Selection in initial and return migration: Evidence from moves across Spanish cities

This paper investigates the contribution of migration to the sorting of more productive workers into big cities using administrative data for Spain that follow individuals over their work lives. While migrants to small cities do not exhibit selection of any type, migrants to big cities are positively selected in terms of education, occupational skills, and individual productivity as proxied by their pre-migration position in the local earnings distribution. However, not everyone benefits equally from big cities and this leads to a second round of sorting. Returnees are not only ex-ante less productive than permanent migrants, but are also those who, following the first move, have least boosted up their earnings in big cities. Low realized earnings and unemployment affect return decisions of workers who moved to big cities at younger ages in particular, suggesting that older migrants may face less uncertainty upon moving to big cities.