U.S. Department of Justice Office of Justice Programs National Institute of Justice

Charles B. DeWitt, Director September 1992 The Cycle of

By Cathy Spatz Widom

Does childhood lead to adult While most members of both groups had and interviewing a large sample of the criminal behavior? no juvenile or adult criminal record, being previously abused and neglected children abused or neglected as a child increased How likely is it that today's abused to draw a more complete picture of the the likelihood of arrest as a juvenile by 53 and neglected children will become consequences of childhood victimization. percent, as an adult by 38 percent, and for tomorrow's violent offenders? The remainder of this report presents Phase a violent crime by 38 percent. I results in greater detail and introduces In one of the most detailed studies of the preliminary findings from Phase 11. The "cycle of violence" hypothesis sug- issue to date, research sponsored by the gests that a childhood history of physical National Institute of Justice (NIJ) found abuse predisposes the survivor to violence Study design that childhood abuse increased the odds of in later years. This study reveals that vic- future delinquency and adult criminality tims of are also more likely to Several important design features distin- overall by 40 percent. The study followed develop later criminal violent behavior as guish this research from prior efforts to 1,575 cases from childhood through young well. This finding gives powerful support study the intergenerational transmission of adulthood, comparing the arrest records of to the need for expanding common con- violence.' First, by following a large num- two groups: ceptions of . If it is not only ber (1,575) of cases from childhood 0 A study group of 908 substantiated violence that begets violence, but also through adolescence into young adulthood, cases of childhood abuse or neglect proc- neglect, far more attention needs to be this "prospective" study was able to exarn- essed by the courts between 1967 and 1971 devoted to the families of children whose ine the long-term consequences of abuse and tracked through official records over "beatings" are forms of abandonment and and neglect. The sample, drawn from a the next 15 to 20 years. severe malnutrition. An example of inter- metropolitan area in the Midwest, was vention for the prevention of neglect is restricted to children who were 11 years or 0 A comparison group of 667 children, described later in this Research in Brief. younger at the time of the incident of abuse not officially recorded as abused or ne- or neglect. At the time that juvenile and glected, matched to the study group ac- The first phase of this study relied on arrest criminal records were checked, subjects cording to sex, age, race, and approximate records to measure delinquency and crimi- ranged in age from 16 to 33; most were family socioeconomic status. nality. A second phase calls for locating

I From the Director Family violence-particularly violence methods have been considered for estimat- year dealing with the consequences of child against children-is a critical priority for ing its extent. None has examined its effect abuse. In addition, NU will support a criminal justice officials, political leaders, on the later behavior of children as does the multisite study in this fiscal year of child and the public we serve. Statistics indicate NIJ study reported in this Research in Brief. abuse prosecution and a study on how the that the Nation experiences around a half- justice system has addressed this critical million instances of family abuse annually; Some of the findings are startling. For ex- problem. more than 20,000 of these cases involve ample, being abused or neglected as a child children as victims. increased the likelihood of arrest as a juve- nile by 53 percent, as an adult by 38 per- Family violence can be considered from a cent. andfor a violent crime by 38 percent. Charles B. DeWitt variety of different perspectives: criminal Director justice, psychology, sociology, and econom- I have made a priority at NIJ, National Institute of Justice ics. Studies have produced varying estimates and this is the first in a series of five Re- of the magnitude of family violence; various search in Brief reports NIJ will publish this between ages 20 and 30, with a mean age sexual desires," "fondling or touching in swer, shown in exhibit 3, was evident: of 25. an obscene manner," rape, sodomy, and those who had been abused or neglected as incest. children were more likely to be arrested as Matching members of the study group to juveniles (26 percent versus 17 percent), as others whose official records showed no Neglect cases represented extreme fail- adults (29 percent versus 21 percent), and childhood abuse or neglect was an equally ure to provide adequate food, clothing, for a violent crime (1 1 percent versus 8 important feature of the research. This shelter, and medical attention to children. percent). The abused and neglected cases design allowed the