Read the following articles on couples treatment: Intimate Partner Violence: Making the Case for Joint Couple Treatment and Intimate Partner Violence and the Victim-Offender Overlap. In your short paper, address the following elements: Evaluate the effica

Article

Intimate Partner Violence and the Victim-Offender Overlap

Marie Skubak Tillyer1 and Emily M. Wright2

Abstract Objectives: Examine the prevalence and correlates of intimate partner vio- lence (IPV) victimization and offending, as well as the overlap of these experiences. Method: Data from wave 4 of the National Longitudinal Study of Adolescent Health were analyzed to examine IPV among adults ages 24 to 33. A multinomial logistic regression model was estimated to determine whether the correlates of IPV vary across victims, perpetrators, and victim- perpetrators. Results: Approximately 20% of respondents reported some IPV involvement in the past year, one-third of whom reported victimization and perpetration. The victim-offender overlap was observed for males and females across various measures of IPV. Bivariate correlations suggest victimization and perpetration have common correlates. Multivariate analysis, however, reveals considerable differences once we distinguish between victims, offen- ders, and victim-offenders and control for other variables. Perpetrators and victim-perpetrators were more likely to live with a nonspouse partner; feel isolated; display negative temperaments; and report substance use problems.

1 Department of Criminal Justice, University of Texas at San Antonio, TX, USA 2 University of Nebraska at Omaha, NE, USA

Corresponding Author:

Marie Skubak Tillyer, Department of Criminal Justice, University of Texas at San Antonio,

501 W. Cesar E. Chavez Blvd., San Antonio, TX 78207, USA.

Email: [email protected]

Journal of Research in Crime and Delinquency

2014, Vol 51(1) 29-55 ª The Author(s) 2013

Reprints and permission: sagepub.com/journalsPermissions.nav

DOI: 10.1177/0022427813484315 jrcd.sagepub.com

‘‘Victims only’’ were more likely to live with children and have lower household incomes. Conclusions: The victim-offender overlap exists for IPV across a variety of measures. Though perpetrators and victim-perpetrators have similar characteristics, those who are victims only appear distinctly differ- ent. We discuss the implications for theory, policy, and research.

Keywords victim-offender overlap, intimate partner violence

Strangers do not always present the greatest threat of criminal perpetra-

tion—in fact, we are more likely to be victimized in our own homes by our

loved ones rather than attacked by strangers on the street (Straus, Gelles,

and Steinmetz 2006). Estimates from the National Violence against Women

and the National Intimate Partner and Sexual Violence surveys report that

up to 36 percent of females and 29 percent of males have experienced some

form of violence by their intimate partner during their lifetime (Black et al.

2011; Tjaden and Thoennes 2000). Further, recent data show that while the

majority of cases of intimate partner violence (IPV) processed by state

courts involve a male defendant and a female victim, approximately 12

percent involve a female offender and a male victim (Smith and Farole

2009). These figures challenge traditional stereotypes about the gender of

IPV victims and offenders (e.g., Straus 2011) and raise the question of

whether those involved in IPV can be neatly categorized as ‘‘victims’’ and

‘‘offenders.’’

Research on criminal victimization and offending more generally

demonstrates that victims and perpetrators of crime are not necessarily

distinct groups. Rather, there is considerable overlap in these populations

(Jennings, Piquero, and Reingle 2012; Lauritsen and Laub 2007). Several

explanations have been offered for this phenomenon, including the notion

that victims and offenders share common routine activities and similar traits

that create opportunities for both criminal victimization and offending.

There has been little empirical attention paid, however, to whether the

victim-offender overlap exists among those involved in IPV. This omission

in the research may be due in part to long-held beliefs that IPV is an expres-

sion of male domination over women (see the feminist perspective on

domestic violence, e.g., Dobash and Dobash 1979; Dobash et al. 1992;

Lawson 2012); individuals, therefore, are assumed to be victims or perpe-

trators of IPV.

30 Journal of Research in Crime and Delinquency 51(1)

Yet, if we consider that at least some instances of IPV might be better

understood as conflicts that are not necessarily used for domination and

control purposes but which nonetheless occur within intimate settings,

an investigation of the victim-offender overlap with respect to IPV

becomes more tenable. The family violence perspective, for instance, sees

‘‘conflict between family members as universal and inevitable, and vio-

lence between any family members (including violence between spouses)

is viewed as one method utilized by those members to resolve this predict-

able conflict’’ (Lawson 2012:575). Though a family violence perspective

hints that there might be considerable victim-offender overlap with

respect to IPV, this issue has received little empirical attention by

criminologists.

Given that a substantial body of research has established the victim-

offender overlap in general (Berg et al. 2012; Jennings et al. 2010; Jen-

nings, Piquero, and Reingle 2012; Lauritsen and Laub 2007; Lauritsen,

Sampson, and Laub 1991), violence by loved ones is a somewhat common

occurrence (Black et al. 2011; Straus, Gelles, and Steinmetz 2006), and

some violence between partners has been characterized as ‘‘mutual’’

(Johnson 1995; Straus, Gelles, and Steinmetz 2006), the existence of the

victim-offender overlap within IPV seems plausible and warrants investi-

gation. Further, the very nature of intimate partners—that is, individuals

who have self-selected into relationships that routinely bring them into

contact with one another (Carbone-Lopez and Kruttschnitt 2010)—sug-

gests that an initial act of violence may be followed by subsequent vio-

lence: The conditions that led to the initial violence may remain

unresolved and/or the violence itself might produce opportunities and

motivations for additional fights. Because much empirical research on

IPV focuses on female victimization or male offending exclusively, it is

also unclear whether the established correlates of IPV apply to all types

of involvement in IPV, or if there are important differences that distin-

guish those who are both victims and offenders of IPV from those who are

involved in IPV solely as victims or offenders.

We aim to fill these voids by exploring the prevalence of IPV victimiza-

tion and perpetration in adulthood, as well as the overlap of these experi-

ences, in a nationally representative sample of adults ages 25 to 33. To

this end, we draw on the existing research in the areas of IPV and the

victim-offender overlap to examine the extent to which the correlates for

IPV vary across victims, perpetrators, and victim-perpetrators using multi-

nomial logistic regression analysis. Our findings suggest a number of impli-

cations for theory, policy, and future research.

Tillyer and Wright 31

IPV

The existing research on IPV suggests that victims and offenders have many

common characteristics. Like other forms of violence, IPV is inversely

related to age (Rennison and Welchans 2000). Young couples are less likely

to be married or to have been in a relationship for a long period of time, and

it has been suggested that they may lack the skills and experience needed to

successfully resolve arguments and reach compromise in conflicts and may

not yet understand each other’s boundaries for acceptable behavior

(DeMaris et al. 2003). Younger people are also generally more violent and

aggressive than older people (Hirschi and Gottfredson 1983), perhaps

explaining why younger couples are more volatile in their relationships.

Minorities also appear to engage in or report IPV more often than non-

minorities (Rennison and Welchans 2000). This may be due to cultural dif-

ferences in the meaning of partner violence and/or stress and frustration that

arise when they are faced with limited opportunities for upward mobility

due to their position in society (Cloward 1959; Gelles and Straus 1988).

Likewise, couples in the lower socioeconomic strata, as indicated by low

educational or occupational attainment or low income, may experience

more stressors arising from financial difficulties, or may experience more

frustration due to limited opportunities—all of which might increase the

likelihood of IPV occurring (Holtzworth-Munroe, Smutzler, and Bates

1997). Individuals with low education or occupational attainment may also

be more prone to resort to violence when verbal discussion fails them

(Gottfredson and Hirschi 1990), or may lack effective negotiation skills

(Feldman and Ridley 2000; Ridley and Feldman 2003).

Violence within relationships has also been linked to other problems,

particularly alcohol and drug use, as well as anger and hostile attitudes. IPV

is approximately three times more likely to occur when drug use is involved

(Moore et al. 2008). It has been suggested that substance use releases inhi-

bitions regarding the use of violence against one’s partner or that it is used

as an excuse to justify behaviors (such as IPV) that are normally considered

unacceptable (Kaufman Kantor and Straus 1987). Furthermore, substance

use may be used as a coping mechanism for stress and frustration, poor

relationship quality, or may be used among victims as a result of partner

violence (Kilpatrick et al. 1997). Attitudes condoning violence in general,

and against partners specifically, are positively associated with IPV, as are

feelings of anger, hostility, and negative emotionality (Moffitt et al. 2000;

Norlander and Eckhardt 2005; Stith et al. 2004). It may be that persons will-

ing to use violence in general are more apt to use violence within their

32 Journal of Research in Crime and Delinquency 51(1)

relationships as well or that persons unable to regulate their emotions are

more likely to lash out at others, even intimate others (Norlander and

Eckhardt 2005; Sugarman and Frankel 1996).

