After watching the video and completing the readings for this week, conduct further research into ways in which trauma is handled in one of the three contexts for this course (mitigation/aggravation,

Received 17 August 2011 Received in revised form 3 September 2011 Accepted 7 September 2011 Available online 16 September 2011 Keywords:

Offending Victimization Overlap Violence Theoretical and empirical research investigating victimization and offending has largely been either‘victim- focused’or‘offender-focused.’This approach ignores the potential theoretical and empirical overlap that may exist among victims and offenders, otherwise referred to as‘victim–offenders.’This paper provides a compre- hensive review of the research that has examined the relationship between victimization and offending. The review identified 37 studies, spanning overfive decades (1958–2011), that have assessed the victim–offend- er overlap. The empirical evidence gleaned from these studies with regard to the victim–offender overlap is robust as 31 studies found considerable support for the overlap and six additional studies found mixed/lim- ited support. The evidence is also remarkably consistent across a diversity of analytical and statistical tech- niques and across historical, contemporary, cross-cultural, and international assessments of the victim– offender overlap. In addition, this overlap is identifiable among dating/intimate partners and mental health populations. Conclusions and directions for future research are also discussed.

© 2011 Elsevier Ltd. All rights reserved.

Contents 1. Introduction............................................................... 16 1.1. Theoretical perspectives of the victim–offender overlap....................................... 17 2. Methodological approaches to identifying the victim–offender overlap.................................... 19 3. Normative assessments of the victim–offender overlap........................................... 22 3.1. Historical documentation of the overlap............................................... 22 3.2. Contemporary documentation of the victim–offender overlap..................................... 22 3.3. The overlap in dating violence................................................... 23 3.4. Cross-cultural investigations of the victim–offender overlap..................................... 23 3.5. International explorations of the victim–offender overlap....................................... 23 3.6. Intersections of the victim–offender overlap and mental health.................................... 24 4. Conclusions and directions for future research............................................... 24 4.1. The need for more and potentially better theory........................................... 24 4.2. The need for continued application and development of advanced methods . ............................. 24 4.3. The need for expanded consideration of offense types........................................ 24 4.4. The need for studies to collect victimization and offending data.................................... 24 References.................................................................. 25 1. Introduction Of all criminological facts, to include the strong, patterned rela- tionships between crime, age, sex, race, socioeconomic status, neigh- borhood disadvantage, and individual differences, perhaps none is also as consistent but less recognized as the link between offenders (offending) and victims (victimization) (Gottfredson, 1981; Maxfield, Aggression and Violent Behavior 17 (2012) 16–26 ⁎Corresponding author at: University of Texas at Dallas, Program in Criminology, EPPS, 800 W. Campbell Road, GR31, Richardson, TX 75080, United States.

E-mail address:[email protected](A.R. Piquero).

1PhD, Professor, Program in Criminology, EPPS, University of Texas at Dallas, Adjunct Professor Key Centre for Ethics, Law, Justice, and Governance, Griffith University Australia, Co-Editor,Journal of Quantitative Criminology. 1359-1789/$–see front matter © 2011 Elsevier Ltd. All rights reserved.

doi:10.1016/j.avb.2011.09.003 Contents lists available atSciVerse ScienceDirect Aggression and Violent Behavior 1987). Yet, for most of its tenure, thefield of criminology has tended to concentrate on the patterning and predictors of offending (of- fenders only)orthe patterning and predictors of victimization (vic- tims only), without due recognition and theoretical and empirical scrutiny afforded to the overlap between victims and offenders. This specific population of overlapping individuals, shown inFig. 1, oc- cupies an important area of research and is the focus of the current review.

This article is designed to provide readers with a broad overview of the relationship between victimization and offending. In so doing, we begin with a general presentation of the theoretical perspectives most often applied to the victim–offender overlap, and then follow this introductory material with a review of the methodological ap- proaches that have been used to study the overlap. The article then provides both an historical and current documentation of the overlap and key researchfindings to include a discussion of some under- studied topics in the area such as dating violence, mental health populations, as well as cross-cultural and international studies. We close with a summary of keyfindings and a list of important direc- tions for future research.

1.1. Theoretical perspectives of the victim–offender overlap Several sociological and criminological theories attempt to explain the overlap between violent offending and violent victimization. Rou- tine activities theory is by far the most recognizable of these theoretical perspectives (Schreck, Stewart, & Osgood, 2008; Smith & Ecob, 2007; Taylor, Freng, Esbensen, & Peterson, 2008). The theory explains victimi- zation and offending at the confluence of a suitable target, motivated of- fender, and the absence of a capable guardian. As it relates to the victim– offender overlap, the theory centers on the influence that opportunity structures and risky lifestyles have on increasing the likelihood for com- mitting an offense or experiencing victimization.

Extending from this routine activities perspective,Osgood, Wil- son, O'Malley, Bachman, and Johnston (1996)developed a theoretical framework referred to as“unstructured socializing”that has direct relevance for explaining the victim–offender overlap. Specifically, Osgood et al. argue that it is not necessarily the time, in general, that a youth spends with delinquent peers that is conducive for offending but instead the amount of time that a youth spends with delinquent peersin the absenceof adult supervision that produces a situation ripe for offending (Osgood et al., 1996).Schreck, Fisher, and Miller (2004)have since reported that delinquent peers often do not provide the best source of protection from victimization. As such, the greater amount of time a youth spends with delinquentpeers in the absence of adult supervision the greater their likelihood of participating in violent offending and experiencing violent victimization.

Delinquent peers can function as agents that provide an individual with tangible and intangible rewards for delinquent/criminal behavior.

For instance, these peers may be able to afford the youth with an increase in social status or reputation for their involvement in delinquent/ criminal activity. These peers can also serve as a readily available source to draw‘accomplices’or co-offenders from that make the performance of delinquency/criminal activity easier. This delinquent/criminal involve- ment places an individual at an elevated risk for experiencing victimiza- tion as well (Felson, 1986; Jensen & Brownfield, 1986; Mustaine & Tewksbury, 2000; Osgood et al., 1996; Sampson & Lauritsen, 1990; Schreck et al., 2008; Smith & Ecob, 2007; Taylor et al., 2008). In short, the routine activities framework provides a view of the characteristics of the situation that may increase the likelihood of victimization and offending.

Beyond the routine activities perspective, other notable sociological and criminological theories to explain the victim–offender overlap in- cludeGottfredson and Hirschi's (1990)general theory of crime with its focus on the individual characteristic of self-control, as well as the more general set of subcultural theories (Anderson, 1999; Stewart, Schreck, & Simons, 2006). With regard to low self-control,Gottfredson and Hirschi (1990)argue that a lack of socialization leads to delinquent/criminal activity. Specifically, if a child's parents do not effectively monitor their child's behavior, recognize deviant behavior when it occurs, and punish such behavior in responseto its occurrence then this failed socialization process produces an inadequately socialized youth.

Inadequate socialization is viewed as the source of development for low self-control and the subsequent involvement in delinquent/criminal activity over the life course. The link between low self-control and offending has been well established in the literature (Gottfredson, 2009; Pratt & Cullen, 2000), and more recently, a number of studies have extended this theoretical perspective to account for victimization experiences (Baron, Forde, & May, 2007; Forde & Kennedy, 1997; Higgins, Jennings, Tewksburry, & Gibson, 2009; Holtfreter, Reisig, & Pratt, 2008; Jennings, Higgins, Tewksbury, Gover, & Piquero, 2010; Piquero, MacDonald,Dobrin, Daigle, & Cullen, 2005; Schreck, 1999; Schreck, Wright, & Miller, 2002; Stewart & Power, 2002; Stewart et al., 2006).

Comparatively, subcultural theories place an emphasis on the role culture and the environment play in creating opportunities where both victimization and offending are possible. For example,Anderson (1999)has argued that a“code-of-street”exists in minority neighbor- hoods where importance is placed on a might-makes-right attitude. Fig. 1.Overlap of victimization and offending.17 W.G. Jennings et al. / Aggression and Violent Behavior 17 (2012) 16–26 As such, if an individual in this neighborhood who is being socialized under these prevailing cultural norms seeks to increase their status in the neighborhood then they must embrace these cultural norms and exert their physical prowess on other persons in the neighborhood.This action, in and of itself, is an event that leaves them vulnerable to being preyed on and victimized by someone else in the neighborhood who also aspires to be recognized as someone with high social status or“street credit.”A growing literature has supported the association Table 1 Methodological techniques for evaluating the victim–offender overlap.

