The Causes of Wife Abuse

views updated

chapter 3

Although researchers have studied wife abuse for about thirty years, scholars from different intellectual traditions often disagree on its origins and the actions required to prevent and address the problem. Sociologists and anthropologists interpret data differently from economists and political scientists. Psychologists and other therapists perceive different facets of the problem, as do social service and shelter workers. Abused women have their own perspectives on the problem.


In the past, domestic violence was viewed as a phenomenon exclusively affecting the lower classes. But when researchers began investigating the causes of family violence in the 1970s, they noticed that although lowerclass women at first appeared to make up the majority of victims, domestic violence, in reality, spanned all social and economic groups.

Middle- and upper-class women were also abused, the researchers found, but they often did not turn to hospital emergency rooms and shelters for help. Instead, they utilized private facilities and remained largely unknown, unreported, and uncounted by the public agencies that attempt to measure the rates of domestic violence and aid victims.

While women of any social class may be victims of abuse, general population studies find that women with lower incomes and less education, as well as minority women, are more likely to be the primary victims of domestic violence. Still, researchers note, classification is not exclusive. Just about anyone, rich or poor, male or female, may be a victim of domestic violence.

The Bureau of Justice Statistics defines an intimate partner as a spouse, former spouse, or a current or former boyfriend or girlfriend, either of the same sex or the opposite sex. (See Table 3.1.) The National Crime Victimization Surveys found that in 2001 an estimated 588,490 violent


Definitions of an intimate partner
Intimate partner relationships involve current spouses, former spouses, current boy/girlfriends, or former boy/girlfriends. Individuals involved in an intimate partner relationship may be of the same gender. The FBI does not report former boy/girlfriends in categories separate from current boy/girlfriends. Rather, they are included in the boy/girlfriend category during the data collection process.
The FBI, through the Supplementary Homicide Reports (SHR), and BJS, using the NCVS gather information about the victim's and offender's relationship, using different relationship categories. In this report responses to the victim-offender question from both data sets are collapsed into four relationship groups: intimate, friend/acquaintance, other family, and stranger. These groups are created from the following original response categories:
NCVS categoriesSHR categories
source: "Definitions of Intimate Partner," in Intimate Partner Violence, U.S. Department of Justice, Bureau of Justice Statistics, May 2000, (accessed September 23, 2004)
Ex-spouseCommon-law husband or wife
Homosexual relationship
Someone at work/customerEmployer
Other non-relativeOther known
Other familyParent or step parentMother/father
Own child or stepchildSon/daughter
Other relativeIn-law
Other family
Known by sight only

crimes—rape, sexual assault, aggravated assault (assault with a weapon), and simple assault victimizations (assault without a weapon and resulting in minor injuries)—were committed against women by their intimate partners. (See Table 3.2.) About 85.1% of all intimate partner violent crimes were committed against women.


Violence against women by intimate partners, 2001
Intimate partner violence
NumberRate per 1,000 personsNumberRate per 1,000 femalesNumberRate per 1,000 males
— Based on 10 or fewer sample cases.
source: Callie Marie Rennison, "Table 1. Violence by Intimate Partners, by Type of Crime and Gender of Victims, 2001," in "Intimate Partner Violence, 1993–2001," Bureau of Justice Statistics Crime Data Brief, February 2003, NCJ 197838, (accessed September 22, 2004)
Overall violent crime691,7103.0588,4905.0103,2200.9
Rape/sexual assault41,7400.241,7400.4
Aggravated assault117,4800.581,1400.736,3500.3
Simple assault471,8602.1421,5503.650,3100.5


Like victims of domestic abuse, batterers come from all socioeconomic groups and all ethnic backgrounds. They may be male or female, young or old. They share a common characteristic—they all have personal relationships with their victims.

In the Bureau of Justice Statistics report Criminal Victimization (Washington, DC, September 2004), statistician Shannan M. Catalano analyzed general crime trends and confirmed that most female violent crime victims in 2003 knew their offenders, while most men were victimized by strangers. Rape and sexual assault victims were the most likely victims to know their assailants. Of the 5.4 million violent crimes that took place in 2003, Catalano found that intimates were offenders in 19% of the violent assaults on females; intimates were involved in only 3% of violent assaults on males. (See Table 2.5 in Chapter 2.)

