j interpers violence 2002 mcmahon 1002 19

19
http://jiv.sagepub.com/ Journal of Interpersonal Violence http://jiv.sagepub.com/content/17/9/1002 The online version of this article can be found at: DOI: 10.1177/088626050201700906 2002 17: 1002 J Interpers Violence Jennifer McMahon and Jody Clay-Warner Gender Child Abuse and Future Criminality: The Role of Social Service Placement, Family Disorganization, and Published by: http://www.sagepublications.com On behalf of: American Professional Society on the Abuse of Children can be found at: Journal of Interpersonal Violence Additional services and information for http://jiv.sagepub.com/cgi/alerts Email Alerts: http://jiv.sagepub.com/subscriptions Subscriptions: http://www.sagepub.com/journalsReprints.nav Reprints: http://www.sagepub.com/journalsPermissions.nav Permissions: http://jiv.sagepub.com/content/17/9/1002.refs.html Citations: What is This? - Sep 1, 2002 Version of Record >> at Alexandru Ioan Cuza on June 29, 2014 jiv.sagepub.com Downloaded from at Alexandru Ioan Cuza on June 29, 2014 jiv.sagepub.com Downloaded from

Upload: lauralungu

Post on 19-Jul-2016

217 views

Category:

Documents


2 download

DESCRIPTION

wtwatt

TRANSCRIPT

Page 1: J Interpers Violence 2002 McMahon 1002 19

http://jiv.sagepub.com/Journal of Interpersonal Violence

http://jiv.sagepub.com/content/17/9/1002The online version of this article can be found at:

 DOI: 10.1177/088626050201700906

2002 17: 1002J Interpers ViolenceJennifer McMahon and Jody Clay-Warner

GenderChild Abuse and Future Criminality: The Role of Social Service Placement, Family Disorganization, and

  

Published by:

http://www.sagepublications.com

On behalf of: 

American Professional Society on the Abuse of Children

can be found at:Journal of Interpersonal ViolenceAdditional services and information for    

  http://jiv.sagepub.com/cgi/alertsEmail Alerts:

 

http://jiv.sagepub.com/subscriptionsSubscriptions:  

http://www.sagepub.com/journalsReprints.navReprints:  

http://www.sagepub.com/journalsPermissions.navPermissions:  

http://jiv.sagepub.com/content/17/9/1002.refs.htmlCitations:  

What is This? 

- Sep 1, 2002Version of Record >>

at Alexandru Ioan Cuza on June 29, 2014jiv.sagepub.comDownloaded from at Alexandru Ioan Cuza on June 29, 2014jiv.sagepub.comDownloaded from

Page 2: J Interpers Violence 2002 McMahon 1002 19

JOURNAL OF INTERPERSONAL VIOLENCE / September 2002McMahon, Clay-Warner / CHILD ABUSE AND CRIMINALITYThis study examines the extent to which family disorganization moderates the effect of social ser-vice placement on juvenile and adult arrests. Using Widom’s prospective data containing 749substantiated cases of child abuse and neglect, the authors test hypotheses relating to two mea-sures of family disorganization: family separation and family moves. They find that removing anabused or neglected child from the home increased the likelihood of adult arrest for children whoexperienced a recent family separation. Placement reduced likelihood of arrest for males whoexperienced frequent moves and increased risk of adult arrest for females who experienced fre-quent moves. The authors conclude that gender differences in placement outcomes should beexplored, and they discuss the implications of this research for social service agencies.

Child Abuse and Future CriminalityThe Role of Social Service Placement,Family Disorganization, and Gender

JENNIFER MCMAHONJODY CLAY-WARNER

University of Georgia

The serious consequences of child abuse and neglect are well documented.Victims of child abuse or neglect often suffer from poor attachment forma-tion (Crittenden & Ainsworth, 1989), academic problems (Oats, 1989),behavioral problems (Salter, Richardson, & Kairys, 1985), and feelings ofbetrayal thatmay lead to anger, hostility, and distrust of others (Wolfe, 1999).Considerable attention has also been given to the link between child abuseand neglect and subsequent criminal behavior (e.g., Smith & Thornberry,1995; Widom, 1989a; Zingraff, Leiter, Myers, & Johnsen, 1993).

In cases of child abuse, social service workers have the difficult task ofdeciding how to intervene so as to protect the child from harm (Thoburn,1988). One of the most important decisions to be made in this regard iswhether to remove the child from the home. In most instances, social serviceworkers seek to keep the family intact and recommend out-of-home place-ment only when there is a clear risk to the child’s immediate safety(Steinhauer, 1991). In some cases, the social worker determines that thethreat of continued abuse or neglect is so great that placement is the onlyoption.

1002

Authors’ Note: An earlier version of this article was presented at the 2001 meetings of theSouthern Sociological Society in Atlanta, Georgia.

