Research
Research
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 depending on the empirical methods used. We aim to reconcile these divergent estimates by explicitly accounting for the role of heterogeneity in the returns to migration. We develop a correlated random coefficient model that allows for location-specific skills and heterogeneous returns, estimated using rich longitudinal data from Indonesia, China, and Tanzania. 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. Our results reveal considerable heterogeneity in the returns to migration and show a clear pattern in the relationship between absolute and comparative advantage across countries: those with the lowest productivity in rural areas stand the most to gain from migrating. This suggests that migration is a pro-poor strategy but that barriers to migration may prevent workers from realizing their potential. As such, individuals appear to be inefficiently sorted across space; therefore, encouraging migration could lead to large returns.
The wage premium of migrants and return migrants: Internal migration in Brazil (with James Hamlette)
Abstract: We study the wage premium of internal migrants in Brazil, exploring different migrant definitions in large cross-sectional and longitudinal datasets. By differentiating between current and return migration both across and within states, we shed light on the composition of the 9-10% wage premium of interstate migrants. Our findings are consistent with a model where migrants (i) are positively self-selected based on ability, (ii) have a person-specific matching with their locations, and (iii) face migration costs increasing with distance. Panel regressions with individual fixed effects suggest that around 5/6 of the wage premium of current migrants is due to unobserved ability and 1/6 to the person-place matching. Additional results show that internal migrants are more likely to have positive labor income and work longer hours, and the difference in the premia of current and return migrants—the returning penalty—shrunk to near zero between 2000 and 2010.
The impact of teacher strikes on high school students in the United States (with Dustin White, Jamie Wagner, and Megan Harris)
Abstract: Teacher strikes reduce the number of school days and interrupt instruction, yet there is scarce evidence on how teacher strikes impact observable measures of students’ performance. Using data from the High School Longitudinal Study of 2009 and information gathered about teacher strikes from the 2009-10 to 2012-13 school years, we find that each day of school missed due to teacher strikes is associated with a 0.015-point decline in GPA for affected students. The typical strike corresponds to a 0.158-point GPA drop for affected students. Low-performing students are also significantly less likely to graduate after a strike. We leverage information on parent and student expectations of their educational attainment to show that those expectations are unaffected by strikes, suggesting that omitted variables are unlikely to cause the observed changes in GPA and graduation rates.
Internal migration and the spread of long-term impacts of historical immigration in Brazil (with Philipp Ehrl, Daniel Lopes, and Leonardo Monasterio)
Abstract: This paper investigates how the inter-regional spread of immigrants' descendants from the 1850-1960 period affects current labor market outcomes in Brazil. We use linked employer-employee data, apply a surname-based classification to identify workers with non-Iberian ancestry (i.e. descendants) and estimate wage regressions with individual fixed effects and an instrumental variable. The data reveal pronounced differences between municipalities and ancestry groups. We find that the concentration of descendants in interior municipalities of northern and center-western Brazil is positively associated with higher wages. These wage gains, however, accrue only to the predominant population group with Iberian surnames. The same pattern, but with much less pronounced wage gains is observed in municipalities with high descendant rates, i.e. those close to immigrant settlements in the past. Our results are in accord with a simple model in which descendants and non-descendants are imperfect substitutes in the production function.
Intergenerational mobility and regional inequality in Brazil (with Daniel Lopes and Leonardo Monasterio)
Abstract: We adapt the methodology developed by Güell et al. (2015) and reproduce in Brazil the study of intergenerational mobility (IM) and regional inequality done by Güell et al. (2018) in Italy. Our study leverages unique features of the Brazilian context, such as the availability of schooling and income outcomes in datasets informing surnames, detailed measures of ancestry and internal migration, and wide regional variation. Thus, we circumvent data limitations in the two original studies conducted in two developed European countries and provide an overview of the relationship between IM and inequality in a middle-income Latin American country. We also make a novel contribution by investigating how IM correlates with historical factors and contemporaneous indicators of socioeconomic development in Brazil. Our results show that past land inequality and slavery are associated with lower long-term mobility, which in turn is associated with indicators of poor economic performance, such as lower income per capita and educational attainment. Unlike other studies of IM in Brazil, we find no clear geographic pattern, and we find a positive (though weak) correlation between social mobility and current inequality.
Parental leave and career gender gaps (with Ana Paula Melo)
Short description: This paper investigates the impact of parental leave on the career progression of working parents, their coworkers, and employers, allowing for differential impacts for male and female coworkers in an event-study design using matched employer-employee panel data from Brazil.