Authors Location(s) Population Analytical technique Broidy, Daday, Crandall, Klar, and Jost (2006)Bernalillo County, New Mexico All known offenders and victims involved in homicides in Bernalillo County, NM between 1996 and 2001Logistic regression; generalized estimating equations; GLM ANOVA Chang et al. (2003)Nationwide, United States High school seniors Cross-tabulations; Chi-squared tests; logistic regression Daday, Broidy, Crandall, and Sklar (2005)Bernalillo County, NM Victims and offenders involved in the same aggravated battery incidentsConditional logistic regression Dobrin (2001)Prince George's County, MD Homicide victims and random sample of residentsConditional logistic regression; bivariate and multivariate regression Fagan, Piper, and Cheng (1987)Bronx, NY; Dallas, TX; Miami, FL; Chicago, ILSchool-based sample of urban, high-crime neighborhoods, grades 10–12Regression Fiegelman et al. (2000)Baltimore, MD Low-income, predominantly African-Americans ages 9–15Chi-squared; correlation; regression; test for linear trends Gottfredson (1984)England and Wales Nationally representative, household survey of residents 16 years and olderFrequencies, Cross-tabulation Heyman and Smith (2002)Nationwide, United States Nationally representative sample of adults 18 and over who are married or cohabitating, or have a dependent childLogistic regression Hiday, Swanson, Swartz, Borum, and Wagner (2001)One state hospital and three general hospitalsAdult psychiatric patients Logistic regression Jennings et al. (2010)Philadelphia, Portland, Phoenix, Omaha, Lincoln, and Las Cruces6th and 7th graders Group-based trajectory modeling; ANOVA; multinomial regression; cross-tabulation Jennings, Tomisich, Gover, and Akers (2011)South Korea College students Bivariate probit models Jensen and Brownfield (1986)Tucson, AZ; nationally representative, United StatesHigh school students Correlation; Chi-Squared; Regression Klevens et al. (2002)Bogota, Colombia Random sample of the general population (15–60) of BogotaChi-Squared Kuhlhom (1990)Stockholm, Sweden Sentenced offenders and their victims Mean difference Lauritsen and Quinet (1995)Nationwide, United States Random sample of adolescents ages 11–17 Generalized regression Lauritsen, Sampson, and Laub (1991)Nationwide, United States Random sample of adolescents ages 11–17 Regression Linder, Crick, and Collins (2002)Mid-Western University College students Correlation; regression Maldonado-Molina et al.

(2010)Bronx, NY and San Juan, Puerto RicoPuerto Rican adolescents (5–13 years) SEM; cross-tabulation Maldonado-Molina, Piquero, Jennings, Bird, and Canino (2009)Bronx, NY and San Juan, Puerto RicoPuerto Rican adolescents (5 –13 years) Group-based trajectory modeling; ANOVA; multinomial regression Mawby (1979)Sheffield, England School-based sample of children 11–15 Correlation Mayhew and Elliott (1990)England and Wales Nationally representative, household survey of residents 16 years and olderLogistic regression Mustaine and Tewksbury (2000)Four states, United States College students involved in criminal assault Logistic regression Piquero et al. (2005)California, United States Four cohorts of male California Youth Authority ParoleesRare events logistic regression Regoeczi (2000)Canada Homicide cases among victims ages 12–17. Frequencies Reingle, Jennings, Maldonado-Molina, Piquero, and Canino (2011)Bronx, NY Puerto Rican adolescents (5–13 years) Negative binomial regression; mean difference Reingle, Staras, Jennings, Branchini, and Maldonado-Molina (2012)Nationwide, United States Nationally representative sample of adolescents ages 15–26Multinomial regression Sampson and Lauritsen (1990)England and Wales Random household survey of residents ages 16 and olderLogistic regression Savitz et al. (1977)Philadelphia, PA Boys attending public or catholic schools Frequencies Schreck et al. (2008)Nationwide, United States School-based sample of adolescents in 7–12th gradesStructural regression; HLM Silver (2002)Pittsburgh, PA Discharged psychiatric patients 18–40 and non-patients from the same neighborhoodsLogistic regression Silver, Piquero, Jennings, Piquero, and Lieber (2011)Worcester, MA and Kansas City, MOAcute inpatient mental health patients over age 18Bivariate probit models Singer (1981)Review articleLogistic regression Sparks, Glen, and Dodd (1977)London, England General population of adults Frequencies; Correlation Tewksbury and Mustaine (2000)Four states, United States College students involved in assault Logistic regression Van Dijk and Steinmetz (1983)Netherlands and Canada General population Correlation Wittebrood and Nieuwbeerta (1999)Netherlands Random sample of Dutch population over age 15Multilevel logistic regression Wolfgang (1958)Philadelphia, PA Homicides listed by police Frequencies 18W.G. Jennings et al. / Aggression and Violent Behavior 17 (2012) 16–26 between subcultural norms, exposure to neighborhood violence, and participation in offending and experiencing victimization (Anderson, 1999; Baskin & Sommers, 1997; Eitle & Turner, 2002; Felson, 1992; Jacobs & Wright, 2006; Kennedy & Baron, 1993; Nofziger & Kurtz, 2005; Singer, 1987; Stewart et al., 2006).

2. Methodological approaches to identifying the victim–offender overlap Several methodological techniques have been utilized to examine the presence and magnitude of the overlap between victimization and offending. The most common method involves simple bivariate examinations of prevalence, including frequencies, correlations, cross-tabulations, Chi-squared tests, and tests of mean difference (Chang, Chen, & Brownson, 2003; Fiegelman, Howard, Xiaoming, & Cross, 2000; Jensen & Brownfield, 1986; Klevens, Duque, & Ramirez, 2002; Kuhlhom, 1990; Mawby, 1979; Regoeczi, 2000; Savitz, Lalli, & Rosen, 1977; Wolfgang, 1958). Many studies employ bivariate methods for examining the extent of the victim–offender overlap, and expand on thefindings by using regression methods to predict being both a victim and offender (Bryant, Schulenberg, O'Malley, Bachman, & Johnston, 2003; Fiegelman et al., 2000; Jensen & Brownfield, 1986; Maldonado-Molina, Jennings, Tobler, Piquero, & Canino, 2010)(see Table 1).

Many studies use regression methods exclusively in examining the presence of a victim–offender overlap. For example,Reingle et al. (2011)used negative binomial regression to examine the gender differences in exposure to violence (including victimization of vio- lence). Negative binomial regression was used because the depen- dent variable (exposure to violence) was count-based and the distribution was not appropriate for a Poisson regression model due to the overdispersion of exposure to violence. Their analysis indicated that delinquency was strongly predictive of exposure to violence for both males and females. Further, this association was robust and ob- served across time suggesting that the link between exposure to vio- lence including personal victimization and involvement in offending could not be explained away by a number of familial, peer, and con- textual factors.

Victimization and violent offending are often dichotomized for an- alytical purposes, as an individual may be considered having been victimized or not and having offended or not. In these cases, standard logistic regression procedures are frequently used to examine wheth- er offending predicts victimization, and vice versa. The use of this an- alytical methodology is widespread in the literature assessing the victim–offender overlap (Broidy, Daday, Crandall, Klar, & Jost, 2006; Heyman & Smith, 2002; Hiday et al., 2001; Mayhew & Elliott, 1990; Sampson & Laub, 1990; Silver, 2002; Singer, 1981; Wittebrood & Nieuwbeerta, 1999), with the vast majority of the literature using vic- timization as a risk factor for delinquency (Chang et al., 2003; Dobrin, 2001; Fagan, Piper, & Cheng, 1987; Fiegelman et al., 2000; Heyman & Smith, 2002; Hiday et al., 2001; Maldonado-Molina et al., 2009; Mayhew & Elliott, 1990; Singer, 1981) rather than the reciprocal or bidirectional relationship (Gottfredson, 1984; Jensen & Brownfield, 1986; Lauritsen & Quinet, 1995; Lauritsen et al., 1991; Reingle et al., 2011; Sampson & Lauritsen, 1990; Savitz et al., 1977; Silver, 2002; Wittebrood & Nieuwbeerta, 1999).