JOURNAL OF INTERPERSONAL VIOLENCE, Vol. 17 No. 9, September 2002 1002-1019© 2002 Sage Publications

at Alexandru Ioan Cuza on June 29, 2014jiv.sagepub.comDownloaded from

Page 3: J Interpers Violence 2002 McMahon 1002 19

Other times, however, the consequences of the child remaining in thehome are unclear. In these cases, social service workers must weigh the emo-tional costs of placement against the risk of abuse/neglect in the home. Indoing so, they often take into account family and social characteristics(Abner, 1980; Katz, Hampton, Newberger, Bowles, & Snyder, 1986). Place-ment may also have long-term consequences for the child, however. Forexample, because research indicates that experiencing abuse and neglectincreases a child’s risk of future criminality (e.g., Widom, 1989b), to theextent that placement reduces the child’s exposure to the abusive environ-ment, placement may also reduce the risk of future criminality. Research hasfound, however, that placement has little effect on the likelihood that theabused or neglected child will engage in criminal activity (Jonson-Reid &Barth, 2000; Runyan & Gould, 1988; Widom, 1991a; Widom & Maxfield,1996). We suggest that although placement may not have direct effects oncriminality, family characteristics moderate the role of placement in deter-ring future criminal behavior. Here we consider the role of familydisorganization.

Specifically, we examine the effects of out-of-home placement when chil-dren have experienced family separation and frequent family moves. Webegin by discussing the link between child abuse/neglect and criminality andthen review the literature on the role of out-of-home placements in the childabuse/neglect-criminality relationship.We then use data collected byWidom(1994) to examine the effect that removing an abused or neglected child fromhis or her home has on the likelihood of future arrests.

The Relationship Between Child Abuse/Neglect and Criminality

Early research reported a strong relationship between child abuse/neglectand future criminality (Alfaro, 1981; Bolton, Reich, & Gutierres, 1977;Geller & Ford-Somma, 1984; Lewis, Shanok, Pincus, &Glaser, 1979).Morerecent research finds that although the relationship is not as strong or direct asonce believed, the link between abuse/neglect and criminal behavior doesexist. In her prospective study of victims of childhood abuse and neglect,Widom (1989b) found that abused or neglected children had higher rates ofjuvenile arrests than did the nonabused comparison group, although the dif-ference between the groups was not as large as reported in previous studies.Smith and Thornberry (1995) found that 45% of the maltreated sample intheir study had an official report of delinquency, whereas 31.7% of the non-maltreated comparison group had an official arrest record (see also Zingraffet al., 1993).

McMahon, Clay-Warner / CHILD ABUSE AND CRIMINALITY 1003

at Alexandru Ioan Cuza on June 29, 2014jiv.sagepub.comDownloaded from

Page 4: J Interpers Violence 2002 McMahon 1002 19

Examining juvenile violent offenses, Widom (1991b) found that, overall,the abused and neglected children did not have significantly higher rates ofoffending than did those in the comparison group. Abused and neglectedfemales, however, had significantly higher violent juvenile offenses com-pared to females in the comparison group, whereas there were no differencesin the rates of reported violent juvenile offenses for the abused or neglectedmales compared to the nonabused males (Widom, 1991b). Widom (1990;Widom&Maxfield, 1996) also reported that the abused and neglected grouphad significantly higher rates of adult arrests than the comparison group aswell as higher overall rates of violent offending.

These studies, however, do not indicate that all or most abused or neglectedchildren commit crimes. Instead, a number of factors affect the relationshipbetween abuse and criminality (Smith & Thornberry, 1995; Widom, 1989b,1989c, 1989d, 1990, 1991a, 1991b; Zingraff et al., 1993). For example,Widom (1989b) andZingraff et al. (1993) found that the type ofmaltreatmentaffected rates of arrest for violent crime. Certain family characteristics, suchas the presence of an alcoholic father, also significantly increased risk ofarrest (Widom, 1991c).One less examined factor is out-of- homeplacement.

Child Abuse/Neglect and Criminality:The Role of Out-of-Home Placement

Few studies have examined whether placement may also protect a childfrom future criminal involvement (Jonson-Reid & Barth, 2000; Runyan &Gould, 1988; Widom, 1991a; Widom & Maxfield, 1996). Widom (1991a;Widom &Maxfield, 1996) found that arrest rates for victims removed fromtheir homes and placed in a different setting (i.e., foster home, relative’shome, etc.) were similar to the arrest rates of victims who remained in theirhomes. However, placement characteristics were associated with differentialarrest rates. For example, those who were placed at a younger age had thelowest rates of arrest. Furthermore, childrenwho spent less time (4 to 6 years)in their first placement had the highest arrest rates. Children who had experi-enced four ormore placements and thosewho had behavior problems had thehighest rates of arrests.1

Similar to Widom, Runyan and Gould (1988) found no significant differ-ence in the total number of juvenile offenses committed by abused andneglected children who received in-home care from social service workersand those who were placed into foster care for at least 3 years. They did find,however, that the foster care childrenwere at an increased risk for committing

1004 JOURNAL OF INTERPERSONAL VIOLENCE / September 2002

at Alexandru Ioan Cuza on June 29, 2014jiv.sagepub.comDownloaded from

Page 5: J Interpers Violence 2002 McMahon 1002 19

assaults and that the risk of delinquency, in general, for the foster care groupincreased as the number of placement moves increased.

Jonson-Reid andBarth (2000) found that childrenwith a reported incidentofmaltreatmentweremore than twice as likely as those in the general popula-tion to enter the California Youth Authority (CYA), which is California’sjuvenile justice system for serious young offenders. They also found that thelevel of services (no services vs. in-home care vs. placement) received by themaltreated children had no effect on criminal outcomes. They did find, how-ever, that race and gendermoderated the effects of placement on delinquency.AlthoughAfricanAmerican youths were at the greatest risk of incarceration,African American children who received social services after the investiga-tion of the maltreatment incident had lower rates of incarceration than thosewho received no services. Conversely, maltreated Caucasian children whowere placed into foster care had higher rates of incarceration than those whoreceived no services and thosewho received in-home care. Therewere no dif-ferences in the rates of incarceration for maltreated Hispanic children receiv-ing different levels of social services.

