with Clement Imbert.

This paper studies the economic effects of rural-urban migration on Brazilian cities. Using a shift-share IV design, we show that over the span of a decade immigration lowers wages, increases the number of formal firms and jobs, reduces the share of informal workers, and has no effect on unemployment. These findings contradict the traditional view that rural-urban migration contributes to urban informality and unemployment. To explain these surprising results, we develop and estimate a model of firm dynamics and informality. If formal wages are rigid, informality increases in the short run following a migration shock, but declines as wages adjust downwards. In the long run, the model quantitatively replicates the IV results. The informal sector serves as a “stepping-stone” for around 40% of formal firms in the new steady state. However, the overall economic benefits of immigration are larger in a counterfactual scenario with higher enforcement and no informality.