Marital status was a factor in much of the violence. Never married, divorced, and separated men and women experienced higher rates of victimization than persons who were married or widowed. (See Table 3.3.) In addition, rates of violent victimization by an intimate partner toward women increase as household incomes go down, according to Callie Marie Rennison and Sarah Welchans in "Intimate Partner Violence" (U.S. Department of Justice, Bureau of Justice Statistics, May 2000, NCJ 178247).

The Effects of Poverty

Murray Straus, a highly regarded researcher and codirector of the Family Research Laboratory at the University of New Hampshire, found that serious physical acts of wife abuse are more likely to occur in poorer homes. His research shows that for lower levels of violence, such as shoving or slapping, the differences in socioeconomic status are small. For more serious types of violence, the rates increase dramatically as the socioeconomic status drops.

University of Massachusetts researchers Gerald T. Hotaling and David B. Sugarman found that in eight of eleven studies of socioeconomic status, low socioeconomic status was consistently related to wife assault ("A Risk Marker Analysis of Assaulted Wives," Journal of Family Violence, vol. 5, 1990). Hotaling and Sugarman proposed two interpretations of this finding. First, men of lower socioeconomic status are exposed to greater stress and possess fewer resources to cope with it, such as economic security or education. Second, the relationship between lower socioeconomic status and wife abuse is a response to a subculture of violence that makes these individuals more likely to hold values permitting the abuse of women.

The 1985 National Family Violence Survey, based on 6,002 households, provided researchers with the primary data to test their observations against a database large enough to produce statistically significant, valid findings. In the survey, families living at or below the poverty level had a rate of marital violence 500% greater than more affluent families.

More recent research funded by the National Institutes of Health offered additional support for the relationship between socioeconomic status and abuse. Deborah Pearlman et al. presented the findings of an analysis of policereported domestic violence in relation to variables including socioeconomic conditions, age, race, and ethnicity (Neighborhood Environment, Racial Position and Domestic Violence Risk: Contextual Analysis, Academy for Health Service Research and Health Policy, Annual Meeting, June 24, 2002). Researchers found a complex but strong relationship between poverty and domestic violence. They speculated that one explanation for the increased risk of domestic violence in poorer neighborhoods might be differences in law enforcement availability and practices—economically deprived communities might have less police notification, attention, and documentation.


Rates of violent crime and personal theft, by household income, marital status, region, and location of residence of victims, 2003
Victimizations per 1,000 persons age 12 or older
Violent crimes
Characteristic of victimPopulationAllRape/sexual assaultRobberyTotalAggravatedSimplePersonal theft
Note: The National Crime Victimization Survey includes as violent crime rape, sexual assault, robbery, and assault. Because the NCVS interviews persons about their victimizations, murder and manslaughter cannot be included.
*Based on 10 or fewer sample cases.
source: Shannan M. Catalano, "Table 7. Rates of Violent Crime and Personal Theft, by Household Income, Marital Status, Region, and Location of Residence of Victims, 2003," in Criminal Victimization, 2003, U.S. Department of Justice, Bureau of Justice Statistics National Crime Victimization Survey, September 2004, (accessed September 19, 2004)
Household income
Less than $7,5008,335,12049.91.6*9.039.310.828.51.2*
$75,000 or more47,855,86017.50.5*1.715.42.712.61.0
Marital status
Never married76,429,29041.


The struggle for power in a relationship appears to play a significant role in battering. Some researchers suggest that the need to exert control over one's partner begins long before marriage.

In a study by Diane R. Follingstad et al. titled "Risk Factors and Correlates of Dating Violence: The Relevance of Examining Frequency and Severity Levels in a College Sample" (Violence and Victims, vol. 14, no. 4, Winter 1999), 290 male and 327 female college students were questioned about violence in their dating relationships. The authors found that students who used physical force in dating relationships often did so in order to exercise greater control over their dating partners. Compared to students who did not use violence in their dating relationships, respondents who reported using force were more likely to express anger, experience higher levels of jealousy, have poorer communication skills, report more daily stressors, display more irrational behavior and beliefs, and have more difficulty controlling their anger. These same respondents also reported more problems with alcohol, more verbal aggressiveness, and more efforts to control their dating partners than the students who never used force. The researchers concluded that intervention in the area of dating violence should focus specifically on an individual's need to control his or her dating partner and the motivations for the need to control.

