Domestic Violence Studies

This site is copyrighted, supported, and maintained by the Equal Justice Foundation.


 

| EJF Home | Join the EJF | Comments? | Get EJF newsletter | Newsletters |

| DV Home | Abstract | Contents | Authors and Site Map | Tables | Index | Bibliography |

 

| Chapter 4 — Domestic Violence Statistics |

| Next — Methodological Issues: Archival Data vs. Scientific Research by Tom James, Esq. |

| Back — Reading between the numbers |


 

Contents

Studies supported by the National Institute of Mental Health

Pioneering work of Straus, Gelles, and Steinmetz

Causes of domestic violence

National violence against women survey

Domestic violence in the military

Domestic violence as a function of age

Extreme female violence and age

Violence and senility

Actual domestic violence cases in Colorado

Impact on jury system


 

Studies supported by the National Institute of Mental Health

Pioneering work of Straus, Gelles, and Steinmetz

The most widely reported figures, and probably the most reliable, on domestic violence and abuse are based on social studies by Straus and others supported by the National Institute of Mental Health. They report that roughly 6 million women a year are victims of some level of domestic violence in the United States. Of these 6 million women, 1.8 million are seriously assaulted every year. These estimates are based on surveys made in 1975, 1985, and 1992 (see Table 8).

However, it is widely ignored that the same studies consistently found that an equal number of men were experiencing domestic violence and an even greater number, 2.2 million men per year, are being seriously assaulted by their female partners.

Note, however, that only a small fraction of these men and women think of the events social scientists label family, or domestic violence as a crime.

As of 2002 there have been over one-hundred-and-thirty scientific studies (Straus, 1999) supporting the pioneering work of Straus, Gelles, and Steinmetz. Their results are among the most replicated of all studies in the social sciences. See, for example, the bibliography compiled by Fiebert.

All available evidence, e.g., Table 8 based on the data of Straus and Gelles (1986, p. 470) and Straus and Kantor (1994), shows a clear trend toward decreasing levels of interpersonal violence. However, the decrease of male on female violence has been more marked and continuous than the drop in female on male assaults and killings. Also, in some categories female on male aggression may actually be increasing.


 
    Table 8: Rate of violence per 1,000 couples stated as a percentage for the years indicated from social studies .

Husband against wife

1975

1985

1992

As reported by

(for 1992 data)

Overall violence

12.1%

11.3%

11.5%

Wife

8.3%

Husband

Severe violence

3.8%

3.0%

2.3%

Wife

1.7%

Husband

Wife against husband

1975

1985

1992

As reported by

(for 1992 data)

Overall violence

11.6%

12.1%

9.9%

Wife

9.1%

Husband

Severe violence

4.6%

4.4%

5.8%

Wife

3.3%

Husband

Severe violence: To kick, bite, or hit with a fist; burn; scald; to hit or try to hit with an object; to beat up the other; to threaten with a knife, gun, or other deadly weapon; to use a knife, gun, or other deadly weapon.

Top


 

Straus and Gelles followed up their original studies in National Family Violence Surveys with nationally representative samples of 2,143 married and cohabiting couples in 1975 and 6,002 couples in 1985. These data are available from Sociometrics.

Clearly the level of overall violence shown in Table 8 is roughly equal between men and women at ~12% in 1975, decreasing to about 10% by 1992. Note, however, that Table 8 indicates a statistically significant higher level of severe violence by wives against their husbands and other studies consistently support their findings (see the report by the Revs. Sewell ). Note that the researchers report on the actions of the couples, not the resultant injuries that a woman attacking a dangerous animal is likely to receive.

The rates of male-on-female aggression declined between 1975 and 1985 (Table 8) while female-on-male stayed constant. The 1992 data in Table 8 indicates that overall the level of violence is still declining but women's evaluation of severe violence against their husbands shows a sharp increase. Also, wives evaluation of overall violence by husbands does not show any decrease.

The Straus and Gelles surveys also suggest that women suffer injuries at about 7 times the rate of men but that they used weapons such as baseball bats, boiling water, and knives, among other things, to make up for their physical disadvantage in roughly 80% of the incidents, as compared to men using a weapon in ~25% of the assaults.

In the 1985 National Family Violence Survey:

• 3.0% of women who were assaulted reported they needed to see a doctor.

• 0.4% of men who were assaulted reported they needed to see a doctor.