Overall, Jonson-Reid andBarth (2000) found that the abuse incident had agreater effect on the likelihood of incarceration for females than for males.The rate of entry into CYA for females with documented abuse histories wasalmost 3 times higher than those without abuse histories (.2/1000 comparedto .07/1000); the rates of entry for boyswith childwelfare histories was twicethat of the general population (2.9/1000 compared to 1.4/1000). They alsoreported that maltreated females who had higher social service involvementhad higher rates of incarceration. In contrast, there was little difference in therates of incarceration for maltreated males who received different levels ofsocial service involvement.

Although research finds that there are no overall differences in arrest out-comes for abused and neglected children who remain in their homes com-pared to abused and neglected children who are removed from their homes(Jonson-Reid & Barth, 2000; Runyan & Gould, 1988; Widom, 1991a;Widom &Maxfield, 1996), Jonson-Reid and Barth (2000) found significantdifferences when other factors were taken into consideration. They focused,however, on the effects of race and gender and did not take family characteris-tics into consideration. The present study builds on Jonson-Reid and Barth’sfindings by examining the role of family disorganization in moderating therelationship between placement and prevention of criminality. Specifically,we ask whether frequent moves and recent family separation affect the suc-cess of social service placement in preventing future criminal behavior.

McMahon, Clay-Warner / CHILD ABUSE AND CRIMINALITY 1005

at Alexandru Ioan Cuza on June 29, 2014jiv.sagepub.comDownloaded from

Page 6: J Interpers Violence 2002 McMahon 1002 19

Hypotheses: Explaining the Effects of Placement andFamily Disorganization on Likelihood of Arrest

A home in which abuse/neglect occurs is often characterized by insecureattachments between parents and children. When parents are abusive orneglectful, children learn adaptive attachment behaviors thatmake it difficultto form healthy relationships with childhood peers and later with otheradults. Both neglectful and punitive parenting, which are related to the devel-opment of poor attachment patterns, have been found to be correlated withdelinquency (Farrington & West, 1990). Murray (1996) described poorlyattached children as lacking “a sense of agency or secure sense of self,”whichmay manifest itself through behavioral disorders and delinquency (p. 55).Although the first year is considered to be a particularly important period forthe development of these secure attachment patterns, attachment continues tobe important throughout childhood. In fact, although the child develops his orher initial model of attachment through the relationship with parents, thismodel is open, so the child assimilates experienceswith other attachment fig-ures into the model (Crittendon&Ainsworth, 1989). Family disorganizationmay, however, disrupt the development of secure attachments because par-ents are either physically or emotionally unavailable. Placement may alsoaffect these attachments, as removing the child from the home could eitherweaken existing attachments or provide a more stable environment in whichthe childmay develop secure attachments. In light of the role that poor attach-ment may play in criminality, it is important to examine the effects that bothfamily disorganization and placement may have on crime commission.

Parental divorce or separation and the death of a familymember are stress-ful life events that can damage social and emotional relationships betweenfamily members (Wolfner & Gelles, 1993). Children whose parents aredivorced or separated are exposed to their parents’ marital discord(Cummings & Davies, 1994), and, as a result, they may become hostile andaggressive. The divorce or separation may also disrupt their relationshipswith their parents (Erel & Burman, 1995), further damaging attachmentbonds. In particular, parents may be physically and/or emotionally absentand therefore unable to provide the childwith the necessary sense of security.At the same time, the parents who are divorcing or who have suffered the lossof a loved one may also become so overwhelmed that they fail to see theeffects that the situation is having on their children. Thus, when such familyseparation occurs, the needs of the child may go unfulfilled and attachmentssuffer.

Despite the stress caused by such events, Widom (1991c) reported thatfamily separation is not directly linked to criminality. Research, however, has

1006 JOURNAL OF INTERPERSONAL VIOLENCE / September 2002

at Alexandru Ioan Cuza on June 29, 2014jiv.sagepub.comDownloaded from

Page 7: J Interpers Violence 2002 McMahon 1002 19

shown that divorce, separation, and a death in the family are actually factorsthat significantly contribute to child abuse and neglect (Wolfner & Gelles,1993). Abuse/neglect is likely to prevent the development of secure attach-ments, leading to an increased risk of criminal involvement. Thus, removingan abused or neglected child fromhis or her homeunder these conditionsmayprevent further abuse or neglect, provide the childwith amore stable environ-ment inwhich attachmentswith alternate caregiversmay develop, and reducethe risk of future criminality. We predict the following:

Hypothesis 1: Placement will reduce the likelihood of arrests more for abusedor neglected children who have experienced recent family separation than forthose who have not experienced recent family separation.

