OJJ D P B J S O G JUST I C E P R National Institute of Justice

1 Department of JusticeOffice of Justice ProgramsNational Institute of JusticeR e s e a r c h i n B r i e fNational Institute of JusticeJeremy Travis, DEPARTMENTOFJUSTICE OFFICEOFJUSTICEPROGRAMSBJANIJOJJDPBJSOVC July 1999 Issues and FindingsDiscussed in this Brief: Findingsabout partner violence from the lon-gitudinal Dunedin MultidisciplinaryHealth and Development Study of arepresentative birth cohort of 1,037 New Zealand men and womenbornbetween April 1, 1972, andMarch 31, issues: Dunedin researchersdecided to study partner violenceas part of their longitudinal studybecause they suspected that somestudy members continued antiso-cial activities into adulthood in aform not well represented bymeasures of delinquent were included in thephase 21 (age 21) interview todetermine the extent of partnerviolence among study findings: Characteristics ofcohort members who were in-volved in partner violence includethe following.

2 LAlthough both partners in a re-lationship may not recall the sameacts in precisely the same way, 70 80 percent of one partner s reportwas in agreement with the otherpartner s report on whether physi-cal violence took place and on theextent of the factors in childhood andadolescence for male perpetratorsincluded poverty and low academicachievement. Female perpetratorsshowed risk factors of harsh familydiscipline and parental strife. Bothmale and female perpetrators alsohad histories of aggressive About Partner ViolenceFrom the Dunedin MultidisciplinaryHealth and Development Studyby Terrie E. Moffitt and Avshalom CaspiThe 1992 96 National Crime Victimiza-tion Survey (NCVS) indicates that in1996 victimization by intimates spouses, ex-spouses, boyfriends, girl-friends, and former boyfriends andgirlfriends accounts for about 21 per-cent of the violent crime experienced bywomen and about 2 percent of the vio-lence experienced by men.

3 Rates of non-lethal violence are highest among womenages 16 24 and women in These statistics highlightthe importance of studying partner vio-lence among young adults, both marriedand unmarried, who represent allsocioeconomic Dunedin Multidisciplinary Healthand Development Study is a 21-yearinvestigation of a representative birthcohort of infants born between April 1,1972, and March 31, 1973, in Dunedin, acity of approximately 120,000 people onNew Zealand s South Perinataldata for 1,139 births were obtained at de-livery. When these children were tracedat age 3, 91 percent, or 1,037, were as-sessed, forming the base sample for thelongitudinal study. The base sample wascomposed of 535 (52 percent) males and502 (48 percent) females.

4 Fewer than7 percent of the study members identifiedthemselves as nonwhite (Maori orPolynesian). The social class andethnicity of their families matched thoseof the South Island s general Dunedin cohort was reassessed atages 5, 7, 9, 11, 13, 15, 18, and 21, with992 (97 percent) of the 1,020 living mem-bers of the age 3 base sample (51 percentmale) participating in the last assessmentat age 21 in 1993 94. This history ofreassessment at regular intervals is espe-cially important for the research on part-ner violence. Since their childhood, studymembers have revealed problem behav-iors; their confidentiality has never beenviolated, and they have learned to expectno intervention from the researchers (un-less imminent serious danger was posedto themselves or others).

5 As a result, bythe time they were 21 years old, theywere comfortable giving frank responsesto questions about partner violence. Thiscircumstance offers a special advantageover studies involving self-reports fromperpetrators held in correctional institu-tions or remanded for Research in Brief summarizes thefindings about partner violence from theDunedin study. Among the results arethat partner violence is strongly linked tocohabitation at a young age; a variety ofmental illnesses; a background of family2R e s e a r c h i n B r i e fIssues and , dropping out of school, andjuvenile aggression; conviction forother types of crime, especially violentcrime; drug abuse; long-term unem-ployment; and parenthood at a study partner violence?

