The Long-run Consequence from Living In a Poor Neighborhood

I examine the effect of neighborhood quality on long-run labor market outcomes among adults who grew up in substantially different public housing projects in Toronto. Subsidized families were assigned to projects throughout the city at the time they applied, with assignment based mainly on the number of bedrooms required. Unlike housing programs used in previous studies, neighborhood quality differences were not attributable to one set of families moving to better neighborhoods and another set remaining at their current residence. I match census data, longitudinal administrative records, and criminal occurrence data to public housing addresses and track participants’ outcomes, in some cases, a decade or more after leaving the program. The main finding is that differences in neighborhood quality play little or no role in determining adult earnings, education attainment, or social assistance participation, but do affect residents’ exposure to crime. Living in contrasting housing projects cannot explain large variances in labor market outcomes, but family differences, as measured by sibling outcome correlations, account for up to 30 percent of the total variance in the data. Overall, the results suggest that policies aimed at improving long-run outcomes among children from low-income households are more likely to succeed by addressing family, rather than neighborhood, circumstances. (JEL: I30, J38). I am very grateful to my advisors, Alan Auerbach, David Card, John Quigley, and Steve Raphael, for their guidance and support. I also wish to thank Miles Corak and Sophie Lefebre for assistance with Statistics Canada’s administrative data. Frances Beard and Barbara Watson from Metro Toronto Housing Corporation, Brent Donnelly and Ryner Soegtrop from Cityhome, and Monique Volpe from the Ontario Housing Corporation were instrumental in helping me compile a complete list of subsidized housing projects in Toronto. This research was written while I was employed at the Family and Labor Studies Division of Statistics Canada. Additional funding from the Burch Center for Tax and Public Policy and the Fisher Center for Real Estate and Urban Economics is greatly appreciated. I am solely responsible for the contents of this paper.