Another indication of family disorganization is frequent residentialmobility. Moving two or more times within 1 year has been linked to child-hood emotional, behavioral, and academic problems (Wood, Halfon,Scarlata, Newacheck, & Nessim, 1993). Frequent moves may also impedethe development of social ties with alternative caregivers. This is particularlyimportant for abused or neglected children, who often suffer from poorattachment formation. Also, frequent residential mobility affects communityties (Raviv, Keinan, Abazon, & Raviv, 1990). According to Hirschi’s (1969)social bond theory, strong social and community ties (i.e., to friends, teach-ers, neighbors, etc.) instill feelings of belonging to the community and obli-gation or pressure to adhere to the community’s norms and values. Frequentfamily moves, however, may hinder the development and maintenance ofsocial bonds. When abused or neglected children who have experienced fre-quent moves are placed in a more stable environment, they may be able toestablish the necessary social bonds that help to deter criminal behavior. Con-versely, removing an abused or neglected child who has not experienced fre-quent family moves may destroy existing social bonds, placing the child at ahigher risk for criminal behavior.

Hypothesis 2: Placement will reduce the likelihood of arrests more for abusedand neglected children whose family experienced frequent moves prior to theabuse or neglect incident than for abused or neglected children who did notmove frequently.

Given the findings that both placement and abuse/neglect may have dif-ferential effects on male and female criminality (Jonson-Reid and Barth,2000; Widom, 1989b, 1991b), we also conduct separate analyses for malesand females, although we do not make specific predictions about genderdifferences.

McMahon, Clay-Warner / CHILD ABUSE AND CRIMINALITY 1007

at Alexandru Ioan Cuza on June 29, 2014jiv.sagepub.comDownloaded from

Page 8: J Interpers Violence 2002 McMahon 1002 19

METHOD

Sample

The data used for this studywere collected byCathy SpatzWidom (1994),with support from the National Institute of Justice, and made availablethrough the Inter-University Consortium for Political and Social Research(ICPSR). The data set contains detailed information from official juvenileand adult court records on 908 substantiated cases of abuse/neglect commit-ted against children under the age of 12 in a midwestern metropolitan areabetween 1967 and 1971. Included in these records was information about thevictim and offender as well as the results of the court investigation of theabuse or neglect incident. Widom then examined official juvenile and adultarrest records in 1987, 1988, and 1994 to determine which of the participantswere involved in criminal activity. In the original social service report, thereis placement information for 749 of these individuals; these persons consti-tute the sample used in the current analyses. Forty-nine percent of the partici-pants aremale, and 33%are non-White (Black orHispanic). Almost all of theparticipants had passed through the years of peak offending by the end of thedata collection (Widom & Maxfield, 1996).

Variables

The dependent variables for this study are binarymeasures of arrest. Juve-nile arrest includes any official arrest for an offense committedwhile the indi-vidual was under the age of 18 (excluding traffic offenses). Twenty-six per-cent of the participants had at least one juvenile arrest. Adult arrest includesany official arrest for an offense committed while the individual was over theage of 18 (excluding traffic offenses). Thirty percent of the participants had atleast one adult arrest.

The independent variables are placement, family separation, and frequentmoves.Widom coded these data based on information contained in the origi-nal court records detailing the abuse or neglect incident. In cases of multipleor ongoing abuse, the variables were coded in relation to the particular inci-dent that brought the abuse/neglect case to the attention of the courts. Place-ment indicates whether or not the individual was removed from his or herhome by a social service agency. Eighty-one percent of the children wereplaced. Family separation represents a divorce, separation, or death occur-ring around the time of the abuse or neglect incident.2 Thirty-eight percent ofparticipants had experienced a family separation. Frequent moves indicateswhether the child’s family moved two or more times during the year prior to

1008 JOURNAL OF INTERPERSONAL VIOLENCE / September 2002

at Alexandru Ioan Cuza on June 29, 2014jiv.sagepub.comDownloaded from

Page 9: J Interpers Violence 2002 McMahon 1002 19

the abuse or neglect incident. Twenty-four percent of the participants experi-enced frequent moves.

We include controls for gender (0 = female, 1 =male) and race (0 =White,1 = non-White). The original data set does not include information on age attime of abuse or socioeconomic status. Therefore, we could not control forthese factors.

RESULTS

Weconducted binary logistic regression analyses using one-tailed signifi-cance tests when testing directional hypotheses and two-tailed tests for allother analyses. In all of the regression models, both control variables weresignificant, indicating thatmales and non-Whites had higher rates of juvenileand adult arrest. Placement was marginally significant in the main effectsmodel presented in Table 1 (p < .1).3 Placement was unrelated to adult arrestin both main effect models.

Hypothesis 1 predicted that placementwill reduce the likelihood of arrestsmore for abused or neglected children who have experienced a recent familyseparation than for those who have not experienced a recent family separa-tion. The results failed to provide support for the hypothesis (see Table 1).Although the regressionmodel for adult arrests indicated amarginally signif-icant interaction (B = 1.13, p = .054),4 a closer examination of the effectsdemonstrated that placement is associated with an increased risk of adultarrest for those who have suffered a recent family separation. By partialingout the main effects, we calculated the odds ratio associated with the interac-tion term (seeKleinbaum, 1994). Childrenwhowere placed following a fam-ily separation were 1.8 times more likely to be arrested as adults than thosewho had experienced a family separation butwere not placed. The interactioneffects were similar for males and females.

Hypothesis 2 predicted that placement reduces the likelihood of arrestsmore for abused or neglected children whose family experienced frequentmoves prior to the abuse or neglect incident than for abused or neglected chil-drenwho did notmove frequently. In reviewing juvenile arrests, we found theresults provide support for this hypothesis (B = –.89; p < .05, one-tailed).Children who had experienced frequent moves and were placed were 38%less likely to be arrested as juveniles than those who were placed but had notexperienced frequent moves. Overall, placement reduced the likelihood ofjuvenile arrests by 28% for those who had experienced frequent moves (seeTable 2).

