Spatial dynamics of urban development: School competition and public housing policy

Samantha Snyder, Purdue University

Abstract

While the debate about the optimal degree of policy intervention continues, urban public policy can be generally categorized into either place- or people-based initiatives, each accompanied by their respective sets of strengths and weaknesses. The research contained in this dissertation aims to provide additional insight into the complex ways in which place- and people-based policy interventions and private market decisions interact to shape outcomes in an urban environment. Using the context of Chicago, Illinois, this dissertation specifically examines the impacts of two broadly-implemented public policies aimed at improving outcomes for urban residents. In the first empirical chapter, we examine the effects of school choice policy on student outcomes. Touted as a quasi-free market solution to poor public school performance, school choice programs and the resulting competition fostered between schools should result in higher performance overall. We test this hypothesis using a smooth transition autoregressive (STAR) model that employs a school-level measure of competition to allow the competitive environment to alter the education production function. We find that competition does not improve student outcomes and in some cases likely lowers performance for schools serving disadvantaged populations. Accounting for the disproportionate use of the choice program by households of higher socioeconomic status as compared with their disadvantaged counterparts, we suggest that privileged households may locate in areas with substandard schools and satisfy preferences for good schools by using the choice program. If this is true, then areas of the city experiencing gentrification would likely not benefit from an increase in school performance to match the increase in resident socioeconomic profile. Our results show that in the event that gentrifying households disproportionately use school choice, schools in competitive environments suffer the most. The second empirical chapter examines the external effects of public housing on neighborhood property values. One might expect that public housing developments exert a negative impact on surrounding home prices. However, it is important to consider the non-random location of public housing within a city as failure to account for non-random treatment assignment can potentially bias estimates of any treatment effect. We use propensity score matching methods to provide a more balanced sample after which we estimate a spatial hedonic home price model. The characteristics of public housing developments vary across the three types commonly used in Chicago, and our results reveal that the impact of public housing strongly depends on the type of development. Home price impacts also vary with distance from the public housing development. Mixed income developments tend to have negative effects on home prices that extend as far as one mile from the site. Scattered site housing tends to increase nearby property values, and especially at close distances. While the results of this analysis indicate potential spillovers from place-based policies, the methods point to refinements in the matching schemes used in spatial econometric modeling. In sum, the research presented in this dissertation provides further evidence that policy interventions and private household decisions frequently interact and often result in unexpected outcomes. In the case of school choice, policymakers have clearly adopted a people-based strategy that simultaneously spotlights the role of household decisions in impacting student outcomes. However, the broad consequences appear to run counter to the stated policy objectives and in fact point to deleterious effects on school performance. In contrast, public housing policy is distinctly place-based and clearly aims to serve the needs of low income residents. The results presented here, however, highlight the significant spillovers, both negative and positive, that accrue to households not directly targeted by the policy.

Degree

Ph.D.

Advisors

Florax, Purdue University.

Subject Area

Education Policy|Economics|Agricultural economics|Urban planning

COinS