A small number of case–control studies have used conditional lo- gistic regression procedures in order to examine the overlap between victimization and offending. For example,Dobrin (2001)used a sam- ple of homicide victims matched with a general population sample of non-victims from the same county in Maryland. The victims and non- victims were matched on age, sex, and race. Because of this matching, cases and controls are not independent and conditional logistic re- gression models must be used (Hosmer & Lemeshow, 1989). His anal- ysis indicated that previous offending increases the risk of homicidevictimization. The case –control methodology has been used similarly in other studies (Daday, Broidy, Crandall, & Sklar, 2005).

There are alternative regression methods that have been often used to assess the victim–offender overlap. For example, multinomial regression procedures have been used in studies where individuals have been grouped into more than two categories (e.g., victims only, offenders only, victims and offenders, and non-victims and non-offenders). In this case, the dependent variable is a non-ordinal categorical outcome. Several studies have employed this methodolo- gy in predicting group membership (Jennings et al., 2010; Maldonado-Molina et al., 2009; Reingle et al., 2012).

Bivariate probit models are also appropriate when modeling two separate outcomes (e.g., victimization and offending) jointly (Greene, 1997). This analytic strategy has been employed byJennings, Tomisich, Gover, and Akers (2011) and Silver et al. (2011)to evaluate the victim– offender overlap. Specifically,Jennings et al. (2011)used these models to evaluate predictors of the joint occurrence of psychological and physical dating violence perpetration and victimization (Jennings et al., 2011) and found a strong degree of overlap between dating vi- olence perpetration and victimization that could not be explained by social learning and self-control constructs.Silver et al. (2011)also used a similar methodology to examine the victim–offender overlap among a sample of patients recently released from a psychiatric hos- pital. These authors also found substantial overlap between violence and victimization and a set of mutual and distinct risk factors for groups of victims and offenders. In addition,Piquero et al. (2005) employed a rare events logistic regression strategy when examining predictors of violent offending and homicide victimization, respec- tively, among four cohorts of male California Youth Authority pa- rolees. Their results indicated that approximately 50% of these parolees were re-arrested for a violent offense and 1.5% of these pa- rolees died as a result of a homicide victimizationfive-years post- parole, and that low self-control was shown to predict both violent offending and homicide victimization.

Aside from traditional regression, other methods have been used to examine the victimization/offending relationship. For example, group-based trajectory modeling has been used to evaluate trends in victimization and offending over time (Brame, Nagin, & Tremblay, 2001; Nagin & Tremblay, 2001; Nagin & Tremblay, 2005). Essentially, trajectory methods seek to provide a non-parametric approach for examining behavioral changes in outcomes over age or time. Results from studies using this methodology as applied typically to crime, in- dicate substantial heterogeneity in the shape and volume of offending over the life-course, and typically identify several different groups (low-, medium-, and high-rates) that follow distinct age/crime trajectory profiles (see review inPiquero, 2008). Extending the application of the trajectory method to victimization research, several studies have assessed the degree to which adolescents belong to both high-offending and high-victimization latent groups (Jennings et al., 2010; Maldonado-Molina et al., 2009). For example, Jennings et al. (2010)estimated trajectories of delinquency and vic- timization independently among a longitudinal sample of adolescents who participated in the Gang Resistance Education and Training pro- gram in six United States cities. Once trajectories were estimated, multinomial logistic regression was used to identify which risk and protective factors predicted membership in each trajectory group.

Then, a cross-tabulation was used to assess the degree of overlap be- tween adolescents in the victimization and offending trajectory groups. The authors found a substantial degree of overlap between victimization and offending trajectories (Jennings et al., 2010).

Structural equation modeling has been used in at least one study to evaluate the victim–offender overlap (Maldonado-Molina et al., 2010). In this study, the authors used data from Puerto Rican adoles- cents living in the Bronx, NY, and San Juan, Puerto Rico, to examine the prevalence of the victim–offender overlap and how the associa- tion varies over time. In this case, structural equation modeling was 19 W.G. Jennings et al. / Aggression and Violent Behavior 17 (2012) 16–26 Table 2 Summary of support for the victim–offender overlap.

Authors Support for overlapBrief summary offindings Broidy et al. (2006)+/ More than half (57%) of homicide offenders have prior arrests, and 50% of victims have a prior arrest. Offenders are 45% more likely to have an arrest history than victims. These differences are attributable to the violent histories of offenders.

Chang et al. (2003)+ Repeat victimization was significantly associated with delinquency.

Daday et al. (2005)+ Victims and offenders share histories of violent behavior, property arrests, lifestyle characteristics, and neighborhood characteristics.

Dobrin (2001)+ Victims were 4–10 times more likely to have been previously arrested than non-victims. This association holds for property arrests, violent arrests, and drug arrests. Each arrest increases the odds of homicide by 1.4–5.6 times.

Fagan et al. (1987)+ Students who had been victimized were more likely to have engaged in delinquency; however, the directionality of the relationship is uncertain. Different social processes for each behavior may exist.

Fiegelman et al.

(2000)+ Of perpetrators, only 5% reported not being victimized or witnessing violence, while 74% reported personal victimization and witnessing at least one episode of violence. The correlation (r) between perpetration and victimization was 0.29. Victim status predicted perpetration (b = 0.14).

Gottfredson (1984)+ The strongest association between victimization and offending was between violent crimes and personal victimization. Those participating in delinquent activities were also more likely to be victims, but to a lesser extent.

Heyman and Smith (2002)+ Victimization of family violence as a child is predictive of physical partner abuse perpetration in adulthood.

Hiday et al. (2001)+ The greater victimization participants experienced, the more likely they were to have engaged in afight, assault, or threatened with a weapon and engaged in serious violence.

Jennings et al. (2010)+ Substantial degree of overlap between victimization and offending trajectories. Low self-control is important in distinguishing victimizationand delinquency trajectories.

Jennings et al. (2011)+ Considerable overlap between victimization and offending in dating violence. Social learning and self-control constructs could not explain the overlap.

Jensen and Brownfield (1986)+/ The greater the involvement in criminal behavior, the higher the likelihood of victimization. This does not appear to extend to drug use activity, as use was not associated with increased victimization.

Klevens et al. (2002)+ One-third of the sample has been both victims and perpetrators of violence. An additional one-third has been victims only. Victims only differ demographically and in terms of their routine activities compared to those who are both victims and perpetrators.

Kuhlhom (1990)+ Victims are less likely to have a criminal history, previous violence, and drug abuse compared to offenders; however, the rate of each of these occurrences in victims exceeds the rate in the general population.

Lauritsen and Quinet (1995)+ Individual delinquency significantly predicted victimization for assault, robbery, larceny, and vandalism.

Lauritsen et al. (1991)+ Participation in delinquency is greatest risk factor for assault, robbery, larceny, and vandalism victimization. Substance use was not related tovictimization risk.

Linder et al. (2002)+ There was a significant correlation between romantic relational aggression and romantic victimization.

Maldonado-Molina et al.

(2010)+ Longitudinal overlap between victimization and offending. The association cannot be accounted for by the similarities in risk factors for both groups.

Maldonado-Molina et al.

(2009)+ Exposure to violence (including victimization) predicted membership in trajectories of violent behavior in both samples of Puerto Rican adolescents (living in the Bronx, NY, and Puerto Rico). 20W.G. Jennings et al. / Aggression and Violent Behavior 17 (2012) 16–26 Mawby (1979)+ There was a significant overall correlation between victimization and offender status. This association was observed in all three neighborhoods for males, and one of three neighborhoods for females.

Mayhew and Elliott (1990)+/ Victimization was associated with increased offending only among elderly victims.

