Evidence on the Production of Cognitive Achievement from Moving to Opportunity
This paper performs a subgroup analysis on the effect of receiving a Moving to Opportunity (MTO) housing voucher on test scores. I find evidence of heterogeneity by number of children in the household in Boston, gender in Chicago, and race/ethnicity in Los Angeles. To study the mechanisms driving voucher effect heterogeneity, I develop a generalized Rubin Causal Model and propose an estimator to identify transition-specific Local Average Treatment Effects (LATEs) of school and neighborhood quality. Although I cannot identify such LATEs with the MTO data, the analysis demonstrates that membership in a specific demographic group is more predictive of voucher effects than is the group’s average change in school or neighborhood quality. I discuss some possible explanations.
JEL Codes: C31, C36, C50, D04, I20, I38, R23.
Keywords: Two-Dimensional Treatment, Rubin Causal Model, School Effect, Neighborhood Effect, Local Average Treatment Effect, Education Production Function (EPF), Moving to Opportunity.
Suggested citation: Aliprantis, Dionissi, 2017. “Evidence on the Production of Cognitive Achievement from Moving to Opportunity,” Federal Reserve Bank of Cleveland, Working Paper no. 17-07. https://doi.org/10.26509/frbc-wp-201707.