Contents lists available at ScienceDirect Journal of Criminal Justice journal homepage: www.elsevier.com/locate/jcrimjus Does oending intensify as exposure to violence aggregates? Reconsidering the eects of repeat victimization, types of exposure to violence, and poly- victimization on property crime, violent oending, and substance use Chelsea Farrell, Gregory M. Zimmerman Northeastern University, School of Criminology and Criminal Justice, 409 Churchill Hall, 360 Huntington Avenue, Boston, MA 02115, United States ARTICLE INFO Keywords: Exposure to violence Oending Poly-victimization Victim-oender overlap ABSTRACT Purpose: The relationship between exposure to violence and adverse behavioral outcomes is well-documented. But, heterogeneity in this relationship across dierent operational strategies for exposure to violence is less well understood. This study examines the eects of repeat victimization, exposure to dierent types of violence, and poly-victimization on property crime, violent oending, and substance use. Methods: We analyze two waves of data from the National Longitudinal Study of Adolescent to Adult Health (N = 12,603). We operationalize exposure to violence as: a dichotomous indicator of overall occurrence; ex- posure to multiple incidents of violence (repeat exposure); types of exposure to violence (witnessed, threatened, and experienced violence); and poly-victimization (i.e., repeat exposure to violence and exposure to multiple types of violence). Results: Exposure to violence regardless of how it is measured increases oending risk. The strongest eects are observed for poly-victimization, followed by repeat exposure to violence and exposure to a single episode of violence. There is little variation in eect sizes across types of exposure to violence. Conclusions: The results speak to the utility of preventing the onset of exposure to violence and addressing ongoing exposure to violence in order to interrupt the link between exposure to violence and oending. 1. Introduction Exposure to violence is a particularly consequential aspect of youths' reality. Recent estimates indicate that over 60% of children and ado- lescents under the age of 17 in the United States are exposed to violence each year (Finkelhor, Turner, Ormrod, & Hamby, 2009; Resnick et al., 1997). These youths are at higher risk for subsequent mental health issues (Borowsky, Ireland, & Resnick, 2001; Brown, Cohen, Johnson, & Smailes, 1999; Buka, Stichick, Birdthistle, & Earls, 2001), negative biological responses (Perkins & Graham-Bermann, 2012), and adverse behavioral outcomes (Brezina, Agnew, Cullen, & Wright, 2004; Cleary, 2000; Zimmerman, Farrell, & Posick, 2017). Yet, there is het- erogeneity in the extent and consequences of exposure to violence across studies, in part because researchers have employed dierent conceptual and operational denitions of exposure to violence. Some studies focus on direct exposure to violence, including personal victi- mization and threats of violence (Jennings, Higgins, Tewksbury, Gover, & Piquero, 2010), whereas others note the importance of in- direct experiences with violence, such as witnessing or hearing about violence (Buka et al., 2001; Fagan, Wright, & Pinchevsky, 2014; Richters & Martinez, 1993) to the extent that it substantiates knowledge of the act itself (Zimmerman & Posick, 2016) and leads to a lack of safe haven (Fagan, 2003). There is further variation in the denition of violence,which can be inclusive of the Uniform Crime Report's index crimes (i.e., murder, rape, robbery, assault), as well as less serious, but more prevalent, forms of violence (e.g., ghting) among youth. This study considers both direct (i.e., inter-personal victimization and vio- lent threats) and indirect (i.e., witnessing violence) exposure to vio- lence. Our denition of violence spans ghting and violence with a weapon: respondents answered questions about witnessing a shooting, being threatened with a weapon, and personal victimization. We also consider exposure to multiple incidents of a single type of violence (repeat exposure to violence) and exposure to multiple incidents of multiple types of violence (poly-victimization). By considering dierent measures of exposure to violence, this study seeks to shed additional light on the relationship between exposure to violence and oending behaviors. Much of the prior research on this topic has focused on the widely documented relationship between victimization and oending, a phenomenon commonly referred to as the victim-oender overlap (Jennings, Piquero, & Reingle, 2012). But, http://dx.doi.org/10.1016/j.jcrimjus.2017.09.004 Received 11 August 2017; Received in revised form 12 September 2017; Accepted 13 September 2017 Corresponding author at: Northeastern University, School of Criminology and Criminal Justice, 417 Churchill Hall, 360 Huntington Avenue, Boston, MA 02115, United States. E-mail addresses: farrell.che@husky.neu.edu (C. Farrell), g.zimmerman@neu.edu (G.M. Zimmerman). Journal of Criminal Justice 53 (2017) 25–33 0047-2352/ © 2017 Elsevier Ltd. All rights reserved. MARK