Conference Program

Previous studies show that the characteristics of both families’ residential neighborhoods (e.g., Burdick-Will et al., 2024; Candipan, 2020; Scandurra et al., 2020) and schools’ surrounding areas (e.g., Bell, 2009; Bernelius et al., 2021; Hailey, 2025) influence school choice behavior, but few have analyzed these contexts simultaneously. In this article, we integrate research on neighborhood effects, territorial stigmatization, and school choice to advance the argument that the effect of school surroundings on enrollment decisions operates not only in absolute terms but also relative to families’ place of residence. The neighborhoods in which families live, we suggest, may shape how they perceive and act upon the characteristics of schools’ surrounding areas when making school choices. Importantly, we anticipate that these dynamics may vary by social class, given the salience of class in shaping perceptions of urban space and school choice strategies. Does the influence of schools’ surrounding areas on enrollment decisions depend on families’ residential contexts? Do these relative effects vary across social classes?

To address these questions, we assemble a unique dataset that combines geocoded administrative data from the Chilean School Admissions System with several sources of information at the census block level. We examine how violence and deprivation levels around schools and students’ homes jointly influence secondary school choice. We focus on two outcomes that capture the key stages of the school choice process: the likelihood of applying to a school and the ranking of applications.

Results of our regression models indicate that absolute levels of violence and deprivation around schools reduce the likelihood of applying to a school and the probability that it is ranked higher. Families are also more likely to list and rank higher schools located in comparatively wealthier neighborhoods than their place of residence. The results for relative violence levels are mixed. Deprivation matters more in the construction of application lists, whereas violence has stronger effects in ranking decisions. These effects are more pronounced among high-SES students, and this class-based variation reflects not only structural differences in choice set attributes but also differences in preferences. A set of sensitivity analyses indicates that these findings are robust to the use of many alternative measures of school surroundings, choice sets, and model specifications.

Overall, our findings emphasize the interaction between schools’ geographic surroundings and families’ residential neighborhoods in understanding the link between spatial inequality and school choice. We suggest that the residential neighborhood serves as a baseline context from which parents form judgments about other neighborhoods and the schools located within them, ultimately influencing their enrollment decisions. Beyond school choice, the framework developed in this paper may improve how scholars study contextual effects on individual behavior, by considering the joint influence of the urban contexts in which people and institutions are embedded.