McMahon, Clay-Warner / CHILD ABUSE AND CRIMINALITY 1009

at Alexandru Ioan Cuza on June 29, 2014jiv.sagepub.comDownloaded from

Page 10: J Interpers Violence 2002 McMahon 1002 19

1010

TABLE 1: Unstandardized Coefficients (Beta) and Odds Ratios (OR) From Logistic Regression Predicting Effects of Placement and FamilySeparation on Likelihood of Arrest

Juvenile Arrest Adult Arrest

Main Effects Interaction Model Main Effects Interaction Model

Variable Beta OR Beta OR Beta OR Beta OR

Gender 0.75 2.12*** 0.75 2.12*** 1.44 4.22*** 1.45 4.23***Race 0.83 2.29*** 0.83 2.29*** 0.66 1.92*** 0.65 1.92***Placement 0.41 1.51* 0.44 1.56* –0.33 0.72 –0.56 0.57**Family separation –0.24 0.79 –0.08 0.92 0.13 1.14 –0.87 0.42Placement × Separation –0.18 0.84 1.13 3.09*Model χ2 45.78*** 45.87*** 82.73*** 86.86***

*p < .10. **p < .05. ***p < .01. All two-tailed tests.

at Alexandru Ioan C

uza on June 29, 2014jiv.sagepub.com

Dow

nloaded from

Page 11: J Interpers Violence 2002 McMahon 1002 19

1011

TABLE 2: Unstandardized Coefficients (Beta) and Odds Ratios (OR) From Logistic Regression Predicting Effects of Placement and FrequentFamily Moves on Likelihood of Arrest

Juvenile Arrest Adult Arrest

Main Effects Interaction Model Main Effects Interaction Model

Variable Beta OR Beta OR Beta OR Beta OR

Gender 0.75 2.11*** 0.75 2.12*** 1.44 4.23*** 1.44 4.23***Race 0.82 2.27*** 0.83 2.29*** 0.64 1.91*** 0.64 1.91***Placement 0.35 1.42 0.56 1.74** –0.29 0.75 –0.30 1.20Frequent moves –0.33 0.72 0.42 1.52 0.03 1.03 0.01 1.10Placement × Frequent Moves –0.89 0.41* 0.02 1.02Model χ2 46.72*** 49.63*** 82.25*** 82.23***

*p < .10. **p < .05. ***p < .01. All two-tailed tests.

at Alexandru Ioan C

uza on June 29, 2014jiv.sagepub.com

Dow

nloaded from

Page 12: J Interpers Violence 2002 McMahon 1002 19

Conducting the analysis separately for males and females, however, dem-onstrated that the significant effect of the interaction term is due to the protec-tive effect of placement formaleswhowere placed following frequentmoves(p < .01).Males who experienced frequent moves andwere placedwere 68%less likely to have a juvenile arrest thanweremaleswho experienced frequentmoves and were not placed (see Table 3). The interaction was not significantfor females in the juvenile model (see Table 4).

Therewas no significant interaction effect for adult arrests. However, test-ing this hypothesis for females only, the results were marginally significant(p = .064).4 These marginally significant results, however, were not in thepredicted direction. The results indicated that females who were placed fol-lowing frequent family moves were almost twice as likely to have an adultarrest as those who were not placed (see Table 3). Testing this hypothesis formales only, the results significant for adult arrests (p < .05, one-tailed) in thepredicted direction. Males who had frequent family moves and were placedwere 49% less likely to be arrested as adults than were males whose familieshad moved frequently but had not been placed.

DISCUSSION

We found that removing an abused or neglected child from the homeincreases the likelihood of arrest under certain circumstances yet reducesthe likelihood of arrest under other circumstances. We found that place-ment may act as a contributing factor to criminality for abused or neglectedchildren who have experienced a recent family separation. Abused orneglected children who experienced a family separation and were removedfrom their homeswere almost twice as likely to have an adult arrest as abusedor neglected children who experienced a family separation and were notremoved.

These results suggest that removing a child from his or her home follow-ing a family separation may create further instability in the child’s life.Although the parents may be unable to care for the child’s emotional needs atthis time, the child’s siblings, friends, neighbors, or teachers may provide thechild with the support that he or she needs. If the abused or neglected child isremoved from the home, however, the child will likely be removed from thissupport system, thus increasing the risk of criminality.

On the other hand, an abused or neglected child who has experienced afamily separation may benefit from receiving in-home social service care. Inthis situation, the child could use the existing social support systemwhile thesocial service worker implements a plan to prevent further abuse or neglect.

1012 JOURNAL OF INTERPERSONAL VIOLENCE / September 2002

at Alexandru Ioan Cuza on June 29, 2014jiv.sagepub.comDownloaded from

Page 13: J Interpers Violence 2002 McMahon 1002 19

1013

TABLE 3: Unstandardized Coefficients (Beta) and Odds Ratios (OR) From Logistic Regression Predicting Effects of Placement and FrequentFamily Moves on Likelihood of Arrest for Males

Juvenile Arrest Adult Arrest

Main Effects Interaction Model Main Effects Interaction Model

Variable Beta OR Beta OR Beta OR Beta OR

Race 1.08 2.94*** 1.11 3.04*** 0.76 2.15** 0.78 2.12***Placement 0.19 1.21 0.63 1.88* –0.08 0.93 0.20 1.22Frequent moves –0.29 0.75 1.22 3.37* 0.04 1.04 0.98 2.66Placement × Frequent Moves –1.77 0.17*** –1.10 0.34*Model χ2 26.59*** 33.09*** 12.21** 15.04***

*p < .10. **p < .05. ***p < .01. All two-tailed tests.

TABLE 4: Unstandardized Coefficients (Beta) and Odds Ratios (OR) From Logistic Regression Predicting Effects of Placement and FrequentFamily Moves on Likelihood of Arrest for Females

Juvenile Arrest Adult Arrest

Main Effects Interaction Model Main Effects Interaction Model

Variable Beta OR Beta OR Beta OR Beta OR

Race 0.47 1.59* 0.47 1.60* 0.38 1.46 0.39 1.47Placement 0.47 1.60 0.42 1.52 –0.63 0.54** –0.96 0.38***Frequent moves –0.35 0.70 –0.58 0.56 0.01 1.01 –1.24 0.29Placement × Frequent Moves 0.27 1.31 1.64 5.14*Model χ2 5.67 5.77 6.27* 13.79***

*p < .10. **p < .05. ***p < .01. All two-tailed tests.

at Alexandru Ioan C

uza on June 29, 2014jiv.sagepub.com

Dow

nloaded from

Page 14: J Interpers Violence 2002 McMahon 1002 19

Future research should focus on this issue by examining the effectiveness ofin-home social service care for abused and neglected children who haveexperienced a family separation.

In the second hypothesis, we examined the effectiveness of placementwhen a child had experienced frequent family moves. As suggested byHirschi’s (1969) social bond theory, we expected that placement would bemore effective when children had experienced frequent family moves. Wefound that placement serves as a protective factor for abused or neglectedmaleswhose family experienced frequentmoves in the year prior to the abuseor neglect incident, whereas placement may be deleterious for females whohave experienced frequent moves. Placement significantly reduced likeli-hood of both juvenile and adult arrest for males who had experienced fre-quentmoves. For females, the interaction termwas not significant in the juve-nile arrest model, indicating that placement was no more or less effectivewhen the abused or neglected child hadmoved frequently thanwhen the childhad not moved frequently. The interaction term approached significance inthe adult arrest model, but the effect was not in the predicted direction. Forfemales, placement following frequent family moves increased their risk ofadult arrest by almost 200%. Although this finding only approached signifi-cance, themagnitude of the effect is quite large. The size of this effect, in con-cert with Jonson-Reid and Barth’s (2000) report that females faced a greaterrisk of incarceration as level of social service involvement increased, sug-gests that future research should consider this issue. In particular, researchcould examinewhether other options, such as in-home social service care andfamily counseling, may be more appropriate for some females who haveexperienced abuse or neglect and frequent family moves. Any research, ofcourse, must consider not only the effectiveness of placement alternatives inpreventing future criminality but also the ability of these alternatives to pre-vent future abuse or neglect, as this is the clear priority of any social serviceintervention.

The fact that placement in a presumably more stable environment provedbeneficial formales but detrimental for females suggests that the existence ofsocial bonds may be less critical in the deterrence of female criminality.Research has found, however, that the absence of social bonds is equallyimportant in predicting male and female delinquency (Giordano,Cernkovich, & Pugh, 1986). Thus, it may be that females are better able thanmales to maintain existing social bonds despite frequent moves. It may alsobe that females are better able than males to establish new social bonds, suchas when a familymove involves changing schools and/or neighborhoods.Weare aware of no research that directly addresses the differential development

1014 JOURNAL OF INTERPERSONAL VIOLENCE / September 2002

at Alexandru Ioan Cuza on June 29, 2014jiv.sagepub.comDownloaded from

Page 15: J Interpers Violence 2002 McMahon 1002 19

and maintenance of social bonds among male and female children and sug-gest that this may be an area for future exploration.

Study Limitations and Suggestions for Future Research

One limitation of the present study is that the data (Widom, 1994) repre-sent official reports of child abuse or neglect occurring between the years of1967 and 1971. Although it was necessary to use cases of abuse or neglectreported more than 30 years ago to examine the sample for both juvenile andadult arrests, this presents a few problems for the present analysis. First of all,these cases were reported before the mandatory child abuse reporting lawswere enacted (Widom, 1989b). Therefore, these files represent only themostserious cases of child abuse or neglect. This sample of reported child abuseand neglect cases may be somewhat different from a sample of unreportedcases. As a result, the findings of the present study cannot be generalized tothe entire population of abused or neglected children. Instead, the presentfindings are only representative of the most serious cases of child abuse orneglect.

Another limitation of the present research is that the data used for thisstudy relied on official arrest records to measure the amount of involvementin criminal activities. Relying on official arrest records represents the crimi-nal offenses of abused or neglected children who “got caught” engaging incriminal activities and underestimates the actual extent of criminal behavior.This is a limitation shared by much of the research in this area (e.g., Jonson-Reid & Barth, 2000) and can be addressed only through the use of self-reported or family-reported measures of criminal behavior.