Mustaine and Tewksbury (2000)+ Three distinct groups of those exposed to assault: victims, offenders, and victims–offenders. Each of these groups can be predicted by varying lifestyle measures.

Piquero et al. (2005)+ Among the four cohorts of male California Youth Authority parolees, approximately 50% of these parolees were arrested for a violent offense and 1.5%of these parolees died as a result of a homicide victimization. Low self-control predicted both violence perpetration and homicide victimization.

Regoeczi (2000)+/ Of homicide victims in Canada, 67.4% did not have a previous criminal record. The largest proportion of victims had a property offense (12%), followedby a record of violence (9%). The authors conclude that this small portion of victims with a criminal record provides weak support for the overlap.

Reingle et al. (2011)+ Delinquency increased exposure to violence (including victimization) among both males and females longitudinally over a three-year period.

Reingle et al. (2012)+ 11.9% of a nationally representative sample reported both victimization and perpetration of intimate partner violence. Approximately 30% reported either victimization or perpetration.

Sampson and Lauritsen (1990)+ Participating in violence, theft, vandalism, and drug use increased the risk of violent victimization. These behaviors indicate a criminal lifestyle, which increases propensity for victimization.

Savitz et al. (1977)+/ There was an association between assault victimization and delinquent arrest; however, there was no relationship between theft victimization and arrest status.

Schreck et al. (2008)+/ Results supported adolescents as having a tendency towards either victimization or violent offending. In this case, one role usually predominated.Age and heavy alcohol use predicted one role better than the other.

Silver (2002)+ Mentally disordered people were more likely than non-disordered people to be victimized. Violent behavior significantly predicted victimization of violence. Conflict in social relationships appears to play a part in the effect of mental disease on victimization.

Silver et al. (2011)+ Violent victimization and offending are strongly correlated, with several mutual and distinct risk factors.

Singer (1981)+ Victimization is a potent predictor of offender status.

Sparks et al. (1977)+ Association between violence and victimization of violence. Those who reported being perpetrators of violence were less likely to report victimization to the police.

Tewksbury and Mustaine (2000)+ Those who were victims of vandalism were more likely to be offenders of vandalism. This overlap can be explained by similarities in the daily routinesof victims and offenders.

Van Dijk and Steinmetz (1983)+ Overlap between participation in theft and victimization for theft. Normative values may be weakened by victimization, resulting in perpetrationof the behavior.

Wittebrood and Nieuwbeerta (1999)+ Those who participate in violent behavior are more likely to be victims than those who are non-violent. This overlap may be partially explained by overlap in risk factors for both behaviors.

Wolfgang (1958)+ 64% of homicide offenders and 47% of homicide victims had previous arrest records. Many victims of homicide provoke the incident through some other related (potentially crime-related) behavior.

Total = 37 studies 31 studies provide support 6 studies provide mixed/limited support + Findings provide support the victim–offender overlap.

+/ Findings provide mixed/limited support the victim–offender overlap.

Findings do not support the victim–offender overlap. 21 W.G. Jennings et al. / Aggression and Violent Behavior 17 (2012) 16–26 used to allow for victimization and offending behaviors to vary over time, and simultaneously evaluate the influence of multi-level risk factors on offending, victimization, and the overlap between the two. Findings suggested that the victim–offender overlap was robust and stable over time and that it could not be explained away by shared, multi-level risk factors.

Overall, a wide variety of analytical procedures have been used to examine the victim–offender overlap. The majority of the studies have employed a hybrid approach, using bivariate methods to exam- ine the presence of a relationship and then using regression methods to predict the offending/victimization overlap. However, alternative methods have been used in recent years to allow for changes in vio- lence and victimization over time (Jennings et al., 2010; Maldonado-Molina et al., 2009; Maldonado-Molina et al., 2010).

These models allow for patterns of criminal behavior and victimiza- tion to vary as adolescents age, the identification of distinct trajecto- ries of offending, victimization, and their overlap, and then assessing how risk/protective factors distinguish between trajectories thereby reflecting the emerging prominence of the life-course perspective for understanding development over time.

3. Normative assessments of the victim–offender overlap 3.1. Historical documentation of the overlap The recognition that there is a consistent relationship between victimization and offending and between victims and offenders is not a recent phenomenon, as research identifying the existence of a high-risk group that experiences both victimization and offending dates back to the middle of the 20th century. Much of this pioneering research was conducted in Philadelphia using homicide victims (Wolfgang, 1958), with many of these early studies reporting that victims were likely to have a record of criminal offending themselves (Fagan et al., 1987; Gottfredson, 1984; Jensen & Brownfield, 1986; Wolfgang, 1958). As shown inTable 2, a number of studies have found direct associations between victimization and offending for as- sault, larceny, robbery, vandalism (Lauritsen & Quinet, 1995), vio- lence, theft, drug use (Sampson & Lauritsen, 1990), and offender status in general (Mawby, 1979; Singer, 1981; Sparks et al., 1977).

For example,Van Dijk and Steinmetz (1983)speculated that victimi- zation weakens normative values, resulting in subsequent offending.

Alternatively,Wolfgang (1958)proposed that persons who end up as victims often provoke crime. In his research, Wolfgang found that homicide was often provoked by another crime perpetrated by the victim of the homicide.

The existence of the victim–offender overlap is not an overwhelm- ingly common occurrence. Although it is the case that several studies have found support for the overlap for some behaviors, exceptions have been reported. For example,Jensen and Brownfield (1986) used data from the Monitoring the Future study, as well as an inde- pendent sample of high school students in Tucson, Arizona, and ob- served a direct, positive relationship between criminal behavior and victimization; however, they reported that this relationship did not appear to include drug use.Lauritsen et al. (1991)found that sub- stance use was not related to risk of criminal victimization. Further, Savitz et al. (1977)suggested that there was no relationship between theft victimization and arrest status; however, the relationship be- tween assault victimization and arrest status was strong.

Other studies have found that the victim–offender overlap applies only to certain groups of individuals. In Stockholm,Kuhlhom (1990) found that victims are less likely to have a criminal history and report previous violence and drug abuse when compared to offenders. How- ever, when compared to the general population, victims have higher rates of these behaviors. Mayhew and Elliott also found mixed sup- port for the victim–offender overlap. In their analysis of the British Crime Survey (1982–1984), they found that victimization wasassociated with increased offending among elderly victims only. The authors suggest that this was likely due to the heterogeneous“victim- ization”category which may have concealed strong relationships be- tween specific forms of victimization and offending behavior (Mayhew & Elliott, 1990).

Much of the historical literature suggests that the relationship be- tween victimization and offending is a function of common risk fac- tors associated with both behaviors (Gottfredson, 1984; Jensen & Brownfield, 1986).Sampson and Lauritsen (1990)suggest that a deviant lifestyle is responsible for the overlap between victimization risk and criminal offending. Some situational and contextual variables they identify that predicted both victimization and offending include being non-married and spending nights out on the town. Lower levels of education, younger age, and being male were also associated with both victimization and offending (Sampson & Lauritsen, 1990).

Lauritsen et al. (1991)elaborated on the deviant lifestyle hypothesis using data from the National Youth Survey (Lauritsen et al., 1991).

They found that participation in pro-social activities reduced the risk of criminal victimization, while participation in a“delinquent lifestyle”increased the risk of victimization for robbery, larceny, vandalism, and assault.

In summary, the historical research suggests that a homogenous group of both victims and offenders exists. Yet, it remains unclear whether this typology applies across all forms of criminal behavior.

This body of research also suggests that the observed overlap is attrib- utable to shared risk factors or a delinquent lifestyle prone to victim- ization. More recent work in this area has further supported the existence of the overlap whereas documentation of shared risk fac- tors or the“delinquent lifestyle”hypothesis has received less support but still remains under-studied.