Based on these percentages, female respondents were about 7.5 times more likely to 'report' they needed to see a doctor as a result of being assaulted by their husbands. For the same level of injury, e.g., bruise, cut, welt, scratch, burns, etc., the data suggest that police and hospital statistics under report domestic violence injuries to men by a factor of about 8. To put it another way, when comparing injuries between men and women, figures for men should be multiplied by about 7.5 to achieve a reasonable comparison simply because men don't usually report the injury.

Many women freely admitted during the surveys that their use of weapons was not in self-defense.

Table 8 obviously lumps mutual violence between a man and a woman into the individual categories. Mutual combat occurs in about 50% of domestic violence incidents (McLeod, 1984; Moffitt and others, 2001).

Causes of domestic violence

Top

Drawing on these studies, Gelles (1979, p. 141) suggested that: "families living in large urban areas, minority racial groups, individuals with no religious affiliation, people with some high school education, families with low incomes, blue-collar workers, people under 30, and families where the husband was unemployed had the highest rate of marital violence."

Nisonoff and Bitman found that alcohol was a contributing factor in 26% of the violent incidents in their study. Drug abuse, whether it be prescription abuse or illicit drugs, is another contributing factor. Some studies have placed the involvement of alcohol as high as 75%, with drug abuse estimated to be involved in 10% to 15% of the incidents.

Other factors in dangerous violence by males and females are physical or mental illness. (Dutton, 1995, p. 140-155, Gelles, 1997, p. 80) find that borderline personality disorder (BPD), which is marked by a proclivity for intense relationships, fear of abandonment, and proneness to rage, to be strongly associated with male battering of women. While we know of no studies making such an association with violent women, 75% of the diagnosed cases of BPD are females.

A history of family abuse is found associated with domestic violence in about one third of the cases.

However, study after study shows that the highest correlation of any factor associated with domestic violence is the absence of the biological father during the child's formative years. In examining the influences of race, ethnicity, gender, and place, Lauritsen and White (2001, p. 53) state that: "...the proportion of households with children that are female-headed was the strongest and most consistent community predictor of risk for all forms of violence."


 

National violence against women survey

Top

The National Institute of Justice and the Centers for Disease Control and Prevention undertook a large, 8,000 men and 8,000 women, survey termed the National Violence Against Women Survey (NVAW). The original purpose of the NVAW was to obtain data on assault victimization of women. Only women were to be interviewed. That obvious bias was fortunately eliminated when an equal sample of men was added after the survey began. That initial bias still strongly influenced the survey as Tjaden's research did not ask women about their assaultive behavior against their intimate partners.

It remains an open question about how likely it is that a given male would participate in a telephone survey titled "Violence Against Women." Nowhere in the reports does it indicate how many men and how many women were actually contacted to obtain a sample size of 8,000 of each sex. One suspects they may have had to contact well over 100,000 men in order to get 8,000 male participants. Female rate of participation would likely be much higher, with probably less than 50,000 contacts required to meet the 8,000 participants goal.

Currently only about 20% of people of either sex contacted by telephone agree to participate in any type of survey.

Such disparities obviously bias the results of such surveys and bring sharply into question how representative the results are of the general population.

The survey of women began in November, 1995, and the survey of men began in February, 1996. Both surveys were completed in June, 1996. An extensive summary by Patricia Tjaden and Nancy Thoennes is available, and both these women are with the Denver-based Center for Policy Research. In contrast with other studies, NVAW found that only 1.9% of women said they were physically assaulted at any level in the previous 12 months. Such assaults include acts of both minor and severe violence. That is a level of assault roughly one-sixth (1/6 th ) of the findings of other independent investigators, as noted above. Straus (1999) has examined the reasons for such a low-level compared to the many other independent surveys, and why NVAW found a three-to-one ratio of assaults by men compared to women. The Associated Press on March 2, 1998, also reported on an earlier talk given by Straus claiming the NVAW survey was simply wrong.

In their exhibit 12, Tjaden and Thoennes show that two-thirds (66.6%) of the injuries reported as part of a physical assault were a "scratch, bruise, welt." Most men incur worse in the daily course of events and probably would not report such an "injury" as an assault. That is particularly true under today's laws where the man would almost certainly be arrested when he called the police and, hence, the assault on him would be counted in crime statistics as male-on-female. Those facts lead to a much higher reporting ratio of assault for females versus males. As noted above, injuries for men are probably under reported by a factor of about 8 compared to women. We also note that if two-thirds of such minor injuries were subtracted from the NVAW statistics the level of domestic violence measured by the survey would differ even more radically, on the low side, from other investigations.

