate dynamic premiums. We find that migrants exhibit lower mean but faster pre-move earnings growth than non-migrants. Post-move, the static earnings premium dominates. The observed trajectory is related to frequent pre-move changes of industrial sector, presumably to obtain better job-worker matches. Post-move, these changes occur less frequently. Highly educated females exhibit largest static premiums (34%), less-educated females least (24%), males an intermediate amount. Our findings suggest that cities primarily generate earnings premiums through agglomeration-based efficiencies and superior job-worker matches varying heterogeneously by education and gender.