A common conclusion of nearly two decades of research on radicalization is that challenges stemming from the heterogeneity of extremists, low base rates of offending, and the seemingly prosaic nature of radicalization correlates make it difficult to generate reliable risk factors for violent extremism. At the same time, the need for insights to help identify individuals at risk of radicalizing, as well as to make evidence-based decisions about the rehabilitation and reintegration of extremist offenders, has never been greater. The number of individuals adhering to hate-based and extremist ideologies in the United States has grown considerably in the last decade, while law enforcement agencies and those who are responsible for administering community-based prevention programs continue to struggle with limited resources.
The purpose of this study is to improve the validity, reliability, and utility of risk analyses of violent extremism by creating a relational database that combines individual-level radicalization risk factors with variables at the meso and macro-levels of analysis. In 2012, with support from the National Institute of Justice (Grant Award #2012-ZA-BX-0005), our research team at START began work on a database of political extremists called Profiles of Individual Radicalization in the United States (PIRUS). The PIRUS dataset contains individual-level information on 2,225 violent and nonviolent extremists from across the ideological spectrum who committed crimes in the United States from 1948 to 2018. The database includes 147 variables that cover a wide range of micro-level attributes relevant to an individual’s radicalization process, such as their personal history, basic demographic information, group membership, interaction with online extremist content, mobilization mechanisms, and prior criminal history.
This project expanded PIRUS into a suite of relational datasets that include individual level, network, and community data. We began by mapping the co-offender networks present in PIRUS and we used them to construct a new dataset called the Social Networks of American Radicals (SoNAR). SoNAR allows users to model the co-offending networks in which the subjects in PIRUS were embedded at the times that they committed their crimes. Users can then perform social network analysis (SNA) to understand the dynamics of extremist co-offending relationships in the United States and they can integrate network-level variables alongside individual-level characteristics in their studies of radicalization.
Publication Information
Jensen, Michael, and Gary LaFree. 2022. "A Multi-level Approach to the Study of Violent Extremism." START: College Park, MD.