Dissatisfaction with the amount of power a dating partner felt in a relationship was associated with the use of violence for both men and women, according to "Power and Dating Violence Perpetration by Men and Women," a study of 352 male and 296 female undergraduate college students by Shelby A. Kaura and Craig M. Allen. They found, however, that witnessing parental violence was a stronger predictor (Journal of Interpersonal Violence, vol. 19, May 2004). The study also found that women reported they were the perpetrators of significantly more dating violence than did men. The authors concluded that "dissatisfaction with power in relationships is important for both genders."

The Balance of Power

The desire to dominate one's partner may be manifested using methods other than violence, such as attempts at financial, social, and decision-making control. Some researchers theorize that men of lower socioeconomic status are more likely to batter because they do it to assert the power that they lack economically. Violence becomes the tactic that compensates for the control, power, independence, and self-sufficiency these men lack in other areas.

In "Marital Power, Conflict, and Violence in a Nationally Representative Sample of American Couples," Diane Coleman and Murray Straus analyzed data from the 1975 National Family Violence Survey to determine the characteristics of families most prone to violence (Violence and Victims, vol. 1, no. 2, 1986). They asked respondents to indicate "who has the financial say" when it came to buying a car, having children, choosing a house or apartment, the type of job either partner should take, whether a partner should work, and how much money to spend each week for food. The response options were divided into "husband only," "husband more than wife," "husband and wife exactly the same," "wife more than husband," and "wife only."

Coleman and Straus arranged the responses to create four categories: male-dominant, female-dominant, equal, and divided power. The difference between the equal and divided-power types was that in the former the wife and husband made most decisions jointly, while in the latter they divided responsibility for decisions, with each having the final say for different decisions. The highest likelihood of violence occurred in relationships where one partner was dominant. Equal and divided-power relationships, in contrast, had the lowest likelihood of violence. Coleman and Straus also found that equal relationships could tolerate more conflict before violence erupted than other power relationships. Inequality, they posited, inevitably leads to attempts to even out the relationship, which in turn causes conflicts and perhaps violence. However, the researchers observed, male- or female-dominant relationships in which the partners accepted their status in the relationship usually experienced lower levels of conflict and violence.

Learned Gender Roles

Pointing to history, some researchers see wife abuse as a natural consequence of women's second-class status in society. Among the first to express this viewpoint were Emerson Dobash and Russell Dobash in Violence against Wives (New York: Free Press, 1979). Dobash and Dobash argued that men who assaulted their wives were actually living up to roles and qualities expected and cherished in Western society—aggressiveness, male dominance, and female subordination—and that they used physical force as a means to enforce these roles. Many sociologists and anthropologists believe that men are socialized to exert power and control over women. Some men may use both physical and emotional abuse to attain the position of dominance in the spousal relationship.

Other researchers agree. Violence often grows out of inequality within an intimate relationship and reinforces male dominance and female subordination, according to Kersti Yllöo, in "Through a Feminist Lens: Gender, Power, and Violence," from the book Current Controversies on Family Violence (Newbury Park, CA: Sage, 1994). For Yllöo, violence against women in all of its forms, including sexual harassment and date rape, is a tactic of male control, and domestic violence is not just a conflict of interests, it is domination by men. In their study of thirty-three male batterers titled "Gendering Violence: Masculinity and Power in Men's Accounts of Domestic Violence," Kristin L. Anderson and Debra Umberson wrote: "violence is… an effective means by which batterers reconstruct men as masculine and women as feminine" (Gender & Society, vol. 15, June 2001).

Murray Straus compared data on wife battering with indices to measure gender equality, income, and social disorganization variables as part of his research work in "State-to-State Differences in Social Inequality and Social Bonds in Relation to Assaults on Wives in the United States" (Journal of Comparative Family Studies, vol. 25, no. 1, Spring 1994). Straus evaluated gender equality by measuring twenty-four indicators to determine the extent to which women have parity with men in economic, political, and legal arenas. Income inequality was assessed using census data on family income. Social disorganization measures the level of societal instability, such as geographic mobility, divorce, lack of religious affiliation, female- or male-headed households, and the ratio of tourists to residents in each state.