study to separate the Family members (often parents) were the were also more likely to average nearly 1 effects of known correlates of delinquency primary perpetrators of the abuse and year younger at first arrest (16.5 years and criminality (age, sex, race, and socio- neglect. The most frequent type of perpe- versus 17.3 years), to commit nearly twice economic status) from the experience of trator varied. however. by type of maltreat- as many offenses (2.4 percent versus 1.4 abuse and negiecr. Both groups were ap- ment (see exhibit 2). percent), and to be arrested more proximately two-thirds white and one-third frequently (17 percent of abused and ne- black and were about evenly divided be- Juvenile court and probation records were glected cases versus 9 percent of compari- tween males and females. Most were be- the source of information on the abuse and son cases had more than five arrests). tween 6 and 11 years old at the time the neglect, as well as the characteristics of the abuse was documented (see exhibit 1). family. Arrest data were obtained from Sex. Experiencing early child abuse or Federal, State, and local law enforcement neglect had a substantial impact, even on The study design also featured clear opera- records. Recognizing that much child individuals with little likelihood of engag- tional definitions of abuse and neglect. abuse (as well as later delinquent and ing in officially recorded adult criminal Combined with large sample sizes, this criminal behavior) never comes to the behavior. Thus, although males generally permitted the separate examination of attention of any official authority, Phase I1 have higher rates of criminal behavior than physical abuse, sexual abuse, and neglect, will supplement these official records with females, being abused or neglected in defined as follows: interview results. childhood increased the likelihood of arrest Physical abuse cases included injuries for females-by 77 percent over compari- such as bruises, welts, bums, abrasions, Study findings son group females. As adults, abused and lacerations, wounds, cuts, bone and skull neglected females were more likely to be fractures, and other evidence of physical Of primary interest was the question, arrested for property, drug, and misde- injury. "Would the behavior of those who had meanor offenses such as disorderly con- duct, curfew violations, or loitering, but Sexual abuse involved such charges as been abused or neglected be worse than those with no reported abuse?" The an- not for violent offenses. Females in general "assault and battery with intent to gratify are less likely to be arrested for street violence and more likelv to amear in sta- tistics on violence in the homk: Through Exhibit 1. Demoara~hicCharacteristics interviews, Phase I1 will examine the inci- dence of unreported violence to learn more about the possible existence of hidden cycles of family violence. Race. Both black and white abused and neglected children were more likely to be 70 AbusedINeglected Groups " 1 arrested than comparison children. How- -ison Groups ever, as shown in exhibit 4, the difference between whites was not as great as that between blacks. In fact, white abused and neglected children do not show increased likelihood of arrest for violent crimes over comparison children. This contrasts dra- matically with the findings for black chil- dren in this sample who show significantly increased rates of violent arrests, compared with black children who were not abused or neglected. This is a surprising finding and one that may reflect differences in an array of environmental factors. Phase I1 will investigate a number of explanations for these results, including differences in poverty levels, family factors, characteris- tics of the abuse or neglect incident, access Males Females White Black 0-3 yrs 4-5 yrs 6-11 yrs to counseling or support services, and treatment by juvenile authorities. 7- Juvenile record. Previously abused or " neglected persons were at higher risk of Exhibit 2. Perpetrators of Abuse and~e~lecl beginning a life of crime, at a younger age, with more significant and repeated crimi- nal involvement. Notably, however, among those arrested as juveniles, abused PHYSICAL ABUSE or neglected persons were no more likely Mother to continue a life of crime than other Father children: Stepmother 0 In both groups, roughly the same pro- Stepfather P~GXof children with jwenile &i&S also had arrests as adults (53percent ver- Grandmother sus 50 percent). Grandfather 0 Similarly, in both groups, about the Related Adult same proportion of those with violent juve- Legal Guardian nile arrests also had violent arrests as Known Adult adults (34.2 percent versus 36.8 percent). In short, childhood abuse and neglect had no apparent effect on the movement of Yo juvenile offenders toward adult criminal activity. Distinguishing the factors that promote the onset of criminal behavior from those that affect persistence in a SEXUAL ABUSE criminal career is clearly an important Mother topic for future research.

