Program Profile: Comprehensive Anti-Gang Initiative (CAGI)

A crime focused initiative, designed to address gang-related gun homicides in selected cities. The initiative involved a comprehensive model of suppression (enforcement), prevention, and reentry. The program is rated Promising. There was a significant reduction in gun homicides related to gang crime in cities that implemented the initiative.

A Promising rating implies that implementing the program may result in the intended outcome(s).

Program Goals

The Comprehensive Anti-Gang Initiative (CAGI) was implemented in 12 select cities in response to increasing gang prevalence across the country. The main purpose of the CAGI was to prevent and reduce gang-related crime. It grew out of a larger national program called Project Safe Neighborhoods (PSN). CAGI was developed to support communities in preventing and controlling gang crime by funding a comprehensive model of suppression (enforcement), prevention, and reentry. CAGI was coordinated through the U.S. Attorneys' Offices in the selected jurisdictions, and new partnerships focused on gang prevention and controls were developed in order to fully implement the initiative. Partnerships involved state and federal law enforcement, criminal justice agencies, city governments, social service providers, community groups, and schools.

Program Theory

The CAGI was established from the Comprehensive Community Wide Approach to Gang Prevention, Intervention and Suppression developed by Irving Spergel in 1994. A comprehensive approach to gang reduction, the model assumes violence is a product of social disorganization, and gang problems result from the interaction of sociological, demographic, economic and cultural factors along with the social instability and lack of economic support. The Spergel model, also known as the Comprehensive Gang Model, assesses youth needs and provides individualized support services by involving families and local organizations in communities where key organizations are inadequately integrated and have insufficient resources to target gang-involved youth (Spergel, et al. 2003).

Program Components

The CAGI was comprised of specific components, including enforcement strategies and partnerships, prevention and intervention strategies, and reentry/outreach programs. Although the initiative outlined a comprehensive model to gang violence prevention and control for jurisdictions to follow, sites had discretion to make adjustments within the specific components to fit their population and community needs and strengths.  

Enforcement strategies varied by jurisdiction, and included increased federal prosecution, increased state and local prosecution, joint case prosecution screening, and directed police patrols and field interrogations. Prevention and intervention strategies also varied depending on the jurisdiction, and included skills building services, education and outreach, school-based prevention, and substance abuse treatment. Reentry programs focused on outreach and linking services to gang-involved inmates who were returning from prison.

Target Population

The CAGI was developed to target large U.S. cities that had problems with gang activity and gang-related crime. Six cities began receiving funding to implement CAGI in 2006, and by 2008 twelve cities had been funded. The cities were: Cleveland, OH; Dallas/Fort Worth, TX; Los Angeles, CA; Milwaukee, WI; regions of Eastern and Middle Pennsylvania; Tampa, FL; Indianapolis, IN; Oklahoma City, OK; Rochester, NY; Raleigh/Durham, NC; Chicago, IL; and Detroit, MI. The CAGI was implemented in specific cities wait a variation in gang related gun homicides. In 2006, although three of the cities were below the national homicide rate, the majority of the cities involved in CAGI were above the national violence crime rate of 473 per 100,000 populations (McGarrell, et al. 2013).

Study 1

McGarrell and colleagues (2013) examined the effects of the Comprehensive Anti-Gang Initiative (CAGI) using a matched case-control, quasi-experimental design. Data on citywide gun homicides between 2002 and 2009 across U.S. cities with a population of 100,000 or higher was used in the analysis. Data from the Federal Bureau of Investigation (FBI) Uniform Crime Report (UCR) Supplementary Homicide Reports (SHR) were used to measure the primary source for the outcome variable, which was the number of gun-related homicides in the selected cities. City population information was acquired through the 2000 U.S. Census, which also captured structural factors that may have influenced homicide rates.

Cities that received the CAGI were compared with demographically similar control cities that did not receive the CAGI. A group-based trajectory analysis (GBTA) and propensity score matching were used to construct a comparison group of non-CAGI cities with similar population size and similar rates of crime associated with gang violence. The GBTA analysis was used to model 3 years of pre-intervention data to identify and control for preexisting trends relevant to the gun-related homicides. Using a linear model, the researchers analyzed gun homicides in cities with a population of 100,000 or higher were n clusters of 8 to compare model fit statistics. The propensity-score matching model was used to balance structural covariates, which are indicators of higher level of crime, comparing CAGI cities with matched non-CAGI cities. Structural covariate measures included poverty, median home income, unemployment, minority percentage, and homeownership.

To measure levels of implementation, researchers collected data in June 2010 from a survey given to all sites participating in the CAGI. Information collected in a survey from site coordinators was compared with reports of crime from the U.S. Department of Justice, and from site visits and phone interviews.

Binomial regression models were used in the analysis of time series data.. Since the outcome measure is an event count, a Poisson regression binomial model was used to account for time-related lags in implementation factors. Control dummy variables were added to the Poisson regression models to examine whether time variables had an influence on gun-violence event counts.

The National Gang Center (NGC) is the principal training and technical assistance provider for the Gang Reduction Programs. The center provides structured and ad hoc training and technical assistance through meetings, site visits, telephone contacts, materials, and e-mail. Please visit the NGC Web site for more information: https://www.nationalgangcenter.gov/

Subgroup Analysis

In the evaluation examining the effectiveness of the Comprehensive Anti-Gang Initiative (CAGI), McGarrell and colleagues (2013) also measured effects of different components of CAGI dosage on gun homicide rates. Dosage referred to the level of implementation of law enforcement strategies, research integration, and prevention efforts. Law enforcement strategies included increased state/local prosecution of gangs, use of offender notification sessions, comprehensive police partnerships, and an increase in patrol of the CAGI target areas. Prevention programs included outreach strategies to reduce and prevent gun violence and homicide, such as street worker initiatives, skills training, and job placement.

The largest effect in reducing gun homicide rates was the association of law enforcement strategies. The results indicated a significant and immediate effect of law enforcement strategies, resulting in an 8.1 percent decline in homicide rates at immediate consent of CAGI. However, there was no evidence of a sustained impact of these strategies. Furthermore, the majority of the prevention programs implemented at the target sites did not demonstrate a significant impact on gun homicide rates.

These sources were used in the development of the program profile:

These sources were used in the development of the program profile: