Shay Bilchik, Administrator   FACT SHEET # 21  December 1994.

Violent Families and Youth Violence
Terence P. Thornberry

Violent Families and Youth Violence
Compared to other industrialized nations, America's rates of criminal violence
are unacceptably high.  Pervasive violence adversely affects our streets,
schools, work places, and even our homes. 

While we have come to recognize the extent of family violence, we know much
less about its consequences, particularly its effects on children growing up
in violent families.  This fact sheet examines this issue for one outcome,
involvement in violent behavior during adolescence. It addresses two
questions.

First, are children who are victims of maltreatment and abuse during childhood
more apt to be violent when they are adolescents?  And second, are children
who are exposed to multiple forms of family violence not just maltreatment
more likely to be violent?

Methods
Data from the Rochester Youth Development Study are used in this analysis. 
This ongoing study of delinquency and drug use began with 1,000 7th and 8th
grade students attending the public schools of Rochester, New York in 1988. 
Youngsters at high risk for serious delinquency were oversampled, but the data
presented here are weighted to represent the cohort of all 7th and 8th
graders.  The youths and their primary caretaker were interviewed every six
months until the adolescents were in the 11th and 12th grades.  Students who
left the Rochester schools were also contacted.  The overall retention rate
was 88 percent.  In addition to personal interviews, the project collected
data from schools, police, social services, and related agencies.1

Delinquency is measured by self-reports of violent behavior.   Every six
months the interviewed youths indicated their involvement in six forms of
violent behavior, ranging from simple assault to armed robbery and aggravated
assault.  The measure used in this analysis is the cumulative prevalence of
such behavior over the course of the interviews.

Child Maltreatment and Delinquency
Practitioners and researchers have long been interested in whether early
childhood victimization is a significant risk factor for later involvement in
violence.  To examine this issue,  information was collected on maltreatment
from the Child Protective Service files of the Monroe County, New York,
Department of Social Services for all study subjects.  Maltreatment includes
substantiated cases of physical or sexual abuse or neglect.  To examine prior
victimization as a risk factor for later violence, we have considered only
those instances of maltreatment that occurred before age 12.

Sixty-nine percent of the youths who had been maltreated as children reported
involvement in violence as compared to 56 percent of those who had not been
maltreated (Figure 1).2  In other words, a history of maltreatment increases
the chances of youth violence by 24 percent.
Other analyses of these data indicate that maltreatment is also a significant
risk factor for official delinquency and other forms of self-reported
delinquency; for the prevalence and frequency of delinquency; and for all
these indicators when gender, race/ethnicity, family structure, and social
class are held constant.3

Multiple Family Violence
If direct childhood victimization increases the likelihood of later youth
violence, does more general exposure to family violence also increase the
risk?  To address this question, three different indicators of family violence
were examined:  partner violence, family climate of hostility, and child
maltreatment.

Partner violence was measured by the Violence Subscale of the Conflict Tactics
Scale (Straus, 1988).  It was based on parent interview data and indicates the
level of violence between the subject's parent and his or her spouse.4  The
family climate of hostility  scale   also taken from the parent interview  
measures the extent to which there was a) generalized conflict in the family,
and b) family members physically fought with one another.  The child
maltreatment measure is similar to the one used earlier, but now includes
cases of maltreatment in which any children in the subject's family are
victimized, not just the study participant.

Figures 2 through 4 demonstrate that, for each type of family violence,
adolescents who live in violent families have higher rates of self-reported
violence than do youngsters from non-violent families.  The results for
partner violence illustrate this finding. Seventy percent of the adolescents
who grew up in families where the parents fought with one another
self-reported violent delinquency as compared to 49 percent of the adolescents
who grew up in families without this type of conflict.  Similar patterns can
be seen for the other two indicators of family violence.

The final issue we examined was the consequences of growing up in families
experiencing multiple forms of violence (Figure 5).  While thirty-eight
percent of the youngsters from non-violent families reported involvement in
violent delinquency, this rate increased to 60 percent for youngsters whose
family engaged in one of these forms of violence, to 73 percent for those
exposed to two forms of family violence, and further increased to 78 percent
for adolescents exposed to all three forms of family violence.  Exposure to
multiple forms of family violence, therefore, doubles the risk of
self-reported youth violence.

Summary
This analysis examined the relationship between family violence and youth
violence.  Adolescents who had been direct victims of child maltreatment are
more likely to report involvement in youth violence than non-maltreated
subjects.  Similarly, adolescents growing up in homes exhibiting partner
violence, generalized hostility, or child maltreatment also have higher rates
of self-reported violence.  The highest rates were reported by youngsters from
multiple violent families.  In these families, over three-quarters of the
adolescents self-reported violent behavior.  In other words, children exposed
to multiple forms of family violence report more than twice the rate of youth
violence as those from nonviolent families.

References
1 See David Huizinga, Rolf Loeber, and Terence Thornberry, Urban Delinquency
and Substance Abuse. Washington, DC: U.S. Department of Justice, Office of
Juvenile Justice and Delinquency Prevention, 1993. A technical report is also
available that provides a detailed discussion of sampling, attrition, and data
collection procedures.

2 This relationship is significant at the .01 level.  All other relationships
reported in this paper are significant at the .001 level.

3 Carolyn Smith and Terence P. Thornberry, The Relationship Between Childhood
Maltreatment and Adolescent Involvement in Delinquency and Drug Use.  Working
Paper No. 17 . Criminal Justice Center. State University of New York at
Albany.

4 If the subject's parent did not have a spouse or other partner, Partner
Violence equals zero since the adolescent was not exposed to this type of
family violence, at least during the course of our study.


This Fact Sheet was prepared by Dr. Terence P.Thornberry, Professor, School of
Criminal Justice, State University of New York at Albany.  He is also Director
of the Rochester Youth Development Study.  The work was supported by OJJDP
grant #86-JN-CX-0007.  The Rochester Youth Development Study is one of three
coordinated research projects carried out under OJJDP's Program of Research on
Causes and Correlates of Delinquency. 
FS-9421