Stepmother Does only violence Stepfather beget violence? Grandmother To test the notion that childhood victims of Grandfather violence resort to violence themselves in Related Adult later years, violent criminal behavior was Legal Guardian examined as a function of the type of mal- - treatment experienced as a child. The Known Adult I results are presented in simplified form Unknown Adult below. Percent Arrested % for Violent Abuse Group Number Offense Physical abuse NEGLECT only 76 15.8% Mother Neglect only 609 12.5 Father Physical abuse Stepmother and neglect 70 7.1 Stepfather Sexual abuse and other Grandmother abuse or neglect 28 7.1 Grandfather Sexual abuse only 125 5.6 Related Adult Comparison group 667 7.9 Legal Guardian Known Adult The physically abused (as opposed to neglected or sexually abused) were the Unknown Adult most likely to be arrested later for a violent crime. Notably, however, the physically abused group was followed closely by the neglected group. 6 ^ ment and adult criminality, New York City Exhibit3. Extent of Involvement in Delinquency, Adult Criminality, instituted new procedures for police re- and Violent Criminal Behavior sponse and followup in cases involving suspected child abuse and negle~t.~ Abused and Neglected Comparison Group (n = 908) (n = 667) Out-of-home placement and criminal consequences Type of arrest (%I (w Not all abused and neglected children Juvenile 26.0 16.8 grow up to become delinquents, adult Aduit 28.6 21.1 nminals, or violent criminal offenders. Violent crime 11.2 7.9 What are some of the possible mediating variables that act to buffer or protect abused and negkcted children? Placement Note: All differences significant. outside the home is one possible buffer that was investigated with Phase I data. Scholars and practitioners have often criti- cized out-of-home placements (foster care, Exhibit 4. Involvement in Criminality by Race in particular). Children placed outside the home are considered a particularly vulner- able group, since they have experienced Abused and Neglected Comparison Group Significance both a disturbed family situation and sepa- (n = 908) (n a 667) ration from their natural parents. Accord- ingly, child welfare policies today often seek to avoid removing the child from Juvenile home and instead to mitigate negative Black 37.9 family situations through counseling and related support. White 21.1 In contrast to today's practices, the vast Adult majority of a sample of the children abused Black 39.0 and neglected roughly 20 years ago were placed outside the home during some White 24.4 portion of their childhood or early adoles- cence. Year-by-year information was Violent available from juvenile court and probation Black 22.0 records on 772 cases. For these children, White 6.5 out-of-home placements included foster care, guardian's home, and schools for the retarded or physically handicapped. Only 14 percent of these abuse and neglect cases had no record of having been placed up through age 18. The average amount of Because different types of abuse and ne- (provided the abuse involves no neurologi- time in placement was about 5 years, and glect are not distributed evenly by age, cal impairment). In one study of the influ- sometimes lasted through childhood and race, and sex, these frequencies present an ence of early malnutrition on subsequent adolescence. oversimplified picture. Even after control- behavior, previously malnourished chil- As exhibit 5 shows, there was remarkably ling for age, race, and sex, however, a dren had attention deficits, reduced social little difference between the arrest records relationship between childhood neglect skills, and poorer emotional stability than a of those who remained at home and those and subsequent violence remained evident. comparison group.3 Other researchers have who were placed outside the home due to found an array of developmental differ- This finding offers persuasive evidence for abuse and neglect. (Predictably, both of ences associated with childhood negle~t.~ the need to take concerted preventive these groups were strikingly different from This study now suggests that those differ- action. Nationwide, the incidence of ne- those placed outside the home due to delin- ences include a greater risk of later crimi- glect is almost three times that of physical quency as well as abuse and neglect.) At nal violence. abuse (15.9 per 1,000 children in 1986, least for this sample, then, an out-of-home compared to 5.7 per 1,000 for physical Research findings show how imperative placement did not lead to negative effects abuse, and 2.5 per 1,000 for sexual ab~se).