In a 1999 chapter , Straus makes the point that:

"Men are also less likely to call the police, even when there is injury, because, like women, they feel shame about disclosing family violence. But for many men, the shame is compounded by the shame of not being able to keep their wives under control. Among this group, a 'real man' would be able to keep her under control. Moreover, the police tend to share these same traditional gender role expectations. This adds to the legal and regulatory presumption that the offender is a man. As a result, the police are reluctant to arrest women for domestic assault. Women know this. That is, they know they are likely to be able to get away with it. As in the case of other crimes, the probability of a woman assaulting her partner is strongly influenced by what she thinks she can get away with. [quote taken from preprint]"

For these reasons, we agree with Straus (1999) and believe the NVAW survey is fatally flawed. We suggest the reader rely on the more consistent and repeatable results of Straus, Gelles, and Steinmetz, and subsequent surveys.

We ask you to note that NVAW takes a very liberal definition of rape between intimate partners. What is 'rape' in a marriage? Is it sexual congress forced on an uncooperative wife or a biological imperative necessary for the survival of the species? The marriage contract has always been understood as a sex contract. Black's Law Dictionary defines rape as: "The act of sexual intercourse committed by a man with a woman not his wife and without her consent." How has that changed? The question used in the NVAW survey: "Has your partner ever forced you to have sex?" also does not imply rape if the couple are into bondage and domination. Perhaps Tjaden and Thoennes should have read the Joy of Sex and taken a course in biology before beginning their report, or the survey. Ironically, evidence suggests that rape and violence are more common in the type of matriarchal society, e.g., our inner city ghettos, that feminists appear to be pushing for.

Further, Tjaden and Thoennes would have us believe that stalking is much more serious than previously believed because they claim women are more fearful today. Napoleon's army had a saying: "To lie like a bulletin." That might well apply to studies done by feminists under the Clinton administration.

For comparison, take another Clinton program that is, hopefully, less emotional than domestic violence. Everyone has heard from our present government that global warming is occurring and the results will be catastrophic if we don't take action now. Take a moment and look at the actual data. Now, do you still believe in global warming? NOAA in Boulder, Colorado, does because their funding depends on it. Similarly, the funding of Drs. Tjaden and Thoennes, and the Center for Policy Research, depend on the politically-correct finding that women are much more commonly the victim of domestic violence. The only surprise in their results is the low annual prevalence rate.

In his response to NVAW findings, Straus (1999) notes that: "All types of interpersonal violence have been decreasing in the Western world for centuries, and this includes partner assaults..." Rather than simply blaming men wouldn't it be more productive to examine the contributing factors that are causing violence to decrease? We think it extremely unlikely that the current draconian laws passed as a result of the feminist political agenda are a positive factor in this overall decrease.


 

Domestic violence in the military

Top

Another misconception is that military service leads to an increase in wife battering. In her book Battered Wives , author Del Martin claims that:

"The military is a school for violence... He is taught to idealize aggression and rugged masculinity... Is it far fetched to think that there is a connection between military training or combat experience and wife-beating? I don't think so."

That belief is widespread in the feminist literature.

Phillip Cook (p. 5) has examined this hypothesis using data he obtained from the U.S. Army based on 55,000 randomly-selected married soldiers from forty-seven military installations. Of the total sample, 8,500 males and 1,246 females had experienced some level of domestic violence at some point in their relationship, or about 18% of the soldiers sampled. Whether the soldier's spouse was military or civilian was not reported. Note that the rate per year is not reported so that a direct comparison with NIMH/NFVS studies by Straus and others is not possible here but Cook (p. 5) states that it is not significantly different from the civilian population.

Note that it is widely accepted that overall family violence usually decreases as couples age and are together longer. The couples in the Army survey are typically young and recently married.

The military report concludes that: "The most common pattern is for physical aggression to be reported for both partners." How the violence is divided is reproduced in Table 9.


 
    Table 9: Perpetrator of violence based on U.S. Army data (Cook, 1995).

 

Spouse only

Self only

Both violent

Total

Male soldiers

23%

13-14%

62-64%

8,500

Female soldiers

17-23%

23%

60-64%

1,246

Total sample size was 55,000 married soldiers of whom 8,500 males

and 1,246 females experienced some level of domestic violence.

Top


 

Cook (p. 5) notes that there is a remarkable difference in the amount of violence between installations. In the first regional area surveyed that comprised 18 bases approximately 18% of the couples had experienced some level of domestic violence. The highest ratio in the survey was 48%, very close to the NCADV estimate with the notable difference that over 60% of the violence was mutual combat rather than the male attacking the female. In fact, Table 9 suggests that sole attacks by females on their male partners were significantly (~5% overall) higher than the reverse, with female soldiers reporting ~10% more sole violence against their spouse than the reverse.