Finally, our study is limited by the information contained in the adminis-trative records. First, information on socioeconomic status (SES) was notavailable. Because official records rarely contain SES data, similarlydesigned studies have also been unable to control for SES (Johnson-Reid &Barth, 2000; also see Widom, 1991b, 1991c, which used these data). It islikely, however, that these data overrepresent lower SES children becausethese cases aremore likely to come to the attention of authorities, particularlybefore mandatory reporting. Thus, we cannot determine the extent to whichincreased financial resources might alter the effects reported in this research.Due to the nature of the administrative data,wewere also unable to examine anumber of family characteristics that maymediate the effectiveness of place-ment in deterring criminal activity as well as influence the placement deci-sion, such as parental or sibling criminal involvement, parental alcoholism,family size, and family belief system or religion. As a prospective design that

McMahon, Clay-Warner / CHILD ABUSE AND CRIMINALITY 1015

at Alexandru Ioan Cuza on June 29, 2014jiv.sagepub.comDownloaded from

Page 16: J Interpers Violence 2002 McMahon 1002 19

uses only substantiated cases of child abuse/neglect, however, these data arenot influenced by subject recall or retelling of past events to fit future conse-quences. Thus, these data are widely recognized as the best available sourceof information on the child abuse-delinquency relationship (Zingraff et al.,1993).

Overall, our results indicated that the effect of placement on criminalitymay depend on the presence of family disorganization, asmeasured by recentfamily separation and frequent family moves. We also found that the overallpattern of effects differs for males and females. Youths whose families expe-rienced a recent separation, such as through divorce, marital separation, ordeath in the family, were found to be at a greater risk of adult arrest if theywere removed from the home. On the other hand, placement had differenteffects for males and females when the family experienced frequent moves.This finding is consistent with Jonson-Reid and Barth’s (2000) report thatplacement has different effects on male and female criminality while furtherspecifying factors that moderate this relationship. We also recognize thatother family characteristics and situational factors, such as quality of place-ment, may influence the effect of placement on criminality, and we encour-age researchers to investigate these issues. Such research will assist socialservice workers in their placement decisions, allowing them to consider theneed to eliminate immediate physical and psychological harm as well asreduce the risk of future harm resulting from involvement in criminalbehavior.

NOTES

1. The behavioral problems include general conduct issues that do not necessarily constitutedelinquency. To ensure that the behavioral problems did not reflect preexisting delinquency,Widom removed those youthswhose placementwas in a home for delinquents fromher analysis.Behavioral problems remained a significant predictor of later criminality.

2.Widomdoes not elaborate on the coding of the “recent family disruption” variable in eitherthe code book or published descriptions of the data. Thus, there is no indication as to the specifictime frame necessary for a family disruption to be considered “recent.”

3. We report a marginally significant main effect for placement in the juvenile arrest model(p = .08). Widom (1991a, 1991c) reported, however, that placement is not related to arrest rates.Widom’s resultswere based on bivariate analyses and thusmay vary from the results found in themultivariate models presented here.

4. A two-tailed test of significance was used because the effects were not in the predicteddirection.

1016 JOURNAL OF INTERPERSONAL VIOLENCE / September 2002

at Alexandru Ioan Cuza on June 29, 2014jiv.sagepub.comDownloaded from

Page 17: J Interpers Violence 2002 McMahon 1002 19

REFERENCES

Abner, J. L. (1980). The involuntary child placement decision: Solomon’s dilemma revisited. InG. Ger, C. Ross, & E. Zigler (Eds.), Child abuse: An agenda for action. New York: OxfordUniversity Press.

Alfaro, J. (1981). Report on the relationship between child abuse and neglect and later sociallydeviant behavior. In R. J. Hunner &Y. E.Walker (Eds.), Exploring the relationship betweenchild abuse and delinquency (pp. 175-219). Montclair, NJ: Allanheld, Osmun.

Bolton, F. G., Reich, J., & Gutierres, S. E. (1977). Delinquency patterns in maltreated childrenand siblings. Victimology, 2, 349-359.

Crittenden, P. M., & Ainsworth, M. S. (1989). Child maltreatment and attachment theory. In D.Cicchetti & V. Carlson (Eds.), Child maltreatment: Theory and research on the causes andconsequences of child abuse and neglect (pp. 432-463). New York: Cambridge UniversityPress.

Cummings, E. M., & Davies, P. T. (1994). Children and marital conflict: The impact of familydispute and resolution. New York: Guilford.

Erel, O.,&Burman, B. (1995). Interrelatedness ofmarital relations and parent-child relations: Ameta-analytic review. Psychological Bulletin, 118, 108-132.

Farrington, D. P., & West, D. J. (1990). The Cambridge Study in Delinquent Development: Along-term follow-up of 411 London males. In Criminality: Personality, behaviour and lifehistory. Berlin, Germany: Springer-Verlag.

Geller,M., & Ford-Somma, L. (1984).Violent homes, violent children: A study of violence in thefamilies of juvenile offenders. Report prepared for the National Center on Child Abuse andNeglect. Trenton, NJ: New Jersey State Department of Corrections, Division of JuvenileServices.

Giordano, P., Cernkovich, S., & Pugh,M. (1986). Friendships and delinquency. American Jour-nal of Sociology, 91, 1170-1203.

Hirschi, T. (1969). Causes of delinquency. Berkeley: University of California Press.Jonson-Reid,M., &Barth, R. P. (2000). Frommaltreatment report to juvenile incarceration: The

role of child welfare services. Child Abuse and Neglect, 24(4), 505-520.Katz,M.H., Hampton, R. L., Newberger, E. H., Bowles, R. T., & Snyder, J. C. (1986). Returning

children home: Clinical decisionmaking in cases of child abuse and neglect.American Jour-nal of Orthopsychiatry, 56, 253-262.