In all fifty states studied, Straus found that gender equality was the variable most closely related to the rate of wife assault—states where the status of women was higher were less likely to report high rates of wife abuse. Social disorganization was also related to abuse—the higher degree of social disorganization, the greater the probability that the state would have a high rate of wife assault. Straus found that economic inequality did not appear to be related to wife abuse rates.

Attitudes toward Violence

Some researchers believe attitudes about violence are shaped early in life, long before the first punch is thrown in a relationship. In "The Attitudes Towards Violence Scale: A Measure for Adolescents" (Journal of Interpersonal Violence, vol. 14, no. 11, November 1999), Susan B. Funk et al. asked junior high and high school students attending an inner-city public school in a Midwestern city about their attitudes toward violence. Some students identified themselves as victims of violence and others completed the survey before and after participating in a violence awareness program.

Using the responses of 638 students who took the survey prior to the violence awareness program, the researchers examined the correlation of violence with gender, grade-level, and ethnicity. They found that males endorsed more pro-violence attitudes independent of age, grade-level, and ethnicity, as did those students who identified themselves as victims of violence. African-American teenagers endorsed "reactive violence," or violence used in response to actual or perceived threats, at higher levels than other groups. Endorsement of reactive violence was linked to having violent behaviors in one's repertoire, willingness to act in a violent manner, and supporting the actual choice of a violent response. Hispanic Americans endorsed "culture of violence" measures, reflecting a pervasive identification with violence as a valued activity, at slightly higher levels than the teenagers as a group. "Culture of violence" measures included the conviction that the world is a dangerous place where the best way to ensure survival is to be vigilant and prepared to take the offensive. European Americans scored lower on measures of "reactive violence" as well as "total pro-violence attitudes."

The study found that gender, ethnicity, and selfidentification as a victim of violence were all related to proviolence attitudes. Males, regardless of cultural background, were more likely than females to endorse pro-violence attitudes. The researchers concluded that a combination of biological, environmental, and social influences were responsible for these findings.

Does a Patriarchal Society Breed Violence?

Donald Dutton, a professor of psychology and director of a treatment program for batterers, questioned the role of male domination in wife battering and offered alternative explanations for violence in "Patriarchy and Wife Assault: The Ecological Fallacy" (Violence and Victims, vol. 9, no. 2, 1994). According to the patriarchal model, societies that place a high value on male dominance should have high rates of abuse. But Dutton and other investigators cite studies that contradict this premise. For example, Coleman and Straus found that in marriages where spouses agreed that the husband should be dominant, violence levels were low.

Other research discounts the weight of the patriarchal theory of abuse. David B. Sugarman and Susan Frankel, in "Patriarchal Ideology and Wife-Assault: A Meta-Analytical Review" (Journal of Family Violence, vol. 11, no. 1, 1996), examined studies for evidence of a relationship between patriarchy and violence. They measured whether violent husbands had a higher acceptance of violence than nonviolent men and whether they believed that women should exhibit traditional gender roles of obedience, loyalty, and deference. The researchers also measured whether assaultive men were more likely to possess a traditional "gender schema," an internal perception of an individual's own levels of masculinity, femininity, or androgyny. They also considered whether assaulted wives held more traditional gender attitudes than wives who were not battered and whether battered wives held more traditional feminine gender schemas.

Overall, their analysis found support for only two of the five hypotheses. Predictably, assaultive husbands found marital violence more acceptable than nonviolent husbands, and battered wives were more likely to be classified as having "traditional" feminine gender schemas than wives who were not assaulted. Sugarman and Frankel concluded that their findings offered only partial support for the patriarchy theory.

Mark Totten, however, found another link between patriarchy and violence. He concluded in a study of thirty male adolescents, primarily gang members, that under-privileged males in society use violence toward women in response to their lack of access to the traditional benefits of patriarchy ("Girlfriend Abuse as a Form of Masculinity Construction among Violent, Marginal Male Youth," Men and Masculinities, vol. 6, July 2003). Totten posited that the ideals of patriarchy—and the inability of these disenfranchised boys to wield any patriarchal power outside of their gangs or family groups—led them to be violent toward their girlfriends as one way to define their masculinity. "Violence," he wrote, "was one of the few resources over which they had control." On the other hand, he wrote, "Men with more resources can commit different, less visible forms of abuse."