Couples who are cohabiting but who are not married are more likely to

experience IPV (DeMaris et al. 2003; Stets 1991); scholars have suggested

that this relationship may reflect lower commitment levels between the

partners (Stets 1991) or that unmarried cohabitating couples are more likely

to be younger and therefore more violent (Magdol et al. 1998). Violence

between partners occurs most often inside of the couple’s (or victim’s)

house (Rennison and Welchans 2000; Thomas, Dichter, and Matejkowski

2011), with few or no witnesses of the violence, with the exception of young

children living in the household (Holt, Buckley, and Whelan 2008). Theo-

retically, however, the presence of adult third parties may deter violence or

lessen the severity of abuse through shaming, other control mechanisms

(Stets 1991; Van Wyk et al. 2003; Wilkinson and Hamerschlag 2005), or

by increasing the likelihood that victims reach out to others for help

(Browning 2002; Coker et al. 2002). When the third parties are minor

children, however, it is possible that their presence may add to relationship

stress and actually increase violence (Wilkinson and Hamerschlag 2005).

Finally, social support, and its antithesis, social isolation, have been sug-

gested as important factors that may inhibit or increase IPV. Social support

may be related to a reduced risk of victimization because supportive rela-

tionships can provide the opportunities, financial means, or physical alter-

natives needed to escape a violent relationship. For instance, support may

increase the likelihood that the victim seeks and receives help from others

(e.g., Browning 2002; Cullen 1994), since they have people to turn to for

emotional, physical, or financial assistance (Van Wyk et al. 2003). Social

isolation, on the other hand, is related to increased risk of violence within

partnerships. It has been suggested that social isolation may lead to less

support and control within relationships, which in turn may lead to more

aggression (Stets 1991), or that social isolation keeps violence within rela-

tionships private, while also increasing dependence between the partners

(Van Wyk et al. 2003). Further, social isolation may leave victims with few

resources with which to leave their violent partners (MacMillian and

Gartner 1999; Van Wyk et al. 2003).

In sum, many consistencies exist between the predictors of IPV perpetra-

tion and victimization. As noted above, studies indicate that 29 percent of

men and 36 percent of women report having experienced some form of

violence by their intimate partner during their lifetime (Black et al. 2011;

Tjaden and Thoennes 2000), which suggests that IPV is not limited to male

Tillyer and Wright 33

perpetration and female victimization. The vast majority of the IPV

research, however, examines the predictors of IPV perpetration and victimi-

zation separately, often relying on gendered analyses (e.g., male-only per-

petrators or female-only victims, see Bachman and Saltzman 1995; Tjaden

and Thoennes 1998) that may not fully capture differences across victims,

perpetrators, and victim-perpetrators. We now turn to the literature on the

victim-offender overlap to examine possible explanations for the phenom-

enon and how it might apply to IPV.

The Victim-Offender Overlap

Research has demonstrated a substantial and consistent overlap among

those who have been victimized and those who have perpetrated offenses

against others (Jennings, Piquero, and Reingle 2012; Lauritsen and Laub

2007). Studies examining the victim-offender overlap indicate that offend-

ing increases the likelihood of victimization, but also that victimization

increases the likelihood of offending (Lauritsen and Laub 2007; Lauritsen,

Sampson, and Laub 1991). The victim-offender relationship is robust,

having been found in the United States as well as other countries, over time,

across various contexts, and within various demographic subgroups. Fur-

ther, this relationship has been observed across numerous victimization and

offending measures, and despite the use of a variety of control variables

(Jennings, Piquero, and Reingle 2012; Lauritsen and Laub 2007). Most of

this research has focused on crimes such as violence, delinquency, property

crimes, and arrests, but the greatest overlap appears to be among the most

severe crimes, particularly homicide (Jennings, Piquero, and Reingle 2012;

Lauritsen and Laub 2007).

Although the victim-offender overlap is one of the most consistent find-

ings in criminology (Lauritsen and Laub 2007), the theoretical mechanisms

underlying this relationship are somewhat less agreed upon. It is possible

that an underlying trait—such as impulsivity, risk taking, or negative emo-

tionality—conducive to both victimization and offending accounts for the

overlap (Lauritsen and Davis Quinet 1995; Lauritsen and Laub 2007). The

‘‘principle of homogamy’’ (e.g., Cohen, Kluegel, and Land 1981; Lauritsen,

Sampson, and Laub 1991) implies that persons are more likely to be victi-

mized when they come into contact with other offenders and that offenders

are more likely to become victims because of their disproportionate contact

with offenders. It is also feasible that victimization and offending may

change an individual’s future risk of certain behaviors, such as subsequent

victimization or offending (Lauritsen and Davis Quinet 1995; Lauritsen and

34 Journal of Research in Crime and Delinquency 51(1)

Laub 2007). Indeed, much literature suggests that the relationship between

victimization and offending is due to common risk factors shared between

these two groups (Jennings et al. 2010; Jennings, Piquero, and Reingle

2012). Consistent with these ideas, Jennings, Piquero, and Reingle

(2012:17), in their comprehensive review of the research, contend that

routine activities theory is ‘‘by far the most recognizable’’ of theoretical

perspectives that attempt to explain the victim-offender overlap, in that ‘‘the

theory centers on the influence that opportunity structures and risky life-

styles have on increasing the likelihood for committing an offense or

experiencing victimization.’’ Mustaine and Tewksbury (2000), for exam-

ple, used a lifestyles-routine activities perspective to examine similarities

and differences across college students who participated in assaults as

victims, offenders, and victim-offenders. Their analyses indicate that these

are distinct groups with different sources of criminal involvement, leading

them to conclude that ‘‘routine activity theory rationales offer more useful

explanations for the assault risks of the more heavily criminally involved

individual (those who are both victim and offender) over the individuals

who are involved only as a victim or an offender’’ (Mustaine and Tewksbury

2000:356).

Intimate Partners as Offenders and Victims

Despite the considerable attention that has been devoted to the victim-

offender overlap, few studies have examined this phenomenon among

couples in intimate relationships. To be sure, the ‘‘cycle of violence,’’ or the

idea that victimization within family relationships might beget offending

later in life, has been discussed and empirically examined (e.g., Lussier,

Farringon, and Moffitt 2009; Widom 1989). There are, however, few

empirical investigations of the overlap of victimization and offending

within an intimate relationship. As suggested above, this may be due in part

to the influence of the feminist perspective on the study of IPV. If IPV is

assumed to be the result of male domination over women, it makes little

sense to expect a victim-offender overlap.

It is also possible that this criminological phenomenon has rarely been

explored within IPV because IPV and IPV offenders are viewed differently

than typical ‘‘street’’ offenders and the crimes they commit. Though some

research suggests IPV offenders are similar to non-IPV offenders (Felson

and Lane 2010), there is also evidence that the two groups differ on impor-

tant characteristics (Moffitt et al. 2000). For example, violence within rela-

tionships involves intimate knowledge of the victim/offender and emotions

Tillyer and Wright 35

that may not typically be present in stranger crime (Dugan and Apel 2005),

shared characteristics (e.g., routines, violent proclivities, behavioral prefer-

ences) resulting from self-selection of partners (Carbone-Lopez and

Kruttschnitt 2010; Wright et al. 1999), and potentially different motivations

for violence against one’s partner as opposed to violence against strangers

(e.g., control, see Felson and Messner 2000; rage, see Thomas, Dichter, and

Matejkowski 2011).

Interestingly, however, it is the very nature of IPV—that is, violence

between two individuals who are in a relationship and interact on a routine

basis—that makes it particularly well suited for an investigation of the

victim-offender overlap. Most explanations of the victim-offender overlap

rely on common traits and routine activities that contribute to opportunities

for crime (Jennings, Piquero, and Reingle 2012; Lauritsen and Laub 2007;

Mustaine and Tewksbury 2000). In the case of intimate partners, individuals

have self-selected into the relationship, possibly because of their shared

traits (Carbone-Lopez and Kruttschnitt 2010; Wright et al. 1999), and the

relationship repeatedly brings potential victims and perpetrators in close

proximity to one another. General discussions of the ‘‘victim-offender’’

overlap typically refer to an individual whose victim and perpetrator are not

likely the same person (e.g., a drug dealer who is the victim of a robbery).

