study36 finding violence victims some-
what younger, one finding them some-
what older,20 and others21,24 finding no
difference in the ages of abused and non-
abused pregnant women.
Marital Status. Not surprisingly,
women who are separated or divorced
are far more likely to have experienced
domestic violence.7,12-14,18,20,21,34 Several
studies have found single women more
likely to be abused by an intimate than
married women,12,23,32,33 but others have
found single women less likely to be
abused.20,21 Perhaps married women are
more reluctant to admit being abused.
Socioeconomic Status. Although some
studies have found no significant differ-
ences, domestic violence has generally
been found to be more common among
couples of lower socioeconomic status,
as measured by education, income, occu-
pation, or other measures. Some studies
have reported an inverse association
between domestic violence and years of
education,17,20,35 while others have reported
that the highest prevalence is among
individuals with some high school educa-
tion.6,7,10,33 Campbell37 reported no dif-
ferences in education levels between
battered and nonbattered women in
“troubled families.” Results about the
effect of family income were consistent:
income was inversely related to preva-
lence of domestic violence.6,13,17,21,32,35,38
Some studies reported occupational data,
and, in general, the prevalence of domes-
tic violence is somewhat higher in blue
collar occupations,5,8,17 although not con-
sistently,7 and among the unemployed.5,21
Studies that used other indicators of
socioeconomic status also confirmed that
domestic violence is more of a problem
for families in the lower socioeconomic
groups.23,25,39,40
Race and Ethnic Group. Results of
analyses of race and ethnicity are less
clear than for other demographic vari-
ables. Some studies5,7,28,37 reported higher
prevalence among minorities or African-
Americans. In contrast, the National
Crime Surveys have found similar rates
of abuse for African-Americans and
whites,12-14 and Amaro23 found pregnant
African-American women less likely to
experience violence than pregnant white
women. White non-Latinas have been
found to experience more domestic vio-
lence than Latinas.21,32 Other investiga-
tors13,33,35 have found no differences
among racial or ethnic groups.
In light of the consistent differences in
domestic violence according to socioeco-
nomic groups, and the higher proportion
of racial and ethnic minorities in lower
socioeconomic groups, it is important to
control for socioeconomic status when
examining the influence of race or eth-
nicity. Two studies that exercised such
controls found that race was not an inde-
pendent factor. Lockhart39 analyzed the
relationship between violence, socioeco-
nomic status, and race, and found that
among upper or lower class women,
whites were somewhat more likely to
experience violence, while in the middle
class, blacks were more likely to experi-
ence violence. Centerwall40 found that
the higher rates of domestic homicides in
black census tracts could be explained by
the effect of income.
Influence of Alcohol and
Drug Use on Domestic Violence
Several studies have investigated the
relationship between substance use and
domestic violence against women. Gen-
erally, frequent alcohol use by the male
partner has been found to be associated
with increased likelihood of violence in
the home.6,23,35 Alcohol use by the woman
has also been associated with increased
likelihood of violence,21,23 but not consis-
tently,35 and not in all subgroups.21 Each
of these studies used multivariate analyses
to assess the effect of alcohol use. Another
study41 found the husband’s alcoholism
to be an important problem for abused
women receiving individual psychological
treatment, but not for abused women in
conjoint therapy.
While these studies focused on the
usual pattern of alcohol use, others have
examined the use of alcohol at the time
of the violent incident. Results are mixed:
Straus and Gelles6 found that in 76% of
instances, neither member of the couple
was drinking. One-half of Texas women
who had been abused in the previous
year (including those sexually and psy-
chologically abused) reported that abuse
took place when the male partner was
drinking.16 Eberle42 studied women who
had experienced four or more batterings
and found that male partners of 81% of
the women used alcohol at the time of at
least one of the incidents. Only 16% of
the women reported that the man had
used alcohol at the time of each battering.
A recent review by Leonard43 concluded
that in spite of the methodological weak-
nesses of many studies on alcohol and
family violence, the data “strongly sug-
gest that alcoholism and acute alcohol
consumption . . . are related to marital
violence. The research indicates that a
high percentage of those who engage in
marital violence are alcoholics or at least
very heavy drinkers, and that the marital
violence is often associated with alcohol
consumption.”
Discussion
A number of methodological factors help
to explain the variability in prevalence
found in different studies. Definitions of
domestic violence, time frames for the
violent event, populations studied, and
methods of data collection have all var-
ied. Considering these factors, there is
actually considerable agreement on rates.
Careful attention to the variables that
influence prevalence rates has been lack-
ing. As a result both the popular and
professional literatures have misrepre-
sented domestic violence study findings,
frequently overstating the extent of the
problem. For example, the American
Medical Association’s Diagnostic and
Treatment Guidelines on Domestic
Violence,44 states that 22% to 35% of
women seeking care for any reason in
emergency rooms may be battered. One
study of women seen in an emergency
room for any reason found that 24%
were battered.33 Two studies found 30%
to 35% of women with trauma injuries
had been battered.29
Systematic research in domestic vio-
lence is relatively new, and many of the
studies reviewed here were done among
“convenience” samples. Confidence
intervals for prevalence estimates were
generally not reported, and few studies
used multivariate techniques.
Nevertheless, the prevalence of domes-
tic violence among intimate partners in
America is quite high. Population-based
studies report that 8% to 12% of women
experience some form of domestic vio-
lence in a one-year period. Severe vio-
lence with the potential to cause injury is
experienced by 2% to 4% in a year.
Pregnant women do not appear to be
more vulnerable, despite their younger
age and lower socioeconomic status in
most of the studies reviewed. A substan-
tial proportion of injuries among women