Research

Abstract: We study the consequences of worker displacement during the second industrial revolution using linked census data combined with novel archival data on the 19th-century Swedish iron industry. Ironworkers  displaced by the industrial transformation exited the iron industry at higher rates, were more likely to migrate internally, and ended up in occupations that paid around 10% less than non-displaced workers. Regional migration appears to play an important role in mitigating the impact of worker displacement. To our knowledge, this paper is the first to quantify the effects of worker displacement during the second industrial revolution in Europe.

Abstract: Despite the persistent urban-rural income gaps visible across the developing world, it remains unclear to what extent these reflect the causal effect of urban residency. This paper presents new evidence by estimating the effect of urban residency in childhood on labor market outcomes using an age-at-move design combined with Brazilian census data. The analyses show that spending childhood in an urban environment significantly increases earnings and wages in adulthood, with little effect on hours worked or labor force participation. Around half of the effect is explained by an increase in educational attainment. The findings suggest that the previous literature underestimates the impact of urban residency on earnings by 50%.

Abstract: Boys lag behind girls in educational attainment across the developed world. We first show that most of the gender gap in college enrollment in Sweden can be traced back to differences in compulsory school GPA. We then exploit the random assignment of national test subjects in the natural and social sciences across schools to study how providing standardized information on student performance affects the gender gap. Test assignment reduces the gender gap in teacher-assigned grades by 0.046 SD, a third of the gap conditional on test scores. The effect is similar for male and female teachers and consistent with teachers updating their assessments in response to standardized performance information.

Abstract: Parental separation is a common event during childhood, yet the developments and consequences remain poorly understood. In this paper, we use administrative data on residential co-location to identify all types of parental separation events among children born between 1952 and 2007 in Sweden. While high-SES households are as stable today as in the 1960s, household stability has declined sharply among low-SES households since the 1970s. Restricting attention to  divorce severely understates both the SES gradient and changes over time. Turning to children’s outcomes, we find little evidence that parental separations have a long-run impact on learning outcomes as measured by national test scores. In contrast, parental separations significantly reduce teacher-assigned grades, and consequently, academic track enrollment in high school. The findings suggest that school-based interventions, rather than household interventions, may be key to supporting children following parental separations.

Abstract: Places shape children’s educational attainment, yet the reasons remain poorly understood. In this paper, I study the role of spatial differences in educational signals. Using Dutch administrative data combined with high-stakes national exam scores, I show that students receive less ambitious high school track recommendations from teachers in rural areas. The spatial difference is visible across all demographic groups and the ability distribution, and can explain half of the spatial difference in academic track enrollment. The evidence points to two contributing channels: school composition effects due to higher-SES peers in cities and stronger beliefs in educational investment among urban teachers.

Abstract: We use novel administrative datasets to study human capital dynamics from birth to college in Sweden. We focus on the human capital dynamics among two groups: children from migrant families and children from low-SES families. Children from migrant families start out with a large disadvantage early in life but catch up throughout every stage of childhood, ultimately surpassing natives in college enrollment. In contrast, human capital gaps by SES are comparatively small in early childhood but steadily widen, leaving low-SES children only half as likely to attend college compared to high-SES peers. Higher educational ambition and a more favorable location explain the catch-up among children from migrant families, with no evidence that they over-invest in education. Our findings show that human capital trajectories differ markedly across groups often viewed as disadvantaged and highlights avenues for policy to improve educational attainment among disadvantaged groups.

Abstract: Global concerns over affordability and accessibility of the city have led to rapid expansions of social housing in European countries. While social housing can enable access to the city for a given individual, it risks creating negative externalities via the labor market. In this paper, we study how social housing affects effective labor supply and wages within a spatial general equilibrium model. Using extensive administrative data on the Netherlands, we find that low-skilled workers are more productive in cities, but this is not reflected in earnings because of the labor supply distortions introduced by social housing in spatial equilibrium.

Abstract: We study the development of spatial inequality in educational attainment throughout childhood in Sweden. We track ability development using repeated national tests and combine this with data on applications and enrollment. We first show that most of the spatial heterogeneity exists along the density gradient. Children in denser places are significantly more likely to enroll in college conditional on early childhood ability and SES. However, the dynamics differ. In the bottom of the density distribution, most of the spatial differences are explained by differences in high school track application decisions. In contrast, most of the differences at the top of the distribution are explained by differences in ability development throughout compulsory education. Finally, we show large spatial heterogeneity in educational outcomes within cities. Our paper provides novel insight on how places affect educational attainment and can be used to develop targeted interventions to improve the spatial equality of opportunity.

Abstract: Poverty in early childhood is consistently associated with poorer developmental outcomes, but the extent to which such inequalities are evident within welfare states designed to buffer socioeconomic disparities is less well understood. This study examines whether, and how early, developmental trajectories are socioeconomically stratified in relation to relative poverty in Sweden, a context with universal child health services and near-universal preschool participation. Using a population-based cohort of 9,867 children born in 2017–2018 in Uppsala County, we linked developmental assessments at 18 months, 3 years, and 4 years to national register data on household income, parental education and migration background. Growth Mixture Modeling across cognitive, language, social, and motor domains identified four trajectories: Consistently High, Catch-Up, Late Improvement, and Consistently Low. Trajectories differed in both timing and level of development and were stratified by socioeconomic conditions: children in the Consistently Low group were much more likely to live in relative poverty and to have foreign-born and non-tertiary-educated parents than their Consistently High peers, with disparities already present by 18 months (24% vs 8.8% in relative poverty at age four). An independent preschool readiness assessment at age six showed that children in lower-performing trajectories were substantially more likely to score in the bottom decile, whereas children in the Catch-Up group resembled those in the Consistently High group. Findings indicate that even within a high-income welfare state, early developmental pathways are associated with multiple forms of socioeconomic disadvantage, highlighting the importance of early identification of children at developmental risk.

RCT in the field until 2027 

Firm heterogeneity and exports in the Netherlands: Identifying export potential beyond firm productivity. The Journal of International Trade & Economic Development, 29 (2020), 36-68. (With S. Brakman, H. Garretsen and P. Zwaneveld). Link to paper  

"How large are the non-travel time effects of urban highway tunneling? Evidence from Maastricht, the Netherlands." Transportation Research Part A: Policy and Practice 130 (2019): 570-592. (With J. Tijm, T. Michielsen and P. Zwaneveld). Link to paper