Association for Public Policy Analysis & Management

Despite the well documented drop in violent crime in American cities, violence is still part of daily life for many children. A growing body of research shows that neighborhood violence negatively impacts academic performance. Most of this evidence, however, is correlational; it is difficult to disentangle the causal effects of crime from neighborhood sorting, or another unobserved cause. In previous work we derived a causal estimate of the acute effect of violent crime on test scores, and found acute effects for children in grades 3 to 5 and no effect on older students. While illuminating, these estimates do not provide insight into the harmful effects of repeated exposure to neighborhood crime, nor do they examine the contextual factors that may explain variation in effects across students with different characteristics. In this paper we answer the following question: Why do some children experience negative effects of exposure to neighborhood violence on educational outcomes, while others seem to be protected from these harmful effects?

The centerpiece of our empirical work is a regression discontinuity (RD) strategy comparing students exposed to violent crime in the week before a standardized test with otherwise similar students exposed in the week after. First, we investigate whether the acute effect of crime depends on prior exposure to neighborhood violence. Second, we explore variation by school context, including student grade level and school climate. 

We use point specific crime data from the New York City Police Department on daily violent crimes – homicide and assault – on every NYC block from 2004 to 2010. These data are linked to student administrative records and survey data from the NYC Department of Education (NYCDOE). The student level dataset contains a wide range of demographic characteristics, measures of academic performance, and the name of the school students attend. The student-level survey data are from an annual school environment survey administered by the NYCDOE. The survey contains responses from middle and high school students about multiple dimensions of school climate, including perceptions of school safety, disorder, and discipline, and feelings of connectedness to school.

Preliminary results show that the acute effect of violent crime is driven by students with prior crime exposures to violence. Further, we find that students in grades 6-8 attending schools with weak climate who are exposed to violent crime in the week before testing score significantly worse on ELA exams compared to children exposed to violent crime in the week after the test attending similar schools. These results suggest that children who are persistently exposed to crime become more sensitive to violence rather than more resilient to it. Both past crime exposure and school climate are important factors in the relationship between crime and test scores, and help to explain why neighborhood crime is more harmful for some children than others.