On the Other Side of the Tracks: Neighborhood Effects and Segregation in Morocco

Abstract

This paper examines the impact of neighborhood effects using a unique data set from rural Morocco and finds strong evidence that poor households do shift their actions, specifically in the domains of social group membership and investment in education in the next generation, due to living near wealthier households. The estimation of peer effects focuses on comparing the outcomes of two groups shown to be similar across covariates that vary in the distance between their house and that of a third group, specifically those living in the homes of the former European colonists of the town. This identification strategy is rooted in the history of Khouribga, as the segregation of the European and Moroccan residents of Khouribga during the colonial period on separate sides of the railroad track running through town has resulted in the concentration of Khouribga’s wealthiest residents today inside the area where the European colonists used to live. The results suggest that the neighborhood effects contribute to increased participation by adults in work-related professional groups and higher levels of investment in human capital in the next generation of sons. Further support for the presence of neighborhood effects derives from evidence that households where the father participates in a professional group invest even more in the education of their sons, and residents of the former European households are also much more likely to belong to a professional group, facilitating a higher level of interaction between the two groups. Robustness checks focus on two distinct areas. First, alternative hypotheses that may explain this result, such as access to local public goods and other explanations for the limited educational investment by the North are shown to lack support in the data. Second, propensity score matching and utilizing distance to nearest European household as the treatment variable (and dropping observations extremely close to the European households) mitigate concerns about selection bias into the households near the European households driving the results. JEL (I25, N37, R23, 015)