When we consider the victim-offender overlap within the context of IPV,

however, we are essentially investigating mutual violence between only two

individuals, where the possibility of becoming a victim and an offender

exists within a single relationship.

There are a number of explanations for why we might expect to see an

overlap of victims and offenders in intimate relationships. First, the initial

circumstances that produced the violence, if left unresolved, can lead to

additional fights. These circumstances might be external factors that put

stress on the relationship or they could be common traits that the partners

share that make conflict resolution difficult. This explanation for recurrent

and mutual violence is consistent with the principal of homogamy, in that

common traits between the offenders and victims increase the likelihood

that offenders will be victimized. Such a possibility is supported by the

many shared risk factors for IPV perpetration and victimization reviewed

above.

Second, the violence itself can increase the likelihood of additional vio-

lence. The initial crime event—in this case, the violence used against one’s

partner—has the potential to create motivation and opportunity for subse-

quent violence, either during the same incident or in future fights. Not only

might the initial victim feel justified in using violence (e.g., as in the case of

36 Journal of Research in Crime and Delinquency 51(1)

violent resistant partner violence, see Johnson 2008, or in self-defense

efforts), but the initial perpetrator may be left with little recourse—in terms

of turning to others for help—given his or her own perpetration. Thus, it is

plausible that the victim-offender overlap applies to violence that occurs

between intimate partners. The current study explores this possibility and

examines whether the existing correlates of IPV apply to all types of IPV

involvement.

The Present Study

The present study examines the prevalence of IPV victimization and perpe-

tration in adulthood, as well as the overlap of these experiences. Given that

the victim-offender overlap has been observed in numerous studies measur-

ing general offending and victimization, and that most assaults occur

between parties that are acquainted, we argue that such an overlap is plau-

sible. Further, the very nature of IPV—that is, violence between two people

who have self-selected into the relationship and are routinely in contact with

another—presents ongoing opportunities and motivations for future vio-

lence once an initial act of violence has occurred. We draw upon the exist-

ing empirical research on IPV to explore whether the correlates for IPV vary

across victimization, perpetration, and victimization-perpetration preva-

lence. Specifically, the present study aims to answer the following research

questions:

1. What proportion of individuals involved in IPV are victims only,

offenders only, and victim-offenders?

2. Do the established correlates of IPV apply to victims only, offenders

only, and victim-offenders?

Data

The present study uses public-use data from the fourth wave of the National

Longitudinal Study of Adolescent Health (Add Health) to examine the pre-

valence and correlates of IPV victimization and perpetration. Wave 1 data

were collected from a nationally representative sample of adolescents

enrolled in American middle and high schools (Harris et al. 2009). Systema-

tic and stratified sampling techniques were used to select a sample of 80

high schools and 52 middle schools. Students in grades 7 through 12 during

the 1994–1995 school year completed the wave 1 in-school self-report

survey. Wave 1 of the public-use Add Health data includes 6,504 cases.

Tillyer and Wright 37

Wave 4 of the Add Health was designed to examine the developmental and

health trajectories of the Add Health respondents into young adulthood.

Approximately 92.5% of the original wave 1 respondents were located and 80.3% of eligible individuals were interviewed in 2008 (Harris et al. 2009). This resulted in data from 5,114 respondents ages 24 to 33.

1

Because the present study focuses on IPV in heterosexual romantic

relationships, individuals who described themselves as ‘‘bi-sexual’’ or

‘‘homosexual’’ were removed from the sample, resulting in 4,926 cases.

In addition, we removed those who reported being active duty military and

those who were interviewed in prison during wave 4, given that these indi-

viduals may not have had the opportunity to experience IPV in the past year.

This reduced the sample size to 4,795 cases. Listwise deletion of cases

based on missing data for study variables resulted in 4,275 cases for anal-

ysis. The descriptive and inferential statistics presented below use normal-

ized weights to account for the Add Health sampling design. 2

Measures

The dependent variable examined in this study measures the prevalence of

IPV involvement, both as a victim and as a perpetrator. To create our four-

category nominal outcome, we first created two dichotomous variables to

measure IPV victimization and perpetration prevalence. IPV victim is a

dichotomous variable that measures whether the respondent reported

experiencing any of three events in the past year: (1) partner threatened you

with violence, pushed or shoved you, or threw something at you that could

hurt, (2) partner slapped, hit, or kicked you, and (3) you had an injury, such

as a sprain, bruise, or cut because of a fight with your partner. If the respon-

dent reported experiencing any of these events in the past year, IPV victim

was coded as 1. IPV perpetrator is a dichotomous variable that measures

whether the respondent reported engaging in any of three events in the past

year: (1) threatened partner with violence, pushed or shoved partner, or

threw something at partner that could hurt, (2) slapped, hit, or kicked

partner, and (3) partner had an injury such as a sprain, bruise, or cut because

of a fight with you. If the respondent reported experiencing any of these

events in the past year, IPV perpetrator was coded as a 1. IPV involvement

is a four-category nominal variable created from the IPV victim and IPV

perpetrator variables. Respondents were coded as no involvement, victim

only, perpetrator only, or victim and perpetrator.

We operationalized a series of variables that prior IPV studies have

established as correlates of IPV victimization and/or offending. We created

38 Journal of Research in Crime and Delinquency 51(1)

several variables to measure the age, race, and gender of the respondents.

Age is the respondent’s age in years at the wave 4 interview. We created

three dummy variables to measure race and ethnicity—Black, Hispanic, and

Other, with White Non-Hispanic being the omitted reference category.

Male (0 ¼ No; 1 ¼ Yes) measures the respondent’s gender. We also created three variables to measure the socioeconomic status of

the respondents. Education is an ordinal variable that measures the educa-

tional attainment of the respondent (1 ¼ Less than high school diploma, 2 ¼ High school graduate, 3 ¼ Some college or vocational/technical training, 4 ¼ College degree). Employed is a dichotomous variable that measures whether the respondent works for pay for a minimum of 10 hr per week

(0 ¼ No; 1 ¼ Yes). Household income measures the respondent’s total household income before taxes and deductions, with values ranging from

1 ¼ Less than $5,000 to 12 ¼ $150,000 or more. Substance use is a dichotomous variable that measures whether respon-

dents report their drug or alcohol use has caused problems (i.e., problems

with work, school, family, or friends; legal troubles; put them or others at

risk; 0 ¼ No; 1 ¼ Yes). In addition, we created a factor to tap the degree to which the respondent reports feelings of stress, frustration, and irritabil-

ity. Eight survey items that asked respondents the extent to which they

agreed with various statements were used to create the negative tempera-

ment factor, with responses ranging from strongly agree to strongly

disagree. Specifically, respondents were asked the extent to which they

agreed with statements that they have frequent mood swings, get angry eas-

ily, get upset easily, get stressed out easily, lose their temper, are not easily

bothered by things, rarely get irritated, and keep their cool. 3

Factor scores

were created using principal components factor analysis with varimax

rotation. All items loaded on a single factor, with factor loadings ranging

from 0.64 to 0.83 (Eigenvalue ¼ 4.20, a ¼ .87). We also created a series of variables to measure the composition of peo-

ple with whom the respondent resides. Prior IPV research typically distin-

guishes between couples who live together but are not married, and those

who live together and are married, given speculation that unmarried couples

are less committed to each other (Stets 1991) or that unmarried cohabitating

couples are more likely to be younger and therefore more violent (Magdol

et al. 1998). Therefore, we distinguish between these two different types of

relationships by measuring lives with non-spouse partner (0 ¼ No; 1 ¼ Yes) and lives with spouse (0 ¼ No; 1 ¼ Yes).

Lives with adults indicates whether the respondent lives with one or more

nonromantic partner adults (0 ¼ No; 1 ¼ Yes). Lives with children indicates

Tillyer and Wright 39

whether the respondent lives with one or more individuals under the age of

18 (0 ¼ No; 1 ¼ Yes). Beyond living arrangements, we also included variables to measure the

respondent’s social relationships that might prevent IPV victimization and/

or perpetration. Attachment to friends measures the number of people the

respondent reports feeling at ease with, can talk to about private matters,

and can call on for help. Responses ranged from 1 ¼ None to 5 ¼ 10 or more friends. Attachment to parents is the average response to two questions ask-

ing respondents how close they feel to their mother figure and how close

they feel to their father figure (1 ¼ Not at all close, 2 ¼ Not very close, 3 ¼ Somewhat close, 4 ¼ Quite close, and 5 ¼ Very close). Finally, isolated measures how often the respondent feels isolated from others (0 ¼ Never, 1 ¼ Rarely, 2 ¼ Sometimes, 3 ¼ Often). Diagnostics indicate that multicol- linearity is not a concern among the independent variables (variance

inflation factors [VIFs] � 2.06).