Most sociologists and psychologists agree that lower levels of aggression, such as slapping and shoving, can escalate over time into more severe forms of abuse, such as battering and weapon use. However, while most relationships characterized by severe violence begin with milder forms of abuse, many partners limit their aggressive physical behavior to pushing and slapping. K. Daniel O'Leary, in "Through a Psychological Lens: Personality Traits, Personality Disorders, and Levels of Violence" (Current Controversies on Family Violence [Newbury Park, CA: Sage, 1994]) argued that the different levels of aggression—verbal, mild physical aggression, and severe physical violence—are three distinct but related behaviors. Although they exist along a continuum, he contends that men who engage in the milder forms of aggression are not motivated by the same impulses as men who commit severe abuse.

As further evidence of the differences in levels of abuse, O'Leary observed that in marriages with low levels of physical violence, the abuse is often mutual, and that women do not describe their use of force as self-defense. In severely violent relationships, however, women often claim that they use violence in self-defense. O'Leary maintained that this difference is important because while marital therapy may be appropriate and effective treatment for low-level violence, it is neither appropriate nor effective for relationships characterized by severe abuse.

O'Leary found that mildly abusive men scored high on personality tests for impulsiveness, a readiness to defend oneself, aggression, suspicion of others, and a tendency to take offense easily. Men in treatment programs for abuse—generally extremely abusive men—usually have been diagnosed with serious psychological disorders, including schizoid/borderline, narcissistic/antisocial, and possessive/dependent/compulsive personality traits. These men were significantly different from men who were in bad marriages but were not abusive. O'Leary contended that these findings are evidence of a strong psychological component to abuse rather than a social system that promotes the domination of women.

Some feminist researchers disagree with the concept of distinguishing between types of abuse and theories that link the causes of abuse to the severity of the violence. They consider all violence against women unacceptable, and they reject the idea that the pathological personality characteristics of the perpetrators (serious mental health diagnoses) explain or excuse all of their violent behaviors. Feminist researchers and academics question psychological interpretations of violence that portray batterers as psychologically different from the rest of society because many believe that any man has the capability to become a batterer simply by virtue of living in a patriarchal society.


Richard Gelles, the chair of Child Welfare and Family Violence and interim dean of the University of Pennsylvania School of Social Work, thinks it is risky to place too much emphasis on a psychological explanation of abuse. He contended that the picture of a mentally deranged, violent abuser focuses attention on only the most extreme cases of abuse, stereotyped as a psychotic offender and an innocent victim. According to Gelles, only about 10% of abusive incidents are caused by mental illness. The rest, he asserted, cannot be explained by a psychological model. Gelles said a more complete understanding of the causes of abuse may be gained from an examination of sociological models.

General Systems Theory

The general systems theory views violence as a system rather than as a result of individual mental disturbance. It describes a system of violence that operates at the individual level, the family level, and at a societal level.

Straus developed eight concepts to illustrate the general systems theory:

  • Violence between family members has many causes and roots, and personality, stress, and conflicts are only some of the causes of domestic violence.
  • More family violence occurs than is reported.
  • Most family violence is either denied or ignored.
  • Stereotyped family-violence imagery is learned in early childhood from other family members.
  • The family-violence stereotypes are continually reaffirmed through ordinary social interactions and the mass media.
  • Violent acts by violent persons may generate positive feedback; that is, these acts may produce desired results.
  • Use of violence, when contrary to family norms, creates additional conflict.
  • Persons who are labeled violent may be encouraged to play out a violent role, either to live up to the expectations of others or to fulfill their own self-concepts of being violent or dangerous.

The Resource Theory

The second theory in sociological models is known as the resource theory. According to this theory, the more resources—social, personal, and economic—a person can command, the more force he or she can potentially call on. The individual who is rich in terms of these resources has less need to use force in an open manner. In contrast, a person with little education, low job prestige and income, or poor interpersonal skills may use violence to compensate for a real or perceived lack of resources and to maintain dominance.

The Exchange/Social Control Theory

The exchange/social control theory argues that violence can be explained by the principle of costs and rewards. The private nature of the family, the reluctance of social institutions to intervene, and the low risk of other interventions reduce the risk of negative consequences from abuse. This theory maintains that cultural sanction and approval of violence increase the potential rewards for violence.

