Job Market Paper (link to pdf)

Internal migration and the spread of long-term impacts of historical immigration in Brazil (with Daniel Lopes and Leonardo Monasterio)

Abstract: Immigration shocks in the 1850–1960 period left long-lasting positive impacts in southern Brazil. Yet, little is known about how these benefits spread over the country in the following decades. We use a surname-based classification of ancestries to identify descendants of immigrants and investigate the spreading of gains from historical immigration in Brazil. We find that the concentration of descendants of historical immigrants in municipalities in northern and central Brazil is positively associated with several indicators of economic development today, in particular with higher wages. Leveraging individual-level information from linked employer-employee data in which we observe both the individual's wage and ancestry, we find a wage premium of approximately 12% for descendants and positive spillovers between ancestry groups. One additional percentage point in the concentration of descendants in a municipality corresponds to a wage increase of 1% for descendants and 2% for non-descendants. Our results agree with a model where descendants and non-descendants have complementary skills in the production function of the firms, particularly those in the agricultural sector.

Work in Progress

Selection and heterogeneity in the returns to migration (with Marieke Kleemans and Emilia Tjernström)

Abstract: There is considerable debate on the returns to rural-urban migration in developing countries and magnitudes differ sharply depending on the method used. We aim to reconcile these divergent estimates by explicitly accounting for the role of heterogeneity in the returns to migration and sectoral mobility. We exploit rich panel data from over 36,000 individuals in Indonesia whom we observe over a 12-year period. We model self-selection into migration in a multi-period Roy model, which incorporates worker heterogeneity in both absolute and comparative advantage. We then estimate a correlated random coefficient model that considers both types of heterogeneity. This model lets us extrapolate the returns identified from switcher sub-populations to non-switchers­—a group of particular interest to policymakers deciding whether to encourage migration as a development strategy. Finally, we draw on recent developments in the literature on non-parametric panel data identification and employ a group-random coefficient estimator that explicitly tests the parametric assumptions that identify the returns to non-switchers. Our results show little heterogeneity and small returns in earnings to rural-urban migration but larger and more heterogenous returns to switching from agricultural to non-agricultural sectors. Our estimated returns to non-switchers are in the range of 0.45–0.68 log points: considerably higher than the returns to switchers identified using individual fixed effects (0.25 log points) and close to the raw gap between the two sectors (0.66 log points).

The wage premium of migrants and return migrants: Internal migration in Brazil

Abstract: This paper investigates the components and the recent evolution of the wage premium of current and return migrants within Brazil. Using cross-sectional data from repeated household surveys in 2004–2014, I find that the wages of internal migrants are about 12% higher than the wages of non-migrants. For return migrants, wages are 9% higher on average. I also find that the wage premium for migrants decreased during this period, while the wage premium for return migrants increased. Using longitudinal data from linked employer-employee datasets in 2005–2015, I find a 5–10% wage premium for both migrants and return migrants in panel regressions with individual fixed effects for a subsample of formal sector workers. Restricting the sample to those who move at some point in the panel, I find no wage premium associated with the current migrant status and a 4% wage penalty associated with returning. I explore different regression specifications, subsamples, and an instrumental variables strategy based on past migration rates to discuss the role of self-selection, place-specific effects, and learning on these wage premia. My results suggest that the self-selection of internal migrants in Brazil is based more on absolute advantage (migrants earn more in any location) than comparative advantage (migrants earn more in a specific location). My results are also consistent with learning impacting post-migration earnings regardless of a migrant's location.

Surnames and social mobility in Brazil (with Daniel Lopes and Leonardo Monasterio)