Analysis

We begin by examining the frequency of the IPV involvement dependent

variable to assess the overlap in IPV victimization and perpetration. Recall

that our measure of IPV prevalence includes a range of behaviors, including

threats, pushing and shoving, and throwing things that could hurt. Though

our multinomial regression analysis focuses on this more inclusive defini-

tion of IPV, we also examine—by gender—the distributions of the items

used to create our measure of IPV involvement. This allows us to observe

whether the overlap exists for both men and women across all levels of IPV

prevalence severity. Next, we present the bivariate correlations among the

independent variables and IPV perpetrator and IPV victim.

We use multinomial regression to examine the correlates of IPV invol-

vement—either as a victim, perpetrator, or both—controlling for the other

independent variables in the analysis. Multinomial regression was selected

because it allows for categorical dependent variables, such as our four-

category measure of IPV involvement. The analysis produces one set of

coefficients for each category of the dependent variable, minus one for the

omitted reference category (Pampel 2000). For our purposes, ‘‘no involve-

ment’’ in IPV as a victim or perpetrator is the omitted reference category.

Each set of coefficients can be interpreted as the effect of one unit change

in the independent variable on the logged odds of falling into each category

of the dependent variable, relative to the ‘‘no involvement’’ in IPV (i.e., the

omitted reference category of the dependent variable; Pampel 2000).

40 Journal of Research in Crime and Delinquency 51(1)

Findings

Table 1 presents the distribution of IPV among respondents. The frequency

of the dependent variable, IPV involvement, reveals considerable overlap in

terms of IPV victimization and offending. Nearly 80% of respondents reported no involvement in IPV in the past year, while 9.7% identified as an IPV victim only and 3.8% identified as a perpetrator only. The remain- der—6.7%—reported experiencing IPV as both a victim and a perpetrator in the prior year. The remaining columns in Table 1 reveal that even when

we adopt an increasingly narrow definition of IPV, the victim-offender

overlap is still observed, and observed for both genders. These findings

challenge the assertion that females are strictly victims of IPV and any

violence against their partners is purely in self-defense. A higher percentage

of women, relative to men, reported being ‘‘perpetrators only,’’ across all

definitions of IPV. Conversely, a higher percentage of men, relative to

women, reported being ‘‘victims only.’’

Means and standard deviations for all independent variables are pre-

sented in Table 2, as well as the bivariate correlations among the inde-

pendent variables and IPV perpetrator and IPV victim (recall that these

latter two variables do not account for the victim-offender overlap and

are not used in the multivariate analysis). As expected, there is a posi-

tive and statistically significant (p � .01) bivariate correlation between IPV perpetration and victimization. Furthermore, each of these out-

comes maintains statistically significant relationships at the bivariate

level with many of the same independent variables. For example, Black,

Hispanic, substance use, negative temperament, lives with non-spouse

partner, lives with children, and isolated are all positively correlated

with both IPV perpetration and victimization at the bivariate level. Con-

versely, education, household income, lives with spouse, attachment to

friends, and attachment to parents are all negatively associated with

both IPV perpetration and victimization. In addition, IPV perpetration

maintained a negative relationship with male and employed, and a pos-

itive relationship with other race; IPV victimization is positively asso-

ciated with male at the bivariate level.

Table 3 displays the multinomial logistic regression results for IPV

involvement in the past year. Recall that ‘‘no involvement’’ in IPV is the

omitted reference category for the dependent variable. Respondents who

live with children, are male, and are Black were more likely to identify

as a ‘‘victim only’’ of IPV relative to ‘‘no involvement’’ in IPV. Conversely,

older respondents, those with higher levels of education, and those who

Tillyer and Wright 41

T a b

le 1 .

In ti m

at e

P ar

tn e r

V io

le n ce

In vo

lv e m

e n t.

a

IP V

In vo

lv e m

e n t

IP V

In vo

lv e m

e n t

‘‘T h re

at e n e d ,

P u sh

e d /S

h o ve

d ,

T h ro

w n

S o m

e th

in g’

’ ‘‘S

la p p e d , H

it ,

K ic

k e d ’’

‘‘I n ju

ry ’’

A ll

R e sp

o n d e n ts

(% )

F e m

al e s

(% )

M al

e s

(% )

F e m

al e s

(% )

M al

e s

(% )

F e m

al e s

(% )

M al

e s

(% )

F e m

al e s

(% )

M al

e s

(% )

N o

in vo

lv e m

e n t

7 9 .9

8 0 .4

7 9 .4

8 1 .9

8 2 .0

8 9 .3

8 7 .1

9 5 .7

9 5 .6

V ic

ti m

o n ly

9 .7

5 .7

1 3 .8

6 .1

1 1 .7

2 .0

9 .9

2 .0

2 .9

O ff e n d e r

o n ly

3 .8

6 .3

1 .1

5 .2

1 .2

5 .4

0 .4

1 .1

0 .6

V ic

ti m

-O ff e n d e r

6 .6

7 .6

5 .7

6 .8

5 .1

3 .3

2 .6

1 .1

0 .9

N o te

: P e rc

e n ta

ge s

d o

n o t

su m

to 1 0 0

d u e

to ro

u n d in

g. a T

h e

p e rc

e n ta

ge s

ar e

b as

e d

o n

th e

w e ig

h te

d d at

a.

42

T a b

le 2 .

B iv

ar ia

te C

o rr

e la

ti o n s.

a

M (S

D )

1 2

3 4

5 6

7 8

9 1 0

1 1

1 2

1 3

1 4

1 5

1 6

1 7

1 8

1 9

(1 )

IP V

p e rp

e tr

at o r

.1 0

(. 3 1 )

1 .0

0

(2 )

IP V

vi ct

im .1

6 (. 3 7 )

.4 4 **

1 .0

0 (3

) A

ge 2 8 .4

3 (1

.8 4 ) �

.0 2 �

.0 2

1 .0

0 (4

) B

la ck

.1 4

(. 3 5 )

.0 8 **

.1 2 **

.0 5 **

1 .0

0 (5

) H

is p an

ic .1

1 (. 3 1 )

.0 5 **

.0 3 * �

.0 2 �

.1 4 **

1 .0

0 (6

) O

th e r

.0 7

(. 2 5 )

.0 4 *

0 .0

2 .0

1 �

.1 1 ** �

.0 9 **

1 .0

0 (7

) M

al e

.4 9

(. 5 0 ) �

.1 2 **

.0 8 **

.0 7 ** �

.0 2

.0 0

.0 0

1 .0

0 (8

) E d u ca

ti o n

3 .0

1 (. 8 9 ) �

.0 8 ** �

.1 4 **

.0 0 �

.0 8 ** �

.0 9 **

.0 0 �

.1 3 **

1 .0

0 (9

) E m

p lo

ye d

.8 4

(. 3 7 ) �

.0 4 * �

.0 2 �

.0 1 �

.0 4 *

.0 3 * �

.0 1

.1 5 **

.1 1 **

1 .0

0 (1

0 )

H H

in co

m e

8 .0

8 (2

.5 8 ) �

.0 7 ** �

.1 3 **

.0 4 ** �

.2 0 **

.0 4 *

.0 4 *

.0 8 **

.3 3 **

.2 4 **

1 .0

0 (1

1 )

Su b st

an ce

u se

.3 9

(. 4 9 )

.0 8 **

.0 4 * �

.0 5 ** �

.1 5 ** �

.0 5 **

.0 1

.1 4 **

.0 2

.0 4 *

.0 4 *

1 .0

0

(1 2 )

N e ga

ti ve

te m

p .