3.2. Contemporary documentation of the victim–offender overlap The majority of the contemporary literature on the victim–offender overlap indicates widespread support for the existence of this group (Broidy et al., 2006; Chang et al., 2003; Dobrin, 2001; Fiegelman et al., 2000; Heyman & Smith, 2002; Jennings et al., 2010; Jennings et al., 2011; Maldonado-Molina et al., 2009; Maldonado-Molina et al., 2010; Reingle et al., 2011; Reingle et al., 2012; Silver, 2002; Silver et al., 2011; Wittebrood & Nieuwbeerta, 1999). The overlap appears to be most pronounced in the most severe of criminal behaviors: homicide.

In a study of homicide victims in New Mexico,Broidy et al. (2006) found that 57% of offenders had prior arrests and 50% of homicide victims had a prior arrest. Similarly, in the case–control study of ho- micide victims reviewed earlier,Dobrin (2001)found that victims were between four and ten times more likely to have been previous- ly arrested (for property crimes, violent crimes, and drug-related ar- rests) than non-victims. Specifically, Dobrin reported that each arrest increased the risk of homicide 1.4 to 5.6 times.

Among crimes other than homicide, the relationship between vic- timization and offending is strong as well. For instance,Fiegelman et al. (2000)found that only 5% of perpetrators were not victimized or exposed to violence. Conversely, 74% reported either victimization or witnessing at least one episode of violence (Fiegelman et al., 2000). Other studies also provide evidence of victimization as a risk factor for delinquent behavior (Chang et al., 2003; Daday et al., 2005).

The rationale for the overlap remains a subject of debate. Consis- tent with historical work, some authors suggest that the overlap can be predicted through various demographic variables, or a delinquent lifestyle (Daday et al., 2005). One study byJennings et al. (2010)iden- tified a considerable degree of overlap between victimization and offending trajectories. They found a number of contextual influences that differentiated groups of victims compared with offenders, in- cluding school commitment (lower in victims), parental monitoring (lower in victims), biological sex, and low self-control (lower self- control in both delinquent adolescents and victims; however, the 22W.G. Jennings et al. / Aggression and Violent Behavior 17 (2012) 16–26 magnitude of the relationship was stronger and increased as partici- pation in delinquency increased). The authors suggest that low self- control is the most potent differentiator of victim and offender typolo- gies. In another study examining the predictors of the victim/offender overlap,Schreck et al. (2008)found evidence of the overlap, but also noted substantial differentiation in the roles of victim or offender. Indi- viduals' role as an offender or a victim was relatively stable over time, indicating that adolescents were usually a victim or an offender consis- tently over time. Further,Schreck et al. (2008)reported that frequent intoxication and lower school commitment and attachment to parents significantly predicted violent offending rather than victimization, while older teenagers were more likely to report being victims than younger adolescents.

In theBroidy et al. (2006)study that differentiated among homi- cide victims and offenders in New Mexico, homicide offenders were much more likely than victims to have prior violent histories. When comparing offenders and victims with non-violent histories, no sig- nificant differences emerged. Thisfinding adds to the research litera- ture suggesting that previous violence more strongly predicts offending compared to victimization (Kuhlhom, 1990; Schreck et al., 2008).

3.3. The overlap in dating violence A small number of studies have considered the victim–offender overlap as it relates to dating violence in two ways: 1) the inter- generational effect of victimization on future dating violence perpetra- tion and victimization; and 2) the event-specificroleplayedinperpe- tration and victimization of dating violence. Thefirst question has received substantial support, as exposure to or victimization from fam- ily violence as a child increases the risk of physical partner abuse or per- petration in adulthood (Heyman & Smith, 2002; Widom, 1989). The second question assessing the relationship between dating violence vic- timization and perpetration at the event level has received less atten- tion. One study byReingle et al. (2012)reported significant overlap between victims and perpetrators of dating violence in a nationally rep- resentative sample of young adults in the United States. In fact, of the 30% of the sample who reported any exposure to dating violence, 11.9% reported both victimization and perpetration. A number of other studies have been conducted to assess the victim–offender rela- tionship in dating violence internationally, and these studies will be dis- cussed later.

3.4. Cross-cultural investigations of the victim–offender overlap The relationship between victimization and offending has been in- vestigated across cultures, especially among Hispanics living in the United States. This specific ethnic group has recently become the focus of risk-related research, as a number of studies have identified Hispanics who become acculturated to the American lifestyle as‘at risk’for disease, disability and death from a number of high-risk be- haviors (Caetano & McGrath, 2005; Caetano, Ramisetty-Mikler, & Rodriguez, 2008; Caetano, Schafer, Clark, Cunradi, & Raspberry, 2000; Maldonado-Molina, Reingle, Jennings, & Prado, 2011). In addi- tion, Hispanics living in the United States have a unique set of expo- sures, including cultural stress, neighborhood disorganization and transiency, and poverty (Lara-Cinisomo, Xue, & Brooks-Gunn, 2009).

This is known as the“immigrant paradox”or the tendency for foreign-born Hispanics, who would be expected to show poorer signs of health due to immigration and lifestyle disruptions, to report more favorable health indicators than US-born Hispanics (Vega, Rodriguez, & Gruskin, 2009).

At least three studies have evaluated the extent of the victim–of- fender overlap using exclusively Hispanic populations (Maldonado- Molina et al., 2009; Maldonado-Molina et al., 2010; Reingle et al., 2011). Each of the three studies used data from the Boricua YouthStudy, drawing samples from Hispanics living in the Bronx, NY, and/ or San Juan, Puerto Rico. To assess gender differences in exposure to violence (including victimization),Reingle et al. (2011)used the sam- ple of Hispanics residing in the United States and found that delin- quency significantly predicted exposure to violence consistently over time, indicating that individual delinquency was associated with victimization among Hispanics.

The second study was conducted using data from both Puerto Ricans living in the Bronx, NY and those living in San Juan, Puerto Rico. In this study,Maldonado-Molina et al. (2009)sought to describe the trajectories of delinquency across the two contexts and reported both differences and similarities between the samples. Nevertheless, exposure to violence predicted membership in all delinquent trajec- tory groups (compared to the non-delinquent group) across both samples, with the magnitude of the effect being slightly larger among youth in San Juan compared to those in the Bronx.

To provide additional evidence of the victim/offender overlap, Maldonado-Molina et al. (2010)also used the Bronx Puerto Rican youth data to evaluate the extent of the longitudinal victim–offender overlap. The authors found that the overlap exists and persists over time (from childhood into adolescence). The largest group was victims only (32–44% of the sample), followed by non-victims and non- offenders (31–36%), victims and offenders (15–27%), and offenders only (4–9%). Approximately 5% of the sample included“chronic victim– offenders”who reported being both an offender and a victim at each of the three time points. The percentage of youth who were in the “victim–offender”group decreased over time, but members of this group were at exceptionally high risk in terms of individual, familiar, peer, and contextual risk factors.

3.5. International explorations of the victim–offender overlap Evidence of the victim–offender overlap extends beyond research conducted in the United States. Studies of the overlap have been con- ducted in Bogota, Colombia, the Netherlands, and Canada, andfind- ings have been variable across countries. In Bogota,Klevens et al.

(2002)randomly sampled 3500 individuals selected to be represen- tative of the population. They found that 38.6% of the population was victim only, 2.9% were offenders only, and 32.2% were both vic- tims and offenders. A study using data from the Netherlands Survey on Criminality and Law Enforcement (Wittebrood & Nieuwbeerta, 1999) found that offenders of violent crimes, property crimes, and vandalism were significantly more likely than non-offenders to be victimized by the same type of crime committed. The risk was great- est for those who committed violent crimes (OR = 15.21), followed by vandalism (OR = 12.93), and property crimes (OR = 3.90). A third study provides a conflicting interpretation of the presence of a victim–offender overlap group. In Canada, Regoeczi (2000)used ho- micide data from the Canadian Centre for Justice Statistics, 1991– 1995, involving young victims (ages 12–17). Of the 114 victims killed, 67.4% did not have a previous criminal record. However, 13 victims had a record for a violent offense, 17 had a record for a property of- fense, and 4 had a drug conviction. The authors concluded that these statistics did not provide support for the presence of a victim– offender overlap, although no analytical tests of significance were provided.

