Residential segregation and the geography of opportunites: a spatial analysis of heterogeneity and spillovers in education

This research estimates the consequences of socioeconomic residential segregation on educational outcomes in the context of the Chilean voucher system used for education. It is found that the combination of school and socioeconomic residential segregation creates challenges to social mobility and social inclusion of the most vulnerable population. Poverty concentration is understood as the clustering dimension of socioeconomic residential segregation. Its effects are measured by combining quantitative and qualitative methods. Quantitative methods measure the magnitude of two spatial processes: spatial heterogeneity--the contextual differences between neighborhoods--and spatial dependence--by which educational outcomes of one neighborhood depend upon those of adjacent neighborhoods. Spatial processes are tested with multilevel and spatial models implemented in a two step procedure that approximates a hierarchical spatial model. This methodological innovation creates the opportunity for new analytical understanding of the mechanisms driving these spatial processes. A collective case study method of educational communities in three segregated neighborhoods is applied in order to understand the mechanisms driving these spatial processes. More than 16% of the variation in 4th graders' math test scores in Santiago is found to be explained by the characteristics of the neighborhood where the school is located. The effects of concentrated poverty are perceived through the actions of certain social mediators. Whether a student lives with both parents and the strength of the family-school bonds are particularly key factors in predicting educational outcomes in poor and segregated areas. Families in these neighborhoods lack exposure to the middle and upper classes' attitude toward education, which is dominant in formal school settings; thus, a strong school-family bond is a way of bridging this difference in attitude. On the other hand, in poor and segregated areas, teacher job satisfaction is negatively associated with test scores. Some schools adjust their expectations downward about their students' potential outcomes; furthermore, some teachers see themselves as successful social workers but with diminished expectations of students' educational outcomes, which explains this negative correlation. Concentrated poverty affects educational outcomes, but this effect is not deterministic. In fact, some families show successful coping strategies, while others do not. Although further research is needed to explain these differences, this research suggests that the school plays an important role in counterbalancing the negative effects of socioeconomic residential segregation on educational outcomes. Thus, besides neighborhood and school socioeconomic integration, policies aimed at strengthening the mediating role of the school are relevant ways of preventing the negative effects of spatial concentration of poverty on educational outcomes.