The Subculture of Violence Theory

The fourth theory posits that there is a subculture of violence in which some groups within society hold values that permit, and even encourage, the use of violence. This theory is offered as an explanation of why some segments of society and some cultures are more violent than others. This theory is perhaps the most widely accepted theory of violence.

Feminist Theory

Feminist theories of violence against women emphasize that societal patriarchal structures of gender-based inequalities of power are at the root of the problem. The violence, rather than being an individual psychological problem, is instead an expression of male domination of females. Violence against women, in the feminist view, includes a variety of "control tactics" meant to control women.

Structure of Interpersonal Relationships Theory

A more recent theory argued that several key structural features of relationships are conducive to domestic violence. It was presented by Donald Black in "Making Sociological Sense Out of Trends in Intimate Partner Violence" (Violence against Women, vol. 10, June 2004). Black, like Gelles, argued that many of the insights of other theories need to be integrated into a more comprehensive theory of the impact of the structure of relationships on domestic violence. He argued that key risk factors of domestic violence included 1) social isolation of the couple, 2) separate peer support networks, 3) inequality between partners, 4) lack of relational distance, or a high degree of intimacy within a couple, 5) the centralization of authority—in other words, patriarchal dominance within a family, and 6) exposure to violence and violent networks.


The role of alcohol and drug abuse in family violence is featured in many studies, and it is a factor in physical violence and stalking, according to such researchers as Pam Wilson et al. who examined the issue in their article "Severity of Violence against Women by Intimate Partners and Associated Use of Alcohol and/or Illicit Drugs by the Perpetrator" (Journal of Interpersonal Violence, vol. 15, September 2000.) Although researchers generally don't consider alcohol and drug use to be the cause of violence, they find that it can contribute to, accelerate, or increase aggression. A variety of data sources establish correlation (a complementary or parallel relationship) between substance abuse and violence, but correlation does not establish causation. In theory, and possibly even in practice, substance abuse may promote or provoke domestic violence, but both may also be influenced by other factors, such as environmental, biological, and situational stressors. Based on available research, it remains unclear whether substance abuse is a key factor in most domestic violence incidents.

Analyzing 2002 National Crime Victimization Survey data, authors of the 2003 Bureau of Justice Statistics report Drugs and Crime Facts found that 17% of victims of violent assaults believed their offenders had been using alcohol, 4.6% believed offenders had been using both alcohol and drugs, 5.6% believed offenders had been using drugs only, and 1.5% believed offenders had used either alcohol or drugs. (See Figure 3.1.) Only 27.7% of victims believed the offender had not used any drugs or alcohol, while another 43.3% reported they did not know.

While anecdotal evidence suggests that alcohol and drugs appear to be linked to violence and abuse, in controlled studies the connection is not as clear. For example,


some research finds that heavy binge drinking is more predictive of abuse than daily consumption of alcohol. Other research reveals little evidence that drug use directly causes people to become aggressive or violent, and some investigators believe that the substance abuse–violence link varies across individuals, over time within an individual's life, and even in response to environmental influences, such as epidemics of drug use and changing law enforcement policies.

In "Alcohol and Other Drugs Are Key Causal Agents of Violence" (Current Controversies on Family Violence, [Newbury Park, CA: Sage, 1994]), Jerry Flanzer suggested that while not all alcoholics are violent, alcoholism does cause family violence. He observed that personality characteristics of alcoholics and abusers are remarkably similar, marked by behaviors such as blaming others, jealousy and possessiveness, depression, low self-esteem, and "blacking out" critical incidents. Not all alcoholic families are violent and not all violent families include alcohol abusers, but Flanzer speculated that a careful examination of family histories would reveal overlap between violence and alcohol abuse.