6 The Dunedin study began to examinepartner violence because partner violenceresearch is a natural extension of thestudy s earlier research on childhood be-havior problems and teen official crime statistics frompolice and courts suggest that offendingdeclines rapidly during young adulthood,Dunedin researchers suspected that somedelinquents antisocial activities werecontinuing, but in a form not easily de-tected by official crime statistics; for ex-ample, as abuse of family members orintimates in the home. To test this possi-bility, a decision was made to examinepartner violence among study members asthey made the transition to young adult-hood, became involved in serious rela-tionships, and began to form new the unique features of this longitu-dinal study, it became important to applythe study s resources to understanding theorigins of partner Dunedin cohort, it is important tonote, is a birth cohort, not a commu-nity sample.

7 Followup was conductedwith all individuals in the cohort; forexample, those who had and had notused battered women s shelters andthose who had and had not been con-victed of the study members turned 21years old (phase 21 of the DunedinlThe strongest risk factor for bothmale and female perpetrators andvictims was a record of physicallyaggressive delinquent offendingbefore age 15. More than half themales convicted of a violent crimealso physically abused their 27 percent of women and34 percent of men among theDunedin study members reportedthey had been physically abused bytheir partner. About 37 percent ofwomen and 22 percent of men saidthey had perpetrated the violence is most preva-lent among cohabitating percent of females whosuffered serious physical abuse and88 percent of male perpetratorshad one or more mental disorders(as defined by the American Psychi-atric Association in the Diagnosticand Statistical Manual of theAmerican Psychiatric Association[ DSM III R ]).)

8 LWomen who had children byage 21 were twice as likely to bevictims of domestic violence aswomen who were not who had fathered children byage 21 were more than three timesas likely to be perpetrators of abuseas men who were not audience: Mental healthpractitioners; emergency room andgeneral practice medical profes-sionals; victim advocates; juveniledelinquency, substance abuse, andviolence specialists and researchers;public health, juvenile Justice , andcriminal Justice officials and practi-tioners; juvenile court administra-tors; judges; social scientists;researchers; and others concernedabout violence , Incidence, and Conse-quences of Violence Against Women:Findings From the National ViolenceAgainst Women Survey (Research inBrief), by Patricia Tjaden and NancyThoennes, NCJ 172837 Stalking in America: FindingsFrom the National Violence AgainstWomen Survey (Research in Brief),by Patricia Tjaden and NancyThoennes, NCJ 169592 Batterer Programs: What CriminalJustice Agencies Need to Know (Re-search in Action), by Kerry MurphyHealey and Christine Smith, NCJ171683 Legal Interventions in Family Vio-lence.

9 Research Findings and PolicyImplications (Research Report), byNIJ and the American Bar Associa-tion, NCJ 171666 Other Related Publications Supported by the NationalInstitute of Justice and the Bureau of Justice StatisticsUnderstanding Violence AgainstWomen, by Nancy A. Crowell andAnn W. Burgess (eds.), NationalAcademy PressViolence in Families: Assessing Pre-vention and Treatment Programs, byRosemary Chalk and Patricia King(eds.), National Academy PressViolence by Intimates: Analysis ofData on Crimes by Current or FormerSpouses, Boyfriends, and Girlfriends(Bureau of Justice Statistics Factbook),by Lawrence A. Greenfeld, MichaelR. Rand, Diane Craven, Patsy , Craig A. Perkins, CherylRingel, Greg Warchol, Cathy Maston,and James Alan Fox, NCJ 167237 Publications that have NCJ numberscan be obtained from the NationalCriminal Justice Reference Serviceby calling 1 800 851 e s e a r c h i n B r i e faFrom Margolin s Domestic Conflict Scale or Conflict Inventory (Margolin, G.)

10 , B. Burman, John, and M. O Brien, The Domestic ConflictInstrument, Los Angeles: University of Southern California, 1990).bFrom Straus s Conflict Tactics Scales (Straus, , Measuring Intrafamily Conflict and Violence: The Conflict Tactics Scales, Journal ofMarriage and the Family 41(1979): 75 88).Exhibit 1. Item content of physical and psychological partner abuse scalesstudy), questions about partner vio-lence were embedded in a 50-minutestandardized interview about intimaterelationships. Information was gath-ered on both positive and negativeconflict-negotiation behaviors occur-ring during the past 12 months. In-cluded in the interview were the itemsof the Conflict Tactics Scales (CTS)3and other items from published do-mestic violence interviews.