Did cities increase skills during industrialization? : evidence from rural-uban migration

The process of industrialization is typically associated with urbanization and a widening urban-rural skills gap. To what extent were these disparities driven by the direct impact on occupational attainment of living in an urban area or the result of the positive selfselection of more-skilled individuals into cities? In this paper, we leverage exceptional Swedish longitudinal data that allow us to estimate the impact of rural-urban migration on skill attainment during Sweden's industrialization from the 1880s to the 1930s using a staggered treatment difference-in-difference estimator. We attribute roughly half of the gap in urban-rural skills to a direct impact of living in an urban area, whereas the other half is driven by self-selection into cities. A third of the direct impact of residing in cities is explained by a static effect, reflecting better initial matching, while the rest is the result of a dynamic effect as individuals upgrade their skills over time in urban areas. We conclude that cities had a substantial effect on skill development in Sweden around the turn of the nineteenth century that is likely to extend to other European and North American economies that were industrializing around the same time.