Kleinbaum, D. G. (1994). Logistic regression: A self-learning text. New York: Springer.Lewis, D. O., Shanok, S. S., Pincus, J. H., & Glaser, G. H. (1979). Violent juvenile delinquents:

Psychiatric, neurological, psychological, and abuse factors. Journal of the American Acad-emy of Child Psychiatry, 18, 307-319.

Murray, L. (1996). Personal and social influences on parenting and adult adjustment. In S.Kraemer & J. Roberts (Eds.), The politics of attachment (pp. 43-61). London: Free Associa-tion Books.

Oats, R. K. (1989). The consequences of child abuse and neglect. Early Child Development andCare, 42, 57-69.

Raviv, A., Keinan, G., Abazon, Y., & Raviv, A. (1990). Moving as a stressful life event for ado-lescents. Journal of Community Psychology, 18, 130-140.

Runyan, D. K., &Gould, C. L. (1988). Foster care for child maltreatment: Impact on delinquentbehavior. In G. T. Hotaling & D. Finkelhor (Eds.), Coping with family violence: Researchand policy perspectives (pp. 224-237). Newbury Park, CA: Sage.

McMahon, Clay-Warner / CHILD ABUSE AND CRIMINALITY 1017

at Alexandru Ioan Cuza on June 29, 2014jiv.sagepub.comDownloaded from

Page 18: J Interpers Violence 2002 McMahon 1002 19

Salter, A., Richardson, C. M., & Kairys, S. W. (1985). Caring for abused preschoolers. ChildWelfare, 64, 343-356.

Smith, C., & Thornberry, T. P. (1995). The relationship between childhood maltreatment andadolescent involvement in delinquency. Criminology, 33(4), 451-481.

Steinhauer, P. D. (1991). The least detrimental alternative: A systematic guide to case planningand decision making for children in care. Toronto, Canada: University of Toronto Press.

Thoburn, J. (1988). Child placement: Principles and practice. Brookfield, VT: WildwoodHouse.

Widom, C. (1989a). Does violence beget violence? A critical examination of the literature. Psy-chological Bulletin, 106(1), 3-28.

Widom, C. (1989b). The cycle of violence. Science, 244, 160-166.Widom, C. (1989c). Child abuse, neglect, and violent criminal behavior. Criminology, 27(2),

251-271.Widom, C. (1989d). Child abuse, neglect, and adult behavior: Research design and findings on

criminality, violence, and child abuse. American Journal of Orthopsychiatry, 59(3), 355-367.

Widom, C. (1990). Childhood victimization and violent offending. Violence and Victims, 5(1),19-35.

Widom, C. (1991a). The role of placement experiences in mediating the criminal consequencesof early childhood victimization. American Journal of Orthopsychiatry, 6, 195-209.

Widom, C. (1991b). Childhood victimization: Risk factor for delinquency. InM. E. Colton & S.Gore (Eds.), Adolescent stress: Causes and consequences (pp. 201-221). NewYork: Aldinede Gruyter.

Widom, C. (1991c). Avoidance of criminality in abused and neglected children. Psychiatry, 54,162-174.

Widom, C. (1994). Child abuse, neglect, and violent criminal behavior in a Midwest metropoli-tan area of the United States (1967-1988) [Electronic data file]. Compiled by the Depart-ments of Criminal Justice and Psychology, IndianaUniversity. AnnArbor,MI: Inter-Univer-sity Consortium for Political and Social Research.

Widom, C., & Maxfield, G. (1996). A prospective examination of risk for violence amongabused and neglected children. In C. F. Ferris & T. Grisso (Eds.), Understanding aggressivebehavior in children. New York: New York Academy of Sciences.

Wolfe, D. A. (1999). Child abuse: Implications for child development and psychopathology.Thousand Oaks, CA: Sage.

Wolfner, G. D., & Gelles, R. J. (1993). A profile of violence toward children: A national study.Child Abuse and Neglect, 17, 197-212.

Wood, D., Halfon, N., Scarlata, D., Newacheck, P., & Nessim, S. (1993). Impact of family relo-cation on children’s growth, development, school function, and behavior. Journal of theAmerican Medical Association, 270, 1334-1338.

Zingraff,M. T., Leiter, J.,Myers,K.A.,& Johnsen,M.C. (1993). Childmaltreatment and youth-ful problem behavior. Criminology, 31(2), 173-202.

Jennifer McMahon recently received her B.S. with honors from the University of Geor-gia, where she double majored in criminal justice and psychology. She is currently theintake and intervention specialist at the Gwinnett Sexual Assault Center and GwinnettChildren’s Advocacy Center.

1018 JOURNAL OF INTERPERSONAL VIOLENCE / September 2002

at Alexandru Ioan Cuza on June 29, 2014jiv.sagepub.comDownloaded from

Page 19: J Interpers Violence 2002 McMahon 1002 19

Jody Clay-Warner is an assistant professor of sociology at the University of Georgia.Her research interests include violence against women, procedural justice, and victim-ology. Reports of her most recent research have appeared in Violence and Victims andSocial Psychology Quarterly.

McMahon, Clay-Warner / CHILD ABUSE AND CRIMINALITY 1019

at Alexandru Ioan Cuza on June 29, 2014jiv.sagepub.comDownloaded from