.0 0

(1 .0

0 )

.2 0 **

.0 9 **

.0 0 �

.0 2

.0 1

.0 1 �

.1 7 ** �

.1 4 ** �

.0 7 ** �

.1 2 **

.0 4 *

1 .0

0

(1 3 )

L iv

e w

/ p ar

tn e r

.1 9

(. 3 9 )

.0 7 **

.0 7 ** �

.0 8 **

.0 0

.0 1

.0 4 *

.0 2 �

.0 8 **

.0 2 �

.0 3

.0 5 **

.0 1

1 .0

0

(1 4 )

L iv

e w

/ sp

o u se

.4 6

(. 5 0 ) �

.0 4 ** �

.0 8 **

.1 4 ** �

.1 6 ** �

.0 1 �

.0 2 �

.0 8 **

.0 5 ** �

.0 3

.2 3 ** �

.1 3 **

.0 2 �

.4 4 **

1 .0

0

(1 5 )

L iv

e w

/ ad

u lt s

.2 7

(. 4 4 )

.0 2

.0 3 �

.1 0 **

.0 7 **

.0 9 **

.0 8 **

.1 1 ** �

.1 3 ** �

.0 4 *

.0 3

.0 6 **

.0 1 �

.1 1 ** �

.3 5 **

1 .0

0 (1

6 )

L iv

e w

/ ch

ild re

n .5

3 (. 5 0 )

.0 5 **

.0 4 *

.1 4 **

.0 7 **

.0 5 **

.0 0 �

.1 9 ** �

.2 4 ** �

.1 2 ** �

.0 7 ** �

.1 4 **

.1 1 ** �

.0 2

.3 2 ** �

.1 2 **

1 .0

0

(1 7 )

Fr ie

n d s

3 .1

8 (. 9 9 ) �

.0 7 ** �

.0 7 ** �

.0 2 �

.1 4 ** �

.0 6 **

.0 0

.0 7 **

.2 2 **

.1 0 **

.1 6 **

.0 9 ** �

.1 3 ** �

.0 2 �

.0 4 **

.0 1 �

.1 8 **

1 .0

0 (1

8 )

P ar

e n ts

3 .9

8 (1

.0 4 ) �

.0 8 ** �

.0 7 ** �

.0 7 ** �

.0 9 **

.0 1 �

.0 5 **

.0 3

.1 4 **

.0 3 *

.1 4 ** �

.0 4 * �

.1 1 ** �

.0 3 *

.0 6 ** �

.0 1 �

.0 6 **

.1 1 **

1 .0

0 (1

9 )

Is o la

te d

.9 2

(. 9 0 )

.1 4 **

.0 9 **

.0 1

.0 2 �

.0 3 *

.0 5 ** �

.0 3

.0 2 �

.1 2 ** �

.1 5 **

.1 2 **

.2 6 ** �

.0 4 * �

.1 0 **

.0 5 ** �

.0 6 ** �

.1 5 ** �

.1 8 **

1 .0

0

IP V ¼

in ti m

at e

p ar

tn e r

vi o le

n ce

; H

H ¼

h o u se

h o ld

. a T

h e

d e sc

ri p ti ve

st at

is ti cs

an d

b iv

ar ia

te co

rr e la

ti o n s

ar e

b as

e d

o n

th e

w e ig

h te

d d at

a. *p �

.0 5 . **

p �

.0 1 .

43

T a b

le 3 .

M u lt in

o m

ia l L o gi

st ic

R e gr

e ss

io n

fo r

IP V

In vo

lv e m

e n t.

a ,b

,c

V ic

ti m

o n ly

P e rp

e tr

at o r

o n ly

V ic

ti m

an d

P e rp

e tr

at o r

b S E

E x p (B

) b

S E

E x p (B

) b

S E

E x p (B

)

In te

rc e p t

0 .7

3 0 .9

1 —

� 0 .2

0 1 .4

6 —

� 2 .0

1 1 .1

0 —

A ge

� 0 .0

8 *

0 .0

3 1 .0

8 �

0 .1

1 *

0 .0

5 1 .1

2 �

0 .0

1 0 .0

4 —

B la

ck 0 .7

0 **

0 .1

5 2 .0

2 1 .1

9 **

0 .2

4 3 .3

0 1 .0

2 **

0 .1

8 2 .7

7 H

is p an

ic 0 .2

7 0 .1

8 —

0 .6

4 **

0 .2

5 1 .8

9 0 .7

5 **

0 .2

0 2 .1

2 O

th e r

� 0 .0

5 0 .2

4 —

0 .0

8 0 .3

6 —

0 .8

1 **

0 .2

2 2 .2

4 M

al e

1 .0

0 **

0 .1

2 2 .7

1 �

1 .7

8 **

0 .2

4 5 .8

8 �

0 .2

1 0 .1

4 —

E d u ca

ti o n

� 0 .2

0 **

0 .0

7 1 .2

2 �

0 .2

4 *

0 .1

1 1 .2

7 �

0 .2

5 **

0 .0

8 1 .2

8 E m

p lo

ye d

0 .1

5 0 .1

6 —

0 .3

6 0 .2

4 —

� 0 .0

9 0 .1

7 —

H H

in co

m e

� 0 .0

9 **

0 .0

2 1 .0

9 0 .0

2 0 .0

4 —

� 0 .0

3 0 .0

3 —

S u b st

an ce

u se

0 .0

8 0 .1

1 —

1 .0

6 **

0 .1

8 2 .9

0 0 .5

4 **

0 .1

4 1 .7

2 N

e ga

ti ve

te m

p .

0 .0

6 0 .0

6 —

0 .5

3 **

0 .0

8 1 .6

9 0 .4

5 **

0 .0

6 1 .5

6 L iv

e w

/ N

S p ar

tn e r

0 .1

1 0 .1

6 —

0 .6

5 **

0 .2

5 1 .9

2 0 .5

3 **

0 .1

8 1 .6

9 L iv

e w

/ sp

o u se

� 0 .2

4 0 .1

6 —

0 .3

9 0 .2

4 —

0 .0

0 0 .1

8 —

L iv

e w

/ ad

u lt s

� 0 .1

8 0 .1

4 —

0 .0

8 0 .2

2 —

0 .0

3 0 .1

6 —

L iv

e w

/ ch

ild re

n 0 .2

5 *

0 .1

3 1 .2

9 0 .0

7 0 .1

9 —

0 .1

3 0 .1

5 —

F ri

e n d s

� 0 .0

6 0 .0

6 —

� 0 .0

1 0 .0

9 —

� 0 .0

4 0 .0

7 —

P ar

e n ts

� 0 .0

5 0 .0

5 —

� 0 .1

0 0 .0

8 —

� 0 .0

4 0 .0

6 —

Is o la

te d

0 .0

6 0 .0

6 —

0 .1

9 *

0 .1

0 1 .2

1 0 .3

7 **

0 .0

7 1 .4

4 N

ag e lk

e rk

e R

2 0 .1

8 M

o d e l x 2

6 0 7 .3

3 **

a ‘‘N

o in

vo lv

e m

e n t’ ’ in

IP V

is th

e o m

it te

d re

fe re

n ce

ca te

go ry

o f th

e d e p e n d e n t

va ri

ab le

in th

e an

al ys

is .

b T

h e

m u lt in

o m

ia l lo

gi st

ic re

gr e ss

io n

re su

lt s

ar e

b as

e d

o n

th e

w e ig

h te

d d at

a. c O

d d s

ra ti o s

fo r

n e ga

ti ve

co e ff ic

ie n ts

ar e

in ve

rt e d

fo r

e as

e o f in

te rp

re ta

ti o n .

*p �

.0 5 . **

p �

.0 1 .

44

report higher household incomes were less likely to report being ‘‘victims

only’’ of IPV relative to ‘‘no involvement’’ in IPV.

When we examine the predictors associated with being a ‘‘perpetrator

only’’ of IPV relative to ‘‘no involvement,’’ several independent variables

were significant in the expected direction. Respondents who are Black and

Hispanic are more likely to be perpetrators only. In addition, substance use,

negative temperament, living with nonspousal partner, and feeling isolated

were all positively related to being a perpetrator only of IPV. Older respon-

dents, males, and those with higher levels of education were less likely to

fall into the perpetrator only category.

Finally, the third panel in Table 3 presents the coefficients for being a

victim and a perpetrator of IPV in the past year, relative to ‘‘no involve-

ment.’’ All racial and ethnic groups (relative to White non-Hispanics) were

more likely to be both victims and perpetrators of IPV. Substance use,

negative temperament, living with nonspousal partner, and feeling isolated

were all positively related to being a victim and perpetrator of IPV. Finally,

those reporting higher levels of education were less likely to be IPV victims

and perpetrators. Below we discuss the results and the implications for

theory, policy, and future research.