Several studies have been conducted to evaluate the effect of the victim–offender overlap in dating violence internationally (Jennings et al., 2011; Paterson, Feehan, Butler, Williams, & Cowley-Malcolm, 2007). In New Zealand,Paterson et al. (2007)studied a group of new mothers to evaluate past year dating violence and uncovered that few mothers were perpetrators only or victims only. Specifically, 21% reported perpetrating minor violence (but only 2% were only perpetrators), 35% reported minor victimization (only 11% were vic- tims only), 19% were victims of severe violence (1% were victims only), and 11% were perpetrators of severe violence (none were 23 W.G. Jennings et al. / Aggression and Violent Behavior 17 (2012) 16–26 exclusively perpetrators).Jennings et al. (2011)used data from a sample of 1399 South Korean college students to evaluate dating vio- lence experiences and other related behaviors. They found a substan- tial overlap between both physical and psychological victimization and perpetration. Although a few variables explained this relation- ship partially (e.g., experiencing childhood physical abuse), social learning and self-control variables could not explain the observed overlap.

3.6. Intersections of the victim–offender overlap and mental health Adolescents and adults who suffer from mental disorders are known to be at increased risk for violent offending (Link, Andrews, & Cullen, 1992; Hiday, 2006; Hiday et al., 2001; Link & Stueve, 1995) and victimization (Hiday, Swartz, Swanson, Borum, & Wagner, 1999; Hiday et al., 2001; Silver, 2002; Silver, Arseneault, Langley, Caspi, & Moffitt, 2005) compared to those without mental disorders.

Despite these well-documented associations, few studies have evalu- ated the overlap among mentally disordered persons (Hiday et al., 2001; Silver, 2002; Swanson, Borum, Swartz, & Hiday, 1999).

To our knowledge, only three studies have evaluated the overlap between victimization and offending among a mentally disordered population.Silver (2002)used a sample of discharged psychiatric pa- tients and a random sample of non-patients from the same neighbor- hoods. Results from this study found that violent behavior in the past ten weeks increased the odds of violent victimization by more than four. Thisfinding held even when the use of illegal drugs was includ- ed in the model. In a follow-up study,Silver et al. (2011)evaluated the victim–offender overlap in mentally disordered patients using the same sample detailed above. Results indicated that 13% of the sample had committed a violent offense, 19% were victims of violent offenses, and 5.6% were both a victim of violence and committed a vi- olent offense. Bivariate probit models, which refer to a class of models in which two probit models are estimated jointly, revealed that the error terms for victimization and offending were correlated, indicat- ing a degree of overlap. And while victimization and offending had many of the same risk factors (e.g., substance abuse diagnosis, psy- chopathy, perceived stress, and residential moves), others were dis- tinct for each group.

The third study evaluating the effect of prior victimization on fu- ture violent behavior was performed byHiday et al. (2001)using a sample of 331 mentally ill patients from three psychiatric units in general hospitals. Results indicated that regardless of the measure- ment of violence, any violent behavior was associated with victimiza- tion. The reciprocal effect was also detected, as those who were victimized were approximately 76% more likely to be violent than those who were not victimized. The authors suggest that victimiza- tion may be a path to violence among mentally ill individuals.

4. Conclusions and directions for future research This article set out to provide a broad overview of the knowledge base regarding the overlap between offenders (offending) and vic- tims (victimization). In so doing, several theoretical perspectives were discussed, as were methodological applications and reviews of dozens of studies in the more general areas of victimization and offending, as well as among select topics including dating violence, cross-cultural and international studies, and among mental health populations. Overallfindings from this review indicate a rather strong overlap between offenders and victims as well as the risk/protective factors that explain membership in this combined typology, but that the relationship is far from perfect, that it varies across types of of- fenses (stronger for more violent than property offenses), and that risk factors occupy some shared but also some discrepant relation- ships across the various victim/offender permutations. With this in mind, we next outline an expansive research agenda to betterdocument and understand the overlap between victimization and offending.

4.1. The need for more and potentially better theory Currently, two theoretical perspectives dominate thefield's un- derstanding of the victim/offender overlap, routine activities and the general theory of crime. Although we do not take issue with these frameworks and their ability to help describe the overlap, there is a need for better understanding the role of selection and cau- sation in the process that propels the risk for subsequent offending among victims and subsequent victimization among offenders. For example, what is the nature and extent of individuals self-selecting into situations that heighten the risk of one, the other, or both out- comes? Does the mixture of individual and situational characteristics that increase the likelihood of victimization and offending vary over the life course, across different crime/victimization types, and demo- graphic characteristics?

4.2. The need for continued application and development of advanced methods One of the most exciting methodological developments in the social sciences has been the trajectory method. Not only have applications of this technique yielded important insight into the longitudinal pattern- ing of crime over the life course and the characteristics that may distin- guish membership across trajectories, but recent advances in the area have permitted researchers to examine joint trajectories of victimiza- tion and offending. Although there remain very few instances where such analyses have been undertaken, it will be especially important to examine how victimization and offending co-occur throughout the life course. A basic, descriptive accounting of this overlap will be important in and of itself so as to document whether the overlap exists early on, whether it emerges in adolescence or in adulthood, and whether it wanes as adulthood approaches. In short, does the offender/victim overlap mimic that aggregate age–crime curve (seePiquero, Farrington, & Blumstein, 2003)?

4.3. The need for expanded consideration of offense types For the most part, research on the victim–offender overlap has tended to provide cursory information on the concordance between the two outcomes. There is a need to further unpack simple measures of victimization and offending in order to examine whether the rela- tionship is driven by a shared relation among certain behaviors. In this regard, it would be of specific interest to examine whether the experience of one type of victimization may lead to specific forms of subsequent criminal offending, perhaps in a retaliatory manner (i.e., the experience of a verbal insult is followed by the commission of an aggravated assault), and also whether these inter-relationships are spaced closely together or farther apart in time. A concerted focus here should be placed on violent crimes, especially homicide, as a way to understand whether these are isolated events, if they are predictable—both with respect to perpetrating and experiencing, and to understand how situational and individual characteristics are implicated in its overlap.

4.4. The need for studies to collect victimization and offending data Perhaps because researchers have been primarily interested in ei- ther offending or victimization, oftentimes data on both outcomes have not been collected—and not collected longitudinally. Thus, there is a pressing need to collect information on both victimization and offending. There is also a need to consider the collection of more detailed information on victimization and offending not only in shorter time segments (not simple‘past 12 month’items) but to 24W.G. Jennings et al. / Aggression and Violent Behavior 17 (2012) 16–26 also consider questions about the nature of victimization and the na- ture of offending: was one a function of the other, did one happen during the commission/experience of the other? Although this level of detail certainly complicates data collection matters, it is through this kind of detailed lens that researchers will be able to better grasp how and why victimization and offending are co-integrated with one another. Future studies should also make an effort to disag- gregate the constructs for measuring victimization and offending.

Perhaps it is the case that violent offending specifically and not prop- erty or drug offending increases the risk for victimization, particularly violent victimization. However, it is certainly possible that general offending behaviors are, in and of themselves, risky enough to in- crease an individual's likelihood for victimization in any and all forms. Future efforts that unpack the complexity of the relationships between specific types of offending behaviors and victimization expe- riences are a worthwhile endeavor.

In the end, we recognize that we have sketched a comprehensive picture of the kind of research that is needed to further thefield's un- derstanding of the relationship between victimization and offending.

Our hope is that in 20 years time, the issues outlined above will have been studied, important theoretical developments and empirical dis- coveries attained, and another set of research questions will have emerged. This is the process of science, but it can only begin with careful, descriptive research on the overlap between victimization and offending.

References Anderson, E. (1999).Code of the street: Decency, violence, and the moral life of the inner city.New York, NY: W. W. Norton.

Baron, S. W., Forde, D. R., & May, F. M. (2007). Self-control, risky lifestyles, and situa- tion: The role of opportunity and context in the general theory.Journal of Criminal Justice,35, 119–136.

Baskin, D. R., & Sommers, I. B. (1997).Casualties of community disorder: Women's careers in violent crime.Boulder, CO: Westview.

