Research on political extremism in the United States has thus far been unsuccessful in establishing an understanding of the risks and vulnerabilities associated with violent behavior. The challenges associated with analyzing the causes of violent extremism have led many researchers to forgo risk assessments in favor of thick descriptions. However, it may be premature to dismiss the possibility of developing useful risk assessment tools. Indeed, attempts to establish extremism risk factors have been unsuccessful in large part because of small, unrepresentative samples, a lack of non-violent reference groups, and an over-reliance on individual-level variables. In an attempt to improve the accuracy and utility of risk assessment tools, this project proposes an innovative, multi-level exploration of a large sample of US political extremists. By adopting a multi-level framework that considers how micro-level characteristics interact in sequence with meso- and macro-level variables to facilitate or counteract extremist violence, it may be possible to substantially advance a risk-based understanding of violent extremism in the United States. First, the project team will build on the NIJ-funded Profiles of Individual Radicalization in the United States (PIRUS) dataset, which currently contains information on over 1,800 U.S. extremists, by adding approximately 1,000 new cases to the database. Second, the team will apply a social network analysis (SNA) approach to PIRUS by plotting and coding the bi-nodal relationships between individuals within the data, as well as individuals outside of the data who played critical roles in the facilitation of violent acts. The research team will also analyze the structure of discovered network clusters for attributes including organizational hierarchy, network cohesion, and recruitment flows. And finally, the team will overlay the PIRUS data with macro-level data at the city or county-level using US Census data from 1940-2010, allowing researchers to assess the causal role of community-level variables in the mobilization of violence. This will include measures of relative socioeconomic disadvantage, ethnic heterogeneity, crime rates, and residential mobility. The overarching goal of this project is to determine how the components of risk are nested in multi-level structures and to show how those variables work in combination to drive individual-level extremist outcomes. In addition to submitting required reports, participating in academic conferences and practitioner workshops, and authoring peer-reviewed journal articles, the research team will leverage START's training development capabilities to create an online training series aimed at educating law enforcement and CVE practitioners on the risks and vulnerabilities of violent extremism.
To investigate and integrate the range of factors that exist at the micro-, meso-, and macro-levels into a comprehensive risk assessment of violent extremism, the team will utilize cross-classified multivariate statistical modeling, hierarchical linear modeling (HLM), and social network analysis (SNA).