The data Cook examined strongly suggest that high-stress duty has more to do with domestic violence than the fact of military service per se. The article by Connie Smalls suggests the Army is actually ahead of society in general in dealing with domestic violence, primarily because they recognize both men and women are responsible for family violence. She also notes that "...in most years since FY94 active duty male spouse abuse victims outnumber female victims two to one."


 

Domestic violence as a function of age

Top

There is considerable evidence that the rate of domestic violence decreases as couples age. Clearly young men and women are more physically aggressive. However, there is also evidence that the percentage and severity of violence perpetrated by women increases with age in couples that do fight.

A 1988 study by Pillemer and Finkelhor found that in elderly couples wives were more than twice as likely to assault their husband than vice versa. Similar results from Stets and Straus (1989) are shown in Table 10.

The effect is particularly dramatic for cohabitating women going through the change of life in their late forties. Two-thirds of the domestic violence in cohabitating couples aged 45+ was found to be solely female against male in the Stets and Straus (1989) survey (Table 10). Except for cohabiting couples in the 25-44 age group, female only violence is consistently higher than male only violence. That is particularly true for young couples in the 18-24 age group. Mutual combat is most common in younger couples, accounting for about 60% of the reported violence in the 18-24 age group and steadily diminishing to less than 40% for the 45+ age group (Table 10).


 
    Table 10: Perpetrator of violence as a function of age and marital status (Stets and Straus, 1989)

Age

Marital Status

Female only violence

Male only violence

Both

18-24

Cohabiting

29.0%

9.7%

61.3%

Married

29.1%

13.6%

56.4%

25-34

Cohabiting

18.2%

21.2%

60.6%

Married

26.9%

20.7%

52.4%

35-44

Cohabiting

33.3%

40.0%

26.7%

Married

26.9%

28.0%

45.1%

45+

Cohabiting

66.7%

33.3%

0.0%

Married

34.2%

29.4%

36.4%


 

Extreme female violence and age

Top

In compiling vignettes of violence by women both on this site, and on the parent Equal Justice Foundation site, it appeared that the age of violent women centered around 40. One of the EJF members kindly volunteered to go through all these gruesome stories and sort out the woman's age at time of offense, occupation, primary and secondary crimes, the relationship of the woman to her victim(s), weapon(s) used, and state of residence. As of May 1, 2004, from a total of 232 vignettes, the ages of 183 of the women were available. Crimes included women murdering, poisoning, stabbing, kidnapping, cannibalism, burning their partners and children alive, or attempting or conspiring to murder one or more of their male intimate partners, assault and battery, and false allegations, among other nurturing and caring acts.

Table 11 shows the age distribution of the 183 cases where age at the time of the crime is known, grouped in five year increments. The cases used are a crudely random sample of extreme violence by women from 43 states. This limited sample clearly shows that the proclivity of women for extreme violence increases sharply after age 24 and peaks between age 35 and 44, after which extreme female violence sharply declines. Langran and Dawson (1995) found that in the spouse murders they investigated the average age of husbands who had killed their wife was 41; and 37 years for wives who had killed their husbands, and our results are consistent with their work. While the sample size is relatively small the data suggest that what is termed the "lethality index" increases with age. While physical aggression is more frequent in younger couples, it appears that the danger of extreme, or lethal violence in those couples who continue to fight peaks in the 35 to 44 age range, and decreases sharply after age 44. In a March 2011 report Crime and Justice in Colorado 2008-2010 the authors note (p. 46) that the largest proportion of females convicted of crimes (24% of total) are in the age range of 35 to 44, providing further evidence for an age peak in female violence during the onset of perimenopause.


 
    Table 11: Age of women using extreme violence against male intimate partners.

 

Top


 

The women committing these acts were nearly all wives and girlfriends, ex-wives and ex-girlfriends, and, rarely, a sister, daughter, roommate, or prostitute. There is also some bias in the sampling towards Colorado but that isn't considered significant with regard to age.

Murder leads the roster in Table 11 with 83 cases, and 39 cases of attempted murder, solicitation and conspiracy to commit murder as such crimes by women are more likely to be reported and included in our sample. Assault and battery, severe enough to receive public notice (usually involving hospitalization of the male victim), accounted for 71 cases, and false allegations, stalking, and vandalism accounted for the remainder, and these were frequent secondary crimes. No age was usually known for the lesser crimes and such crimes are seriously underreported in our sample.