Alcohol acts as a disinhibitor, allowing an individual to act out emotions, including anger, that were previously held in check. It also impairs an individual's understanding of a situation, which may lead a drinker to respond inappropriately with anger. Alcohol is also frequently used to rationalize erratic behavior and violence, allowing an abuser to avoid responsibility for violent behavior by blaming it on the effects of alcohol. Finally, alcoholism distorts the family system by constantly forcing the family to accommodate the short-term demands of the alcoholic to maintain some measure of family stability. This


Alcohol and drug use by victims and their partners in the year prior to the killing or attempted killing of women or the worst violent incident, 1996
Homicide/attempted homicide
source: Phyllis Sharps, Jacquelyn C. Campbell, Doris Campbell, Faye Gary, and Daniel Webster, "Table 1: Alcohol and Drug Use by Victims and Their Partners in the Year Prior to the Killing or Attempted Killing of Women or the Worst Violent Incident," in "Risky Mix: Drinking, Drug Use, and Homicide,"National Institute of Justice Journal, no. 250, November 2003, (accessed September 10, 2004)
Drunk every day35.111.61.2
Problem drinker13.
Drinks per episode
Ever been in alcohol treatment27.713.513.318.157.119.2
Use drugs18.454.213.425.06.74.3
Ever been in drug treatment20.611.33.512.414.321.4

restructuring of family life establishes an atmosphere that tolerates and accommodates violence.

A 2003 study of patterns of alcohol and drug use in murders and attempted murders of women by their partners showed a relationship between substance use and violence (Phyllis Sharps, Jacquelyn C. Campbell, Doris Campbell, Faye Gary, and Daniel Webster, "Risky Mix: Drinking, Drug Use, and Homicide," National Institute of Justice Journal, no. 250, November 2003). Researchers found that in the year before the violent incident, female victims used alcohol and drugs less frequently and consumed smaller amounts than did their male partners. (See Table 3.4.) Researchers also found that during the homicide or attempted homicide, 31.3% of perpetrators consumed alcohol, 12.6% of perpetrators used drugs, and 26.2% used both. Less than one in three perpetrators (29.9%) used neither alcohol nor drugs. (See Table 3.5.) On the other hand, perpetrators who abused their partners without attempting to kill them consumed alcohol 21% of the time, drugs 6.7% of the time, and both drugs and alcohol 5.8% of the time. These perpetrators had not used substances almost two-thirds of the time (65.8%). Researchers concluded that increased substance use results in more serious violence.

An Excuse, Not a Cause

Gelles disagreed with Flanzer and others who have presented substance abuse as a cause of family violence in "Alcohol and Other Drugs Are Not the Cause of Violence" (Current Controversies on Family Violence [Newbury Park, CA: Sage, 1994]). Gelles argued that although substantial evidence has linked alcohol and drug use to violence, there is little scientific evidence that alcohol or other drugs, such as cocaine, have pharmacological properties that produce violent and abusive behavior. Although amphetamines have been proven to generate increased aggression, there is no evidence that such aggression is routinely expressed as family or intimate partner violence. Gelles maintained that although alcoholism may be associated with intimate violence, it is not a primary cause of the violence.

If alcohol had the pharmacological property of inducing violence, it would do so in all cultures, Gelles argued. Cross-cultural studies of alcohol consumption and violence, however, do not support this correlation between alcohol and violence. In some cultures individuals who drink become passive; in others they become aggressive. In U.S. culture, Gelles noted, drinking disinhibits, permitting violent behavior without responsibility. Social expectations about drinking and drinking behavior exacerbate the problem, teaching people that if they want to avoid accountability or responsibility for their violence, alcohol consumption provides one socially acceptable justification.

Experiments using college students as subjects find that when the students thought they were consuming alcohol, they acted more aggressively than if they were told they had been given nonalcoholic drinks. According to Gelles, it is the expectation of the effects of alcohol that influences behavior, not the actual liquor consumed. He also observed that although abusive families may also abuse alcohol, an analysis of the drinking behavior at the time of the abuse finds that alcohol was not used immediately prior to the abuse in a majority of cases.

The "Drunken Bum" Theory of Wife Beating

Glenda Kaufman Kantor and Murray Straus have tested three commonly held beliefs: that alcohol and wife


Substance use during the killing or attempted killing of women or the worst violent incident, 1996
Homicide/attempted homicideAbuse
source: Phyllis Sharps, Jacquelyn C. Campbell, Doris Campbell, Faye Gary, and Daniel Webster, "Table 2: Substance Use During the Killing or Attempted Killing of Women or the Worst Violent Incident," in "Risky Mix: Drinking, Drug Use, and Homicide," National Institute of Justice Journal, no. 250, November 2003, (accessed September 10, 2004)
Substance use

abuse are related, that wife abuse is more common in blue-collar than white-collar families, and that the acceptance of violence contributes to spousal abuse. In "The 'Drunken Bum' Theory of Wife Beating" (Physical Violence in American Families [Piscataway, NJ: Transaction, 1990]), Kantor and Straus found that the combination of blue-collar work status, drinking, and approval of violence were associated with the highest likelihood of wife abuse. Men with these characteristics had a rate of abuse that was 7.8 times greater than the rate of white-collar men who drank little and disapproved of violence.

