A Comparative Study of Violent Extremism and Gangs

Other START Researchers:

Project Details


Speculation on the similarities between violent extremist groups and criminal gangs has grown in recent years.  There are obvious parallels. Both groups involve illegal activities, especially violence, both are dominated mostly by young men, and both are characterized by a more decentralized organizational structure than is commonly assumed.  Given that criminal justice policy makers have designed and implemented gang prevention and amelioration strategies for decades, there is hope that programs developed for gang interventions might have relevance for reducing violent extremism. Despite the promise of this analogy, there has been surprisingly little empirical research comparing the structures and processes of violent extremist groups and gangs. A major impediment to such analyses has been the lack of comparable data. In this project we provide a plan for collecting and analyzing relevant data at both quantitative and qualitative levels. On the quantitative side, we will build on the Profiles of Individual Radicalization in the United States (PIRUS) database and a life-history gang study of the National Longitudinal Survey of Youth 1997 (NLSY97): both of which have been supported by NIJ in the past. On the qualitative side, we add value by again building on our NIJ-funded PIRUS data which provides detailed case studies on three types of political extremists: Islamic, right- and left-wing. To provide comparable case study data we will undertake new data collection to provide life histories of gang members in five major US cities, building on our established contacts. In Phase One of the proposed project, we plan a systematic analysis of these four data bases (two quantitative and two qualitative) in order to dramatically advance the empirical basis for evaluating similarities and differences between violent extremists and gangs. In Phase Two, we will build on this empirical analysis to provide an assessment of the potential for using community-level gang prevention programs to counter violent extremism. In addition to the research benefits of a more complete understanding of these two complex forms of criminal behavior, we plan to develop policy recommendations that will provide guidance on how we can best focus limited resources on countering violent extremism. In this regard, we will take advantage of the professional networks of the three Co-PIs and also START’s extensive transition channels to ensure that the findings from this study will be disseminated to the public, to relevant policymakers and to the academic community.

Primary Findings:

Offender typologies have long drawn the interest of policymakers, practitioners, and researchers. The study of violent domestic extremist groups and criminal street gangs is consistent with this tradition, owing to the threats of crime, disorder and violence posed by these groups. Despite calls for research on the similarities and differences between violent extremist groups and criminal street gangs, there have been few empirical comparisons. We develop a comparative model that emphasizes explicit, spurious, and indirect linkages between the two groups and use national sources of data on domestic extremists and gang members—the Profiles of Individual Radicalization in the United States (PIRUS) and the National Longitudinal Survey of Youth 1997 (NLSY97)—to compare them across group involvement, demographics, family, religion, and socioeconomic status. Six percent of domestic extremists in PIRUS have gang ties, which constitutes a minimal proportion of domestic extremists but likely the rare exception among the population of gang members. Gang extremists more closely resemble non-gang extremists than they do gang members from a representative sample. While these groups share some similarities, one of the major differences is that gang members are younger than domestic extremists, which likely contributes to many of the other differences we observed between the groups across normative life events and states such as marriage, parenthood, unemployment, and education.


We utilize a range of quantitative methods to examine the similarities and differences between extremists in the PIRUS dataset and the gang members in the NLSY97 dataset. At a minimum key comparisons will be made across (1) demographic and socioeconomic factors; (2) pathways into extremism and gangs; (3) parameters of extremist and gang participation (age-graded onset, duration, disengagement); and (4) the prevalence, frequency, and type of criminal behavior. To our knowledge these four stages of analysis will provide policymakers with the most systematic comparisons across large samples of violent extremists and gang members ever available. We propose to conduct bivariate and multivariate analyses across the two data sources to determine the similarities and differences in background, attitudes, and behaviors. This will allow us to determine if the correlates of extremist group participation are similar to that of gang membership. By analyzing these data in a multivariate context, we will isolate the key correlates of both forms of group participation and associated behaviors. A series of linear and generalized linear (logistic, Poisson) regression models will be used in these analyses, as well as growth curve and group-based trajectory modeling which will allow us to track the patterning of these behaviors over time.


Project Period:

More Information