Most women's estrogen levels begin to decline between age 35 and 44 as they enter perimenopause. The joke "I'm out of estrogen, and I've got a gun!" seems to have some basis in fact. That correlation certainly deserves further research. With modern medicine, extreme, and lesser female violence might well be reduced. Also, reducing the problems of perimenopause has the potential to reduce the many divorces that occur at this stage of women's lives.

If estrogen levels are a significant factor in the age relationship apparent in female violence plotted in Table 11, then the data also contain an uncorrected bias with regard to hysterectomies (surgical menopause). It is unknown how many of the women who committed these crimes had a hysterectomy before the act. The probable bias would be to shift the apparent peak in Table 11 to a younger age, i.e., the actual peak for violent women who had not had a hysterectomy would likely be in their early forties.

In many cases these women had killed, or attempted to kill a previous lover or husband. In those cases the woman's age at the time of her latest crime is tabulated and may bias the results up to roughly 10% toward a peak at a younger age than shown in Table 11. So the known biases shift the peak in Table 11 to a younger age. It seems clear though that for physically aggressive couples past the age of 35 the potential lethality of their disputes is greater than in younger couples. Such factors as personality disorders, substance abuse, head injuries, etc. would further increase the potential for lethal violence in abusive middle-age couples.

Violence and senility

Top

Another fact that is commonly recognized in nursing homes is that many women become violent as they become senile. This effect is probably as common in men but women live considerably longer than men today, so the effect is most noticeable with women. Stories circulate of couples who have been together decades and then one day she starts whacking him with her cane. Steinmetz (1993, p. 229) reviews cases of abuse beginning when a spouse develops Alzheimer's disease or other personality-altering illnesses such as a stroke. Anetzberger (1999) also touches on this issue but published research is scanty.

Clearly any illness or disease of age that causes an alteration of a partner's personality, and there are many such, may lead to abusive or violent behavior.

These are medical problems and should be treated as such. Intervention by law enforcement and the courts under current laws tends to make a bad situation worse.


 

Actual domestic violence cases in Colorado

Top

According to the Colorado Legislative Impact Staff, for fiscal year 1997-1998 there were 12,166 domestic violence cases filed in state courts as a class 6 felony or below (domestic violence is currently a misdemeanor in Colorado). Of these cases, 4,331 were convicted, pled guilty, or no contest. Domestic violence cases for 1998 and later years are provided in the Demographics chapter.

Of those charged with domestic violence in 1997-1998, only 238 were known to have three or more prior domestic violence convictions. Repeat offenders thus make up only 2% of the cases filed and less than 5% of the convictions.

Unlike other criminal charges, prosecutors are prohibited by law (C.R.S. § 18-6-801 (3)) from dismissing or dropping a charge of domestic violence, or plea bargaining it to anything that does not include DV.

Only very rarely does a court dismiss a domestic violence charge. When they do it is usually the day of the trial in a kind of game of chicken the prosecutor and court play with defendants and defense attorneys.

Thus, the fact that only one third of the domestic violence cases resulted in a conviction is very good evidence that the other two thirds of the charges are false or unjustified.

As shown in Table 38, the Colorado population for 1998 was 3,968,967 people. Thus, 0.3% of the population were charged with domestic violence and 0.1% of the population convicted in the 1998 fiscal year, usually because they pled guilty or no contest. As shown in Table 82, the estimate for 1998 continues through later years.

Impact on jury system

Top

The negative impact of domestic violence cases on the jury system is large because men are forced to take their cases to a jury trial in order to clear their names. Incredibly, as shown in Table 71 and Table 72, domestic violence cases, though a rare crime, are now the largest burden on Colorado courts.

The figures shown here suggest most of 930 jury trials in FY 2009 in Colorado were convened because of false or unjustified accusations of domestic violence.

To the best of our knowledge, only rarely do juries convict in domestic violence cases, and often the woman who made the original charge doesn't want a conviction. Juries in such cases are thus left with the impression that their time was being wasted by the trial.

Every year it becomes more difficult to assemble juries in Colorado as citizens become ever more disgusted with the courts.

Top


 

| EJF Home | Join the EJF | Comments? | Get EJF newsletter | Newsletters |

| DV Home | Abstract | Contents | Authors and Site Map | Tables | Index | Bibliography |

 

| Chapter 4 — Domestic Violence Statistics |

| Next — Methodological Issues: Archival Data vs. Scientific Research by Tom James, Esq. |

| Back — Reading between the numbers |


 

This site is supported and maintained by the Equal Justice Foundation.

Last modified 12/20/16