Discussion

Three broad conclusions summarize our findings. First, there is consider-

able overlap in IPV victimization and perpetration. Consistent with research

on the victim-offender overlap in general (see Jennings, Piquero, and Rein-

gle 2012), this overlap was observed for both genders across various mea-

sures of violence. Unlike previous research, however, we found less overlap

for more serious forms of IPV, while the overlap was greatest when more

‘‘minor’’ forms of violence were examined. These findings suggest that

female IPV perpetration cannot be fully explained as ‘‘defensive’’ actions.

Though the purpose of our study was not to test a theoretical model of IPV,

the observed victim-offender overlap is consistent with a ‘‘family conflict/

violence’’ perspective that suggests that most IPV is ‘‘situational or

common couple’’ (Johnson 1995, 2008) violence, characterized by violence

perpetrated by both partners, with males and females engaging in roughly

the same amount of violence (see also Straus 2011; Straus, Gelles, and

Steinmetz 2006).

Second, though IPV victims and perpetrators appear to have similar

characteristics when we examine the bivariate correlations, there are impor-

tant differences once we distinguish between victims, offenders, and

Tillyer and Wright 45

victim-offenders and control for other variables. It seems that the common

bivariate correlates are due, in part, to the fact that victim-perpetrators are

being counted as both ‘‘IPV perpetrators and ‘‘IPV victims’’ in Table 2. The

importance of considering the victim-offender overlap in IPV studies is

highlighted by the fact that IPV ‘‘victims only’’ did not share many of the

same correlates as ‘‘victim-perpetrators,’’ and individuals falling into the

‘‘victim-perpetrators’’ category appear to be more similar to perpetrators.

Thus, it seems that common characteristics, routines, and/or behaviors

shared between offenders and some victims (e.g., Carbone-Lopez and

Kruttschnitt 2010; Lauritsen and Davis Quinet 1995; Lauritsen, Sampson,

and Laub 1991) may help to account for the observed overlap in IPV. This

is consistent with the principle of homogamy (Cohen, Kluegel, and Land

1981; Lauritsen, Sampson, and Laub 1991) discussed in the broader litera-

ture regarding the victim-offender overlap. Being a victim-only of IPV,

however, is explained by correlates not shared with perpetrators and

victim-perpetrators, suggesting that the nearly 10% of individuals who are ‘‘victims only’’ may require different theoretical explanations. For exam-

ple, the negative influence of household income and the positive influence

of living with children suggest that those who are ‘‘victims only’’ may have

fewer options available to them, as they may have children to support and

few financial resources. Such dependency issues have been suggested in the

broader literature on partner violence (Yount 2005). Thus, our study does

not invalidate previous research on correlates of IPV, but rather it clarifies

the influence of various factors by highlighting the importance of consider-

ing the victim-offender overlap.

Third, as interesting as our victim-offender overlap findings might be, of

equal importance is the relatively large proportion of individuals who report

being an IPV victim or perpetrator exclusively. Some of this violence likely

goes beyond the ‘‘common couple’’ conflicts described by the family

violence perspective, and in certain cases, may represent the frequent and

severe violence often described by the feminist perspective. Though the

focus of the present study was on IPV prevalence, we did explore frequency

of IPV in preliminary descriptive analyses. Of note, among those who

reported being ‘‘victims only’’ of IPV injury, six women reported being

injured 20 or more times in the past year. There were no men in that cate-

gory, and in the next category (11 to 20 injuries) there were no women or

men. In short, there appears to be a very small proportion of the sample

in which gendered, ongoing, and serious violence is occurring.

Given our findings, there are likely two (or more) phenomena captured

by our measure of IPV involvement that occur within varied relationship

46 Journal of Research in Crime and Delinquency 51(1)

dynamics: common couple violence with considerable victim-offender

overlap, and violence perpetrated by one partner against the other, which

may be high in frequency and gendered. Our study, using an inclusive mea-

sure of IPV prevalence to explore the victim-offender overlap in a nation-

ally representative sample, is more appropriate for examining the former.

We acknowledge, however, that included within these data are perpetrators

and victims of much more serious and repeated brutal attacks that are not

the focus of the present study, but remain an important area of study for

future research.

The victim-offender overlap in IPV presents challenges for those work-

ing to address violence between intimates. For instance, police officers who

respond to IPV incidents may find it especially difficult to determine the

‘‘offender’’ and ‘‘victim’’ in any single incident, given that both parties may

have a history of both perpetration and victimization. Actions preventing or

responding to IPV perpetration may be relevant for those who are both

offenders and victims, but ‘‘victims-only’’ will likely require distinctly

different services. Treatment providers, police, courts, and other service

providers should be aware of the factors that increase both victimization and

perpetration of IPV, and when possible, reduce or respond to these risk

factors. In particular, counselors and treatment providers should target to

change substance use problems, isolation, and negative temperaments that

were observed for IPV perpetrators and victim-perpetrators. They should

attempt to increase education levels among both offenders and victims.

Additionally, taking steps to reduce the opportunity for violence, such as

enforcing (not just issuing) no-contact orders, may help curb further

violence. Cultural differences should also be considered in the response

to violence in relationships. African Americans and Hispanics were more

likely to be involved in violent relationships as both perpetrators and

victim-perpetrators; perhaps, ‘‘fighting back’’ or mutual violence is more

acceptable among these groups and service personnel who attempt to

address such violence should be aware of this. Conversely, those who are

victims, but not perpetrators, will require different responses. In particular,

variables related loosely to dependence, such as living with children,

education, and income, appear to be important to victims-only of IPV, and

should therefore be targeted for intervention efforts.

Despite the contribution of the present study to our knowledge of IPV

and the victim-offender overlap, there are data limitations that warrant

further discussion and suggest directions for future research. First, we

examined how respondent characteristics affect respondent offending and

respondent victimization. We do not, however, have measures of the

Tillyer and Wright 47

partner’s characteristics. Thus, we cannot examine how the combination of

respondent and partner characteristics affect the outcomes. Future research

should examine the IPV victim-offender overlap in partner samples.

Second, the present study relied on cross-sectional data. This is fairly

common among studies that examine how proximal conditions influence

victimization and offending. Using longitudinal data that measure indepen-

dent variables one or more years prior to the outcomes presents its own

limitations with respect to measurement error. However, it is important to

acknowledge that we cannot establish temporal order with certainty, and

in fact, we suspect that there are likely reciprocal relationships between

some of the independent variables and IPV. Social isolation, for example,

may be both a cause and a consequence of IPV.

Third, because we examine individuals as the unit of analysis, rather than

incidents, we are not able to determine the immediate context which facil-

itates or prevents IPV. It is likely that some of the respondents identified as

victim-perpetrators were actually victims who engaged in violence as self-

defense. Unfortunately, we cannot ascertain which respondents were enga-

ging in retaliatory or defensive violence. This problem is not unique to IPV

or our study; despite our interest in understanding crime incidents, most crime

research uses individual-level data. Studies that use incidents as the units of

analysis (i.e., National Incident Based Reporting System or the National

Crime Victimization Survey) can tell us what crime incidents look like and

the characteristics that predict outcomes of those incidents (e.g., victim

injury, arrest of suspect), but we have no comparison for incidents that do not

result in a crime. There is no system that records noncrime incidents (not to

mention the challenges of defining what constitutes an ‘‘incident’’).

Finally, we want to emphasize that our findings reflect the analysis of

data collected from a general population sample. While this is not a limita-

tion per se, it is important to acknowledge that samples drawn from high-

risk populations, such as shelter samples, may have very different composi-

tions with respect to the victim-offender overlap. Further, it is possible that

examining the frequency and/or severity of violence, instead of prevalence,

may have produced different results regarding the victim-offender overlap.

Our precursory examination of the frequency of violent acts, discussed

above, demonstrates that the nature of IPV is complex and varied. We sug-

gest that future research consider this issue, as the victim-offender overlap

may be less pronounced when the measure of IPV is limited to the most

frequent and severe cases.

An additional avenue for future research is to examine the influence of

neighborhood context on IPV and the victim-offender overlap. Recent

48 Journal of Research in Crime and Delinquency 51(1)

research by Berg and colleagues (2012) demonstrates that the magnitude of

the victim-offender overlap for violence in a sample of African American

youth varied depending on neighborhood context. Prior research suggests

that neighborhood context influences IPV in general (Miles-Doan 1998;

Pinchevsky and Wright 2012; Wright and Benson 2011); whether these

same factors influence the victim-offender overlap for these crimes is

unknown.