Brame, B., Nagin, D. S., & Tremblay, R. E. (2001). Developmental trajectories of physical aggression from school entry to late adolescence.Journal of Child Psychology and Psychiatry,42, 503–512.

Broidy, L. M., Daday, J. K., Crandall, C. S., Klar, D. P., & Jost, P. F. (2006). Exploring demo- graphic, structural, and behavioral overlap among homicide offenders and victims.

Homicide Studies,10, 155–180.

Bryant, A. L., Schulenberg, J. E., O'Malley, P. M., Bachman, J. G., & Johnston, L. D. (2003).

How academic achievement, attitudes, and behaviors relate to the course of sub- stance use during adolescence: A 6-year, multiwave national longitudinal study.

Journal of Research on Adolescence,13, 361–397.

Caetano, R., & McGrath, C. (2005). Driving under the influence (DUI) among US ethnic groups.Accident Analysis and Prevention,37, 217–224.

Caetano, R., Ramisetty-Mikler, S., & Rodriguez, L. A. (2008). The Hispanic American Baseline Alcohol Survey (HABLAS): DUI rates, birthplace, and acculturation across Hispanic national groups.Journal of Studies on Alcohol and Drugs,69, 259–265.

Caetano, R., Schafer, J., Clark, C. L., Cunradi, C. B., & Raspberry, K. (2000). Intimate part- ner violence, acculturation, and alcohol consumption among Hispanic couples in the United States.Journal of Interpersonal Violence,15,30–45.

Chang, J. J., Chen, J. J., & Brownson, R. C. (2003). The role of repeat victimization in ad- olescent delinquent behaviors and recidivism.Journal of Adolescent Health,32, 272–280.

Daday, J. K., Broidy, L. M., Crandall, C. S., & Sklar, D. P. (2005). Individual, neighborhood, and situational factors associated with violent victimization and offending.Criminal Justice Studies,18,215–235.

Dobrin, A. (2001). The risk of offending on homicide victimization: A case control study.Journal of Research in Crime and Delinquency,38, 154–173.

Eitle, D., & Turner, R. J. (2002). Exposure to community violence and young adult crime: The effects of witnessing violence, traumatic victimization, and other stress- ful life events.Journal of Research in Crime and Delinquency ,39, 214–237.

Fagan, J., Piper, E. S., & Cheng, Y. (1987). Contributions of victimization to delinquency in inner cities.The Journal of Criminal Law and Criminology,78, 586–613.

Felson, M. (1986). Linking criminal choices, routine activities, informal control, and criminal outcomes. In D. B. C. R. V. Clarke (Ed.),The reasoning criminal: Rational choice perspectives on offending(pp. 119–128). New York: Springer-Verlag.

Felson, R. B. (1992). Kick 'em when they're down: Explanations of the relationship be- tween stress and interpersonal aggression and violence.The Sociological Quarterly, 33,1–16.

Fiegelman, S., Howard, D. E., Xiaoming, L., & Cross, S. I. (2000). Psychosocial and envi- ronmental correlates of violence perpetration among African-American urban youth.Journal of Adolescent Health,27, 202–209.

Forde, D. R., & Kennedy, L. W. (1997). Risky lifestyles, routine activities, and the general theory of crime.Justice Quarterly,14, 265–294.Gottfredson, M. G. (1981). On the etiology of criminal victimization.Journal of Criminal Law and Criminology,72, 714–726.

Gottfredson, M. R. (1984).Victims of crime: The dimensions of risk.United Kingdom:

Home Office Research and Planning Unit.

Gottfredson, M. R. (2009). The empirical status of control theory in criminology. In F. T.

Cullen, J. P. Wright, & K. R. Blevins (Eds.),Taking stock: The status of criminological theory,Vol. 15. (pp. 77–100)New Brunswick, NJ: Transaction Publishers.

Gottfredson, M. R., & Hirschi, T. (1990).A general theory of crime.Stanford, CA: Stanford University Press.

Greene, W. (1997).Econometric analysis(3rd ed.). Englewood Cliffs, New Jersey: Prentice Hall.

Heyman, R. E., & Smith, A. M. (2002). Do child abuse and interpersonal violence lead to adulthood family violence?Journal of Marriage and Family,64, 684–870.

Hiday, V. (2006). Putting community in perspective: A look at correlations, causes and controls.International Journal of Law and Psychiatry,29, 316–331.

Hiday, V. A., Swanson, J. W., Swartz, M. S., Borum, R., & Wagner, H. R. (2001). Victimization:

A link between mental illness and violence?International Journal of Law and Psychiatry, 24,559–572.

Hiday, V. A., Swartz, M. S., Swanson, J. W., Borum, R., & Wagner, H. R. (1999). Criminal victimization of persons with severe mental illness.Psychiatric Services,50,62–68.

Higgins, G. E., Jennings, W. G., Tewksburry, & Gibson, C. L. (2009). Exploring the link be- tween low self-control and violent victimization trajectories in adolescents.Criminal Justice and Behaviors,36,1070–1084.

Holtfreter, K., Reisig, M. D., & Pratt, T. C. (2008). Routine activities, low self-control, and fraud victimization.Criminology,46, 189–220.

Hosmer, D. W., & Lemeshow, S. (1989).Applied logistic regression.New York, New York:

John Wiley & Sons.

Jacobs, B. A., & Wright, R. (2006).Street justice: Retaliation in the criminal underworld.

New York, NY: Cambridge.

Jennings, W. G., Higgins, G. E., Tewksbury, R., Gover, A. R., & Piquero, A. R. (2010). A longitudinal assessment of the victim–offender overlap.Journal of Interpersonal Violence,25, 2147–2174.

Jennings, W. G., Tomisich, E. A., Gover, A. R., & Akers, R. L. (2011). Assessing the overlap in dating violence perpetration and victimization among South Korean college stu- dents: The influence of social learning and self control.American Journal of Criminal Justice,36, 188–206.

Jensen, G. F., & Brownfield, D. (1986). Gender, lifestyles, and victimization: Beyond rou- tine activity theory.Violence and Victims,1,85–99.

Kennedy, L. W., & Baron, S. W. (1993). Routine activities and a subculture of violence: A study of violence on the street.Journal of Research in Crime and Delinquency,30, 88–112.

Klevens, J., Duque, L. F., & Ramirez, C. (2002). The victim–offender overlap and routine activities: Results from a cross-sectional study in Bogota, Columbia.Journal of Interpersonal Violence,17, 206–216.

Kuhlhom, E. (1990). Victims and offenders of criminal violence.Journal of Quantitative Criminology,6,51–59.

Lara-Cinisomo, S., Xue, Y., & Brooks-Gunn, C. (2009). Hispanic immigrant youth and in- ternalizing behaviors: Examining the links with neighborhood context.RAND Labor and Population, Santa Monica, CA: RAND Corp.

Lauritsen, J. L., & Quinet, K. F. D. (1995). Repeat victimization among adolescents and young adults.Journal of Quantitative Criminology ,11, 143–166.

Lauritsen, J. L., Sampson, R. J., & Laub, J. H. (1991). The link between offending and vic- timization among adolescents.Criminology,29, 265–292.

Linder, J. R., Crick, N. R., & Collins, A. (2002). Relational aggression and victimization in young adults' romantic relationships: Associations with perceptions of parent, peer and romantic relationship quality.Social Development,11,69–86.

Link, B. G., Andrews, H., & Cullen, F. T. (1992). The violence and illegal behavior of men- tal patients reconsidered.American Sociological Review,57, 275–292.

Link, B. G., & Stueve, A. (1995). Evidence bearing on mental illness as possible cause of violence behavior.Epidemiologic Reviews,17, 172–181.

Maldonado-Molina, M. M., Jennings, W. G., Tobler, A. L., Piquero, A. R., & Canino, G.

(2010). Assessing victim–offender overlap among Hispanic youth.Journal of Criminal Justice,38, 1191–1201.

Maldonado-Molina,M.M.,Piquero,A.R.,Jennings,W.G.,Bird,H.R.,&Canino,G.J.

(2009). Trajectories of delinquent behaviors among Puerto Rican children and adolescents at two sites.Journal of Research in Crime and Delinquency,46, 144–181.

