Association for Public Policy Analysis & Management

Recent work by Reardon et al. (2016a, 2016b) has documented substantial variation in achievement levels and in racial/ethnic achievement gaps across school districts and metropolitan areas in the United States. Whereas the mean achievement in school districts such as Brookline, MA and Cupertino, CA is the equivalent to three grades larger than the national mean, academic performance in schools in Camden, NJ and Detroit, MI is more than two grades behind the country average. Similarly, racial/ethnic achievement gaps range from virtually zero in places like Hillside, NJ to more than one standard deviation in Cleveland, OH and Atlanta, GA. Although racial/ethnic differences in parental income, parental education, and racial/ethnic segregation emerge as strong predictors of these achievement disparities across districts, other drivers of inequality remain unexplored. This study will investigate the effect of violent and property crime on mean achievement and on racial/ethnic achievement gaps in math and English Language Arts (ELA) across school districts and metropolitan areas in the United States.

The study of crime as a potential mechanism driving inequality in educational outcomes stems from a large body of research that has shown that exposure to neighborhood violence affects children’s cognitive development and academic performance. Evidence on non-cognitive outcomes has shown that children who grow up in violent neighborhoods are less engaged in the classroom, feel less safe in the school, and are exposed to conflicting cultural frames that influence their educational trajectories. Beyond these individual-level impacts, an extensive literature in criminology and urban sociology has documented the negative effects that crime has at the community level. This body of evidence indicates that neighborhood crime is a salient attribute of children’s environment and has the potential to generate inequalities in academic performance across school districts and metropolitan areas.

This study uses data from the Stanford Education Data Archive (SEDA) on mean achievement and on racial/ethnic achievement gaps from virtually all school districts and metropolitan areas in the United States. These data are merged with agency-level data on crime reports from the Uniform Crime Reporting (UCR) program. To move beyond correlational analyses, this study exploits temporal variation in crime rates and educational outcomes across years and grades within school districts and metropolitan areas, and it employs an instrumental variable strategy that exploits the timing and amount of funds that police departments received through the Community Oriented Policing Service (COPS) grants program. Using longitudinal data from years 2008 to 2012, the 2SLS estimates suggest that a one standard deviation increase in the violent crime rate in the metropolitan area leads to a decline in ELA mean achievement by .41 standard deviations. Similarly, a one standard deviation increase in the property crime rate in the metropolitan area leads to a decline in ELA mean achievement by .27 standard deviations. Effects on math are smaller and non-statistically significant.