~ are improved procedures for the identifica- on the arrest measure for those who were Neglect also is potentially more damaging tion of child abuse and neglect. Refemng removed from their homes due only to to the development of a child than abuse to the connection between child maltreat- abuse and neglect. The study also showed that stability may t be an important factor in out-of-home Exhibit 5. Juvenile and Adult Arrests as a Function of Placement placements. Children who moved three or Experiences for Juvenile Court Cases Only (n = 772) more times had significantly higher arrest rates (almost twice as high) for all types of Arrest (in percent) criminal behaviors-juvenile, adult, and violent-than children who moved less Type of N Any Any Both Juv. Any than three times. In turn, children with Placement Juvenile Adult & Adult Violent multiple placements typically had behavior (n=209) (n=217) (n=i i5) (n=93) problems noted in their files. These nota- tiens cevered I wide specmm of problem No placement 106 15.1 29.2 6.6 10.4 behavior, including chronic fighting, fire Abuselneglect 489 17.8 23.3 8.6 8.4 setting, destructiveness, uncontrollable placement only anger, sadistic tendencies (for example, aggressiveness toward weaker children), Delinquency 96 92.7*** 60.4*** 55.2*** 34.4*** and extreme defiance of authority. placement plus Whether the behavior problems caused the abuselneglect moves, or the moves contributed to the behavior problems, is unclear. In either case, children with numerous placements Note: Adult arrest rates restricted to subjects age 21 and older in March 1988. obviously need special services. *** p.001 These findings challenge the assumption that it is necessarily unwise to remove children from negative family situations. While stability of placement appears to be children. The followup study aims to ex- results of abuse and neglect. The ultimate important, the potential damage of remov- amine the full consequences of maltreat- goal is to provide a base of knowledge on ing an abused and neglected child from the ment as a child and to determine why some which to build appropriate prevention and home did not include a higher likelihood of victims of childhood abuse and neglect treatment programs. arrest or violent criminal behavior. fare well, while others have negative out- comes. The interviews will explore recol- Conclusion and implications Phase II: Followup and lections of early childhood experiences, schooling, adolescence, undetected alcohol Childhood victimization represents a wide- in-person interviews and drug problems, undetected delin- spread, serious social problem that in- While the findings from Phase I demon- quency and criminality, and important life creases the likelihood of delinquency, adult strate convincingly that early child abuse experiences. criminality, and violent criminal behavior. and neglect place one at increased risk for Preliminary Phase I1 findings, based on Poor educational performance, health officially recorded delinquency, adult 2-hour followup interviews with 500 study problems, and generally low levels of criminality, and violent criminal behavior, and comparison group subjects, indicate achievement also characterize the victims a large portion of abused and neglected that other negative outcomes may be as of early childhood abuse and neglect. children did not have official arrest common as delinquency and violent crimi- This study offers at least three messages records. Indeed, the linkage is far from nal behavior. These interviews suggest that to juvenile authorities and child welfare inevitable, since the majority of abused the long-term consequences of childhood professionals: and neglected children did not become victimization also may include: delinquents, adult criminals, or violent Intervene early. The findings of Phase I offenders. However, because the findings ' (depressionand issue a call to police, teachers, and health from Phase I were based on official arrest suicide attempts). workers for increased recognition of the records, these rates may be underestimates Educational problems (inadequate cog- signs of abuse and neglect, and serious ef- of the true extent of delinquency and nitive functioning, extremely low IQ, and forts to intervene as early as possible. The criminality. Phase I findings also do not poor reading ability). later the intervention, the more difficult the tell us about general violent