Maldonado-Molina,M.M.,Reingle,J.M.,Jennings,W.G.,&Prado,W.(2011).Drinking and driving among immigrant and US-bornHispanic young adults: Results from a longitudinal and nationally representative study.Addictive Behaviors,36, 381–388.

Mawby, R. I. (1979). The victimization of juveniles: A comparative study of three areas of publicly owned housing in Sheffield.Journal of Research in Crime and Delinquency,16, 98–113.

Maxfield, M. (1987). Lifestyle and routine activity theories of crime: Empirical studies of victimization, delinquency, and offender decision-making.Journal of Quantitative Criminology,3,275–282.

Mayhew, P., & Elliott, D. (1990). Self-reported offending, victimization, and the British Crime Survey.Violence and Victims,5,83–96.

Mustaine, E. E., & Tewksbury, R. (2000). Comparing the lifestyles of victims, offenders, and victim–offenders: A routine activity theory assessment of similarities and dif- ferences for criminal incident participants.Sociological Focus,33, 339–362.

Nagin, D. S., & Tremblay, R. E. (2001). Analyzing developmental trajectories of distinct but related behaviors: A group-based method.Psychological Methods,6,18–34.

Nagin, D. S., & Tremblay, R. E. (2005). Developmental trajectory groups: Fact or a useful statisticalfiction?Criminology,43, 873–904.25 W.G. Jennings et al. / Aggression and Violent Behavior 17 (2012) 16–26 Nofziger, S., & Kurtz, D. (2005). Violent lives: A lifestyle model linking exposure to vi- olence to juvenile violent offending.Journal of Research in Crime and Delinquency, 42,3–26.

Osgood, D. W., Wilson, J. K., O'Malley, P. M., Bachman, J. G., & Johnston, L. D. (1996).

Routine activities and individual deviant behavior.American Sociological Review, 61, 635–655.

Paterson, J., Feehan, M., Butler, S., Williams, M., & Cowley-Malcolm, E. T. (2007). Inti- mate partner violence within a cohort of Pacific mothers living in New Zealand.

Journal of Interpersonal Violence,22, 698–721.

Piquero, A. R. (2008). Taking stock of developmental trajectories of criminal activity over the life course. In Akiva Liberman (Ed.),The long view of crime: A synthesis of longitudinal research(pp. 23–78). New York: Springer.

Piquero, A. R., Farrington, D. P., & Blumstein, A. (2003). The criminal career paradigm. In Tonry Michael (Ed.),Crime and justice: A review of research,Vol. 30.(pp.359–506).

Chicago: University of Chicago Press.

Piquero, A. R., MacDonald, J. M., Dobrin, A., Daigle, L., & Cullen, F. T. (2005). Self-control, violent offending, and homicide victimization: Assessing the general theory of crime.Journal of Quantitative Criminology,21,55–71.

Pratt, T. C., & Cullen, F. T. (2000). The empirical status of Gottfredson and Hirschi's gen- eral theory of crime: A meta-analysis.Criminology,38, 931–964.

Regoeczi, W. (2000). Adolescent violent victimization and offending: Assessing the ex- tent of the link.Canadian Journal of Criminology,42, 493–505.

Reingle, J. M., Jennings, W. G., Maldonado-Molina, M. M., Piquero, A., & Canino, G.

(2011). Investigating the role of gender and delinquency in exposure to violence among Puerto Rican youth.Journal of Contemporary Criminal Justice,27, 361–377.

Reingle, J. M., Staras, S. A., Jennings, W. G., Branchini, J., & Maldonado-Molina, M. M.

(2012). The relationship between marijuana use and intimate partner violence in a nationally representative, longitudinal sample.Journal of Interpersonal Violence, 27(8).

Sampson, R. J., & Laub, J. H. (1990). Crime and deviance over the life course: The sa- lience of adult bonds.American Sociological Review,29, 609–627.

Sampson, R. J., & Lauritsen, J. L. (1990). Deviant lifestyles, proximity to crime, and the offender–victim link in personal violence.Journal of Research in Crime Delinquency, 27, 110–139.

Savitz, L. D., Lalli, M., & Rosen, L. (1977).City life and delinquency: Victimization, fear of crime, and gang membership.Washington, D.C.: Department of Justice, National In- stitute for Juvenile Justice and Delinquency Prevention.

Schreck, C. J. (1999). Criminal victimization and low self-control: An extension and test of a general theory of crime.Justice Quarterly,16, 633–654.

Schreck, C. J., Fisher, B. S., & Miller, J. M. (2004). The social context of violent victimiza- tion: A study of the delinquent peer effect.Justice Quarterly,21,23–48.

Schreck, C. J., Stewart, E. A., & Osgood, D. W. (2008). A reappraisal of the overlap of vi- olent offenders and victims.Criminology,46, 871–906.

Schreck, C. J., Wright, R. A., & Miller, J. M. (2002). A study of individual and situational antecedents of violent victimization.Justice Quarterly,19, 159–180.Silver, E. (2002). Mental disorder and violent victimization: The mediating role of in- volvement in conflicted social relationships.Criminology,40, 191–212.

Silver, E., Arseneault, L., Langley, J., Caspi, A., & Moffitt, T. E. (2005). Mental disorder and violent victimization in a total birth cohort.American Journal of Public Health,95, 2015–2021.

Silver, E., Piquero, A. R., Jennings, W. G., Piquero, N., & Lieber, M. (2011). Assessing the violent offending and violent victimization overlap among discharged psychiatric patients.Law and Human Behavior,35,49–59.

Singer, S. I. (1981). Homogeneous victim–offender populations: A review and some re- search implications.The Journal of Criminal Law and Criminology,72, 779–788.

Singer, S. I. (1987). Victims in a birth cohort. In T. P. T. Marvin E. Wolfgang, & Robert M.

Figlio (Eds.),From boy to man, from delinquency to crime(pp. 163–179). Chicago, IL:

University of Chicago Press.

Smith, D. J., & Ecob, R. (2007). An investigation into causal links between victimization and offending in adolescents.The British Journal of Sociology,58, 633–659.

Sparks, R. F., Glen, A. G., & Dodd, D. J. (1977).Surveying victims.New York: Wiley.

Stewart, C., & Power, T. G. (2002). Identifying patterns of adolescent drinking: A tri- ethnic study.Journal of Studies on Alcohol,63(2), 156 –168.

Stewart, E. A., Schreck, C. J., & Simons, R. (2006).“I ain't gonna let no one disrespect me”: Does the code of the street reduce or increase violent victimization among Af- rican American adolescents?Journal of Research in Crime and Delinquency,43, 427–458.

Swanson, J. W., Borum, R., Swartz, M., & Hiday, V. (1999). Violent behavior preceding hospitalization among persons with severe mental illness.Law and Human Behav- ior,23, 185–204.

Taylor, T. J., Freng, A., Esbensen, F. -A., & Peterson, D. (2008). Youth gang membership and serious violent victimization: The importance of lifestyles and routine activi- ties.Journal of Interpersonal Violence,23, 1441–1464.

Tewksbury, R., & Mustaine, E. E. (2000). Routine activities and vandalism: A theoretical and empirical study.Journal of Crime and Justice,23,81–110.

Van Dijk, J. J., & Steinmetz, C. H. (1983). Victimization surveys: Beyond measuring the volume of crime.Victimology,8, 291–309.

Vega, W. A., Rodriguez, M. A., & Gruskin, E. (2009). Health disparities in Latino popula- tions.Epidemiologic Review,31,99–112.

Widom, C. (1989). Does violence beget violence? A critical examination of the litera- ture.Psychological Bulletin,106,3–28.

Wittebrood, K., & Nieuwbeerta, P. (1999). Wages of sin? The link between offending, lifestyle and violent victimization.European Journal of Criminal Policy and Research, 7,63–80.

Wolfgang, M. E. (1958).Patterns in criminal homicide.Oxford, England: University of Pennsylvania Press. 26W.G. Jennings et al. / Aggression and Violent Behavior 17 (2012) 16–26