In sum, the present study examined the victim-offender overlap with

respect to IPV. Results suggest that there is considerable overlap, with

approximately one-third of those involved in IPV in the past year reporting

both perpetration and victimization. This overlap was observed across men

and women, and varying levels of violence severity. Our multivariate

analysis revealed that those who are both victims and perpetrators of IPV

are more similar to those who report being IPV perpetrators only than IPV

victims only. Future research is needed to determine whether these relation-

ships are observed in other types of samples and across various contexts.

Acknowledgments

We thank the editor and reviewers for their helpful comments. Special acknowledg-

ment is due Ronald R. Rindfuss and Barbara Entwisle for assistance in the original

design.

Authors’ Note

This research uses data from Add Health, a program project directed by Kathleen

Mullan Harris and designed by J. Richard Udry, Peter S. Bearman, and Kathleen

Mullan Harris at the University of North Carolina at Chapel Hill, and funded by

grant P01-HD31921 from the Eunice Kennedy Shriver National Institute of Child

Health and Human Development, with cooperative funding from 23 other federal

agencies and foundations. Special acknowledgment is due Ronald R. Rindfuss and

Barbara Entwisle for assistance in the original design. Information on how to obtain

the Add Health data files is available on the Add Health website (http://

www.cpc.unc.edu/addhealth). No direct support was received from grant P01-

HD31921 for this analysis.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research,

authorship, and/or publication of this article.

Funding

The authors received no direct financial support for the research and/or authorship of

this article.

Tillyer and Wright 49

Notes

1. Though other waves of the Add Health do include measures of IPV, we examined

wave 4 data exclusively because we were interested in this stage of the life

course, which presumably has notable differences from the adolescent stages

observed in waves 1 and 2 and the late adolescence/young adulthood observed

in wave 3. Given the age of the respondents at wave 4, there is substantial varia-

tion with respect to marriage, cohabitation, children, and employment, all of

which are common correlates of IPV and are of importance to our study. That

being said, future research should examine the victim-offender overlap in IPV

at other stages of the life course.

2. We used the Wave IV weights provided with the Add Health data to create

proportional weights that sum to N (4, 275), so that the proportions between the

weights are correct and the scale effect of the weights is eliminated. This pro-

duces unbiased estimates and correct significance tests. For more information

on how the Add Health created the sample weights, please see Wave IV Weights

at http://www.cpc.unc.edu/projects/addhealth/codebooks/wave4.

3. Responses to the latter three items were reverse coded for the factor analysis.

References

Bachman, Ronet and Linda E. Saltzman. 1995. Violence against Women: Estimates

from The Redesigned Survey. Washington, DC: Bureau of Justice Statistics

Special Report, U.S. Department of Justice.

Berg, Mark T., Eric A. Stewart, Christopher J. Schreck, and Ronald L. Simons.

2012. ‘‘The Victim-Offender Overlap in Context: Examining the Role of Neigh-

borhood Street Culture.’’ Criminology 50:1-31.

Black, Michele C., Kathleen C. Basile, Matthew Breiding, Sharon G. Smith, Mikel

L. Walters, Melissa T. Merrick, Jieru Chen, and Mark R. Stevens. 2011. The

National Intimate Partner and Sexual Violence Survey (NISVS): 2010 Summary

Report. Atlanta, GA: National Center for Injury Prevention and Control, Center

for Disease Control and Prevention.

Browning, Christopher R. 2002. ‘‘The Span of Collective Efficacy: Extending

Social Disorganization Theory to Partner Violence.’’ Journal of Marriage and

the Family 64:833-50.

Carbone-Lopez, Kristin and Candace Kruttschnitt. 2010. ‘‘Risky Relationships?

Assortive Mating and Women’s Experiences of Intimate Partner Violence.’’

Crime & Delinquency 56:358-84.

Cloward, Richard A. 1959. ‘‘Illegitimate Means, Anomie, and Deviant Behavior.’’

American Sociological Review 24:164-76.

50 Journal of Research in Crime and Delinquency 51(1)

Cohen, Lawrence E., J. R. Kluegel, and Kenneth C. Land. 1981. ‘‘Social Inequality

and Predatory Criminal Victimization: An Exposition and Test of a Formal

Theory.’’ American Sociological Review 46:505-24.

Coker, Ann L., Paige H. Smith, Martie P. Thompson, Robert E. McKeown, Lesa

Bethea, and Keith E. Davis. 2002. ‘‘Social Support Protects against the Negative

Effects of Partner Violence on Mental Health.’’ Journal of Women’s Health &

Gender-Based Medicine 11:465-76.

Cullen, Francis T. 1994. ‘‘Social Support as an Organizing Concept for Criminol-

ogy: Presidential Address to the Academy of Criminal Justice Sciences.’’ Justice

Quarterly 11:527-59.

DeMaris, Alfred A., Michael L. Benson, Greer Litton Fox, Terrence Hill, and Judy

Van Wyk. 2003. ‘‘Distal and Proximal Factors in Domestic Violence: A Test of

an Integrated Model.’’ Journal of Marriage and the Family 65:652-67.

Dobash, R. Emerson and Russell P. Dobash. 1979. Violence against Wives: A Case

against the Patriarchy. New York: Free Press.

Dobash, Russell P., R. Emerson Dobash, Margo Wilson, and Martin Daly. 1992.

‘‘The Myth of Sexual Symmetry in Marital Violence.’’ Social Problems 39:

71-91.

Dugan, Laura and Robert Apel. 2005. ‘‘The Differential Risk of Retaliation by Rela-

tional Distance: A More General Model of Violent Victimization.’’ Criminology

43:697-730.

Feldman, Clyde M. and Carl A. Ridley. 2000. ‘‘The Role of Conflict-Based Com-

munication Responses and Outcomes in Male Domestic Violence toward Female

Partners.’’ Journal of Social and Personal Relationships 17:552-73.

Felson, Richard B. and Kelsea Jo Lane. 2010. ‘‘Does Violence Involving

Women and Intimate Partners Have a Special Etiology?’’ Criminology 48:

321-38.

Felson, Richard B. and Steven F. Messner. 2000. ‘‘The Control Motive in Intimate

Partner Violence.’’ Social Psychology Quarterly 63:86-94.

Gelles, Richard J. and Murray A. Straus. 1988. Intimate Violence: The Definitive

Study of the Causes and Consequences of Abuse in the American Family. New

York: Simon and Schuster.

Gottfredson, Michael and Travis Hirschi. 1990. A General Theory of Crime. Stan-

ford, CA: Stanford University Press.

Harris, K. M., C. T. Halpern, E. Whitsel, Jon M. Hussey, J. Tabor, P. Entzel, and J. R.

Udry. 2009. ‘‘The National Longitudinal Study of Adolescent Health: Research

Design.’’ Retrieved November 17, 2011 (http://www.cpc.unc.edu/projects/

addhealth/design).

Hirschi, Travis and Michael R. Gottfredson. 1983. ‘‘Age and the Explanation of

Crime.’’ American Journal of Sociology 89:552-84.

Tillyer and Wright 51

Holt, Stephanie, Helen Buckley, and Sadhbh Whelan. 2008. ‘‘The Impact of Expo-

sure to Domestic Violence on Children and Young People: A Review of the

Literature.’’ Child Abuse and Neglect 32:797-810.

Holtzworth-Munroe, Amy, Natalie Smutzler, and Leonard Bates. 1997. ‘‘A Brief

Review of the Research on Husband Violence: Part III: Sociodemographic

Factors, Relationship Factors, and Differing Consequences of Husband and Wife

Violence.’’ Aggression and Violent Behavior 2:285-307.

Jennings, Wesley G., George E. Higgins, Richard Tewksbury, Angela R. Gover, and

Alex R. Piquero. 2010. ‘‘A Longitudinal Assessment of the Victim-Offender

Overlap.’’ Journal of Interpersonal Violence 25:2147-74.

Jennings, Wesley G., Alex R. Piquero, and Jennifer M. Reingle. 2012. ‘‘On the

Overlap between Victimization and Offending: A Review of the Literature.’’

Aggression and Violent Behavior 17:16-26.

Johnson, Michael P. 1995. ‘‘Patriarchal Terrorism and Common Couple Violence:

Two Forms of Violence against Women.’’ Journal of Marriage and the Family

57:283-94.

Johnson, Michael P. 2008. A Typology of Domestic Violence: Intimate Terrorism,

Violent Resistance, and Situational Couple Violence. Lebanon, NH: Northeast-

ern University Press.

