Examining Differences in Violent Crime Trends in PSN and Non-PSN Cities

• In Event: NIJ: Local Level Implementation and National Evaluation of Project Safe Neighborhoods

Abstract

This presentation describes results of crime trend analyses supplemental to those in Presentation 2 that used the federal district as the unit of analysis, reflecting PSN’s broad mandate to each USAO to reduce violent crime throughout its district. From FY18 on, all federal districts have received PSN formula grants and there are no non-PSN districts with which to compare. This presentation describes subset analyses that instead assess crime trends in large jurisdictions (population over 100,000) that are PSN target areas compared with trends in large non-PSN jurisdictions. Beyond supporting this PSN/non-PSN comparison, these analyses offer the benefit of using a design similar to that used in the first national evaluation of PSN (McGarrell, et al, 2009).
Hierarchical latent growth models compare PSN and non-PSN jurisdictions’ violent crime trends, using the same violent crime, population, and contextual data as in Presentation 2. Indicator variables denote whether each jurisdiction was a PSN target area during each timeframe (including when some jurisdictions were PSN target areas under FY2015 through FY2017 PSN competitive grants awarded to some districts). Beginning with the enhanced PSN model in 2018, crime trends in PSN jurisdictions are modeled using implementation data, similar to the models in Presentation 2.

Author

• James Trudeau, RTI International