behavior, change process becomes. Specialized at- especially unrecorded or unreported family Health and jssues and tention needs to be paid to abused and ne- violence. drug problems). glected children with early behavior problems. These children show the highest 0 Occupational difficulties (lack of work, phase 11 was designed to address many of risk of later juvenile and adult arrest, as employment in low-level service jobs). the unanswered questions from the first well as violent criminal behavior. phase by finding and a large In addition to documenting the broader Develop policies that recognize the hi&, number these people 20 years after the consequences of childhood victimization, risks of neglect as well as abuse. Also im- childhood victimization. Most are now Phase I1 is geared to identify "protective" portant in its implications for juvenile young adults in their early 20's and 30's; factors that mav act to buffer the negative some are beginning to have their own u court and child welfare action is the fact that neglect alone (not necessarily physical Notes abuse) was significantly related to violent criminal behavior. A picture emerges 1. For further information on the design where physical abuse is only one point on and sampling procedures, see Widom, a continuum of family situations that con- C.S., "Child abuse, neglect, and adult tribute to violence. Whether those situa- behavior: Research design and findings on tions result in active physical abuse, or criminality, violence, and child abuse," more passive neglect, it is now quite clear American Journal of Orthopsychiatry, that both foms of child maltreatment are 59(1989):355-367. serious threats. Neglect cases represent the 2. Westat, Inc. Study Findings: Study of majority of cases taxing the child protec- Findings and conclusions of the research re- -National Incidence and Prevalence of ported here are tinose of the researcher and do tion system. Research shows that today's Child Abuse and Neglect: 1988, Washing- victim of neglect may well be a defendant not necessarily reflect the official position or ton, D.C., U.S. Department of Health and policies of the U.S. Department of Justice. in tomorrow's violent criminal case. Human Services. 0 Reexamine out-of-home placement 3. J.R. Galler, F. Ramsey, G. Solimano, The National Institute ofJustice is a compo- policies. This NIJ study focused on cases and W.E. Lowell, "The influence of early during the period 1967-197 1, when out- nentof the Ofice ofJustice Programs, which malnutrition on subsequent behavioral also includes the Bureau of Justice Assist- of-home placements were a common inter- development: II. Classroom behavior," ance, Bureau of Justice Statistics, Oflce of vention. Detailed information available for Journal of the American Academy of Child Juvenile Justice and Delinquency Preven- 772 cases revealed that the vast majority Psychiatry, 24(1983):16-24. tion, and the Office for Victims of Crime. (86 percent) were placed outside their homes for an average of 5 years. This con- 4. See, for instance, R.E. Allen and J.M. trasts sharply with today's efforts to avoid Oliver, "The effects of child maltreatment NCJ 136607 out-of-home placement on the assumption on language development," Child Abuse that separation may aggravate, rather than and Neglect, 6(1982):299-305; B. ameliorate, a child's problems. Yet, there Egeland, A. Sroufe, and M. Erickson, "The was no evidence that those who were sepa- developmental consequences of different rated from their families fared any worse patterns of maltreatment," Child Abuse on the arrest measures than those who re- and Neglect, 7(1983):459-469; A. Frodi mained at home. Though these results are and J. Smetana, "Abused, neglected, and far from definitive, they do suggest that nonmaltreated preschoolers' ability to child protective policies in this area de- discriminate emotions in others: The ef- serve close scrutiny. The assumption that fects of IQ," Child Abuse and Neglect, removal from the home offers additional 8(1984):459465. risk could not be confirmed by this study. 5. Benjamin Ward, Commissioner, New Any policy founded on this assumption York City Police Department, press release ought to be tested through careful local No. 17, May 22, 1989. studies of the full consequences of out- of-home placement.

U.S. Department of Justice I BULK RATE Office of Justice Programs POSTAGE & FEES PAID National Institute of Justice DOJ/NIJ Perm~tNo. G-9 1 I Washington, D.C.20531

Official Business Penalty for Private Use $300