Kaufman Kantor, Glenda and Murray A. Straus. 1987. ‘‘The ‘‘Drunken Bum’’

Theory of Wife Beating.’’ Social Problems 34:213-30.

Kilpatrick, Dean G., Ron Acierno, Heidi S. Resnick, Benjamin E. Saunders, and

Connie L. Best. 1997. ‘‘A 2-year Longitudinal Analysis of the Relationships

between Violent Assault and Substance Use in Women.’’ Journal of Consulting

and Clinical Psychology 65:834-47.

Lauritsen, Janet L. and Kenna F. Davis Quinet. 1995. ‘‘Repeat Victimization among

Adolescents and Young Adults.’’ Journal of Quantitative Criminology 11:143-66.

Lauritsen, Janet L. and John H. Laub. 2007. ‘‘Understanding the Link between Vic-

timization and Offending: New Reflections on an Old Idea.’’ Crime Prevention

Studies 22:55-75.

Lauritsen, Janet L., Robert J. Sampson, and John H. Laub. 1991. ‘‘The Link between

Offending and Victimization among Adolescents.’’ Criminology 29:265-92.

Lawson, Jennifer. 2012. ‘‘Sociological Theories of Intimate Partner Violence.’’

Journal of Human Behavior in the Social Environment 22:572-90.

Lussier, Patrick, David P. Farringon, and Terrie E. Moffitt. 2009. ‘‘Is the Antisocial

Child Father of the Abusive Man? A 40-year Prospective Longitudinal Study on the

Developmental Antecedents of Intimate Partner Violence.’’ Criminology 47:741-80.

MacMillian, Ross and Rosemary Gartner. 1999. ‘‘When She Brings Home the

Bacon: Labor-Force Participation and the Risk of Spousal Violence against

Women.’’ Journal of Marriage and the Family 61:947-58.

52 Journal of Research in Crime and Delinquency 51(1)

Magdol, Lynn, Terrie E. Moffitt, Avshalom Caspi, and Phil A. Silva. 1998. ‘‘Devel-

opmental Antecedents of Partner Abuse: A Prospective-Longitudinal Study.’’

Journal of Abnormal Psychology 107:375-89.

Miles-Doan, Rebecca. 1998. ‘‘Violence between Spouses and Intimates: Does

Neighborhood Context Matter?’’ Social Forces 77:623-45.

Moffitt, Terrie E., Robert F. Krueger, Avshalom Caspi, and Jeffrey Fagan. 2000.

‘‘Partner Abuse and General Crime: How Are They the Same? How Are They

Different?’’ Criminology 38:199-232.

Moore, Todd M., Gregory L. Stuart, Jeffrey C. Meehan, Deborah L. Rhatigan,

Julianne C. Hellmuth, and Stefanie M. Keen. 2008. ‘‘Drug Abuse and Aggres-

sion between Intimate Partners: A Meta-analytic Review.’’ Clinical Psychology

Review 28:247-74.

Mustaine, Elizabeth E. and Richard Tewksbury. 2000. ‘‘Comparing the Lifestyles of

Victims, Offenders, and Victim-Offenders: A Routine Activity Theory Assess-

ment of Similarities and Differences for Criminal Participants.’’ Sociological

Focus 33:339-62.

Norlander, Bradley and Christopher Eckhardt. 2005. ‘‘Anger, Hostility, and Male

Perpetrators of Intimate Partner Violence: A Meta-analytic Review.’’ Clinical

Psychology Review 25:119-52.

Pampel, F. C. 2000. Logistic Regression: A Primer. Thousand Oaks, CA: Sage.

Pinchevsky, Gillian and Emily M. Wright. 2012. ‘‘The Impact of Neighborhoods on Inti-

mate Partner Violence and Victimization.’’ Trauma, Violence, & Abuse 13:112-32.

Rennison, Callie Marie and Sarah Welchans. 2000. Bureau of Justice Statistics

Special Report: Intimate Partner Violence. Washington, DC: U.S. Department

of Justice.

Ridley, Carl A. and Clyde M. Feldman. 2003. ‘‘Female Domestic Violence toward

Male Partners: Exploring Conflict Responses and Outcomes.’’ Journal of Family

Violence 18:157-70.

Smith, Erica L. and Donald Jr. Farole. 2009. Bureau of Justice Statistics Special

Report: State Court Processing Statistics: Profile of Intimate Partner Violence

Cases in Large Urban Counties. U.S. Department of Justice, Office of Justice

Programs, Washington, DC.

Stets, Jan E. 1991. ‘‘Cohabiting and Marital Aggression: The Role of Social Isola-

tion.’’ Journal of Marriage and the Family 53:669-80.

Stith, Sandra M., Douglas B. Smith, Carrie E. Penn, David B. Ward, and Dari Tritt.

2004. ‘‘Intimate Partner Physical Abuse Perpetration and Victimization Risk

Factors: A Meta-analytic Review.’’ Aggression and Violent Behavior 10:65-98.

Straus, Murray A. 2011. ‘‘Gender Symmetry and Mutuality in Perpetration of

Clinical-level Partner Violence: Empirical Evidence and Implications for

Prevention and Treatment.’’ Aggression and Violent Behavior 16:279-88.

Tillyer and Wright 53

Straus, Murray A., Richard J. Gelles, and Suzanne K. Steinmetz. 2006. Behind

Closed Doors: Violence in the American Family. New Brunswick, NJ:

Transaction.

Sugarman, David B. and Susan L. Frankel. 1996. ‘‘Patriarchal Ideology and

Wife-Assault: A Meta-Analytic Review.’’ Journal of Family Violence 11:

13-40.

Thomas, Kristie A., Melissa E. Dichter, and Jason Matejkowski. 2011. ‘‘Intimate

versus Nonintimate Partner Murder: A Comparison of Offender and Situational

Characteristics.’’ Homicide Studies 15:291-311.

Tjaden, Patricia and Nancy Thoennes. 1998. ‘‘Prevalence, Incidence, and Conse-

quences of Violence against Women: Findings from the National Violence

against Women Survey.’’ (No. NCJ-172837). Washington, DC: Department of

Justice, National Institute of Justice.

Tjaden, Patricia and Nancy Thoennes. 2000. ‘‘Prevalence and Consequences of

Male-to-Female and Female-to-Male Intimate Partner Violence as Measured

by the National Violence against Women Survey.’’ Violence against Women

6:142-61.

Van Wyk, Judy, Michael L. Benson, Greer Litton Fox, and Alfred A. DeMaris.

2003. ‘‘Detangling Individual-, Partner-, and Community-Level Correlates of

Partner Violence.’’ Crime and Delinquency 49:412-38.

Widom, Cathy Spatz. 1989. ‘‘The Cycle of Violence.’’ Science 244:160-66.

Wilkinson, Deanna L. and Susan J. Hamerschlag. 2005. ‘‘Situational Determinants

in Intimate Partner Violence.’’ Aggression and Violent Behavior 10:333-61.

Wright, Bradley R. Entner, Avshalom Caspi, Terrie E. Moffitt, and Phil A. Silva.

1999. ‘‘Low Self-Control, Social Bonds, and Crime: Social Causation, Social

Selection, or Both?’’ Criminology 37:479-514.

Wright, Emily M. and Michael L. Benson. 2011. ‘‘Clarifying the Effects of Neigh-

borhood Disadvantage and Collective Efficacy on Violence ‘‘behind Closed

Doors".’’ Justice Quarterly 28:775-98.

Yount, Kathryn. 2005. ‘‘Resources, Family Organization, and Domestic Violence

against Married Women in Minya, Egypt.’’ Journal of Marriage and the Family

67:579-96.

Author Biographies

Marie Skubak Tillyer is an assistant professor at the University of Texas at San

Antonio. Her research involves victimization, crime prevention, and violence. Her

work has appeared in journals such as Criminology, Crime and Delinquency, Crim-

inal Justice and Behavior, Journal of Research in Crime and Delinquency, Journal

of Interpersonal Violence, and Justice Quarterly.

54 Journal of Research in Crime and Delinquency 51(1)

Emily Wright is an assistant professor at the University of Nebraska, Omaha. Her

research involves neighborhoods, intimate partner violence and victimization, expo-

sure to violence, and female offenders. Her work has appeared in journals such as

Criminology, Social Problems, Justice Quarterly, Child Abuse & Neglect, and

Trauma, Violence, & Abuse.

Tillyer and Wright 55