Examining the causal effects of exposure to violence on crime among youth involved in the justice system: Experienced, witnessed, and experienced-witnessed violence.

Abstract

Previous studies on exposure to violence lack a nuanced understanding of the causal effects of different exposure types on offending behaviors. This study, drawing on Pathways to Desistance Study (PDS) data tracking 1354 adjudicated youths aged 14-18 over 7 years, explores the contemporaneous (cross-sectional), acute (after 1 year), enduring (after 3 years), and long-term (after 6 years) causal effects of violence exposure on property and violent offending. The sample, predominantly male (86%), consisted of White (20%), Black (42%), and other (38%) individuals. The generalized propensity score is used to match unbalanced covariates across multiple exposure types, namely noninvolved (n = 392), witnessed (n = 577), experienced (n = 31), and experienced-witnessed violence (n = 305). Results demonstrate the contemporaneous, acute, enduring, and long-term effects of violence exposure on both violent and property offending, with varying durations and strengths across exposure types. The most pronounced risk effects are immediate, diminishing over time and potentially reversing in the long term as youth transition into adulthood. Among exposure types, experienced-witnessed violence exhibits the most potent effects on offending, followed by witnessed violence and then experienced violence-a pattern consistent across the observed time points. Noteworthy is the finding that the impact of violence exposure is more pronounced for violent offending, diminishing more rapidly compared to the effects on property offending.