Kantor and Straus emphasized that in three-quarters of the cases (76%), alcohol was not consumed immediately prior to the instance of wife abuse. In about 14% of the instances, only the male was drinking, in 2% only the female was drinking, and in 8% both the male and female were drinking. The researchers found a definite correlation between the amount of alcohol consumed and violence. Approximately 6.8% of those who abstained from alcohol abused their wives, while almost three times as many binge drinkers (19.2%) used violence. Kantor and Straus, however, underscored the importance of not overlooking the considerable amount of wife abuse perpetrated by nondrinkers and moderate drinkers.

The researchers also found that the major factor determining wife abuse was whether the individual approved of the use of violence against women. Not surprisingly, men who approved of a man hitting his wife were far more likely to have hit their wives than men who disapproved.

The Relationship Is Complex

Glenda Kaufman Kantor warned that considering alcohol as simply a disinhibitor of violence understates the complexity of the problem in her study "Refining the Brushstrokes in Portraits of Alcohol and Wife Assaults" (Alcohol and Interpersonal Violence: Fostering Multidisciplinary Perspectives [Rockville, MD: National Institute on Alcohol Abuse and Alcoholism Research, 1993]). She proposed that the finding that heavy drinking by wives can increase their risk of being abused may be because intoxicated women violate what is considered to be the normal gender role. Furthermore, if women become verbally or physically aggressive under the influence of alcohol, they risk being beaten. Alcoholic women are also more likely to suffer abuse from their partners than nonalcoholic women. However, this finding does not imply that the wife has brought on the abuse by her drinking. Kantor emphasized that violence is initiated by the aggressor, not the victim.

In their study "When Women Are under the Influence: Does Drinking or Drug Use by Women Provoke Beatings by Men?" (Recent Developments in Alcoholism [New York: Plenum, 1997]), Glenda Kaufman Kantor and Nancy Asdigian found little evidence that women's drinking provoked or preceded aggression by husbands. The authors theorized that women drink or use drugs as a means of coping with violent partners. In households where the husband is a substance abuser and abuses his wife, the wife is also more likely to be a substance abuser. The authors proposed that both excessive drinking and spousal abuse have common roots in childhood experiences of physical and/or sexual abuse.

Does Treatment Help?

Timothy O'Farrell and Christopher Murphy, in "Marital Violence before and after Alcoholism Treatment" (Journal of Consulting and Clinical Psychology, vol. 63, no. 2, 1995), examined whether behavioral marital therapy was helpful in reducing violence in abusive relationships. The researchers found the percentage of couples who experienced violent acts decreased from about 65% before treatment to about 25% after treatment. Severe violence dropped from between 30% and 35% before to about 10% after treatment.

Following treatment, recovering alcoholics no longer had elevated violence levels, but alcoholics who relapsed did. Based on women's reports of their partners' violence, 2.5% of nondrinking alcoholics, compared to 12.8% of the nonalcoholic sample, were violent. In contrast, 34.7% of the relapsed alcoholics were violent. O'Farrell and Murphy warned that the data do not permit drawing the conclusion that drinking caused the continued violence since other factors may have influenced behavior. They did conclude, however, that their findings support the premise that recovery from alcoholism can reduce the risk of marital violence.

A more recent study supported these findings. After intensive inpatient treatment of male batterers for alcoholism, both alcohol consumption and levels of violence within families decreased, according to Gregory L. Stuart, et al. in "Reductions in Marital Violence Following Treatment for Alcohol Dependence" (Journal of Interpersonal Violence, vol. 18, October 2003). Not only did the frequency of husband-to-wife physical and psychological abuse decrease, but the frequency of wife-to-husband marital violence also decreased significantly.


Research about intimate partner violence reveals that violence does not stop when women become pregnant. The Centers for Disease Control and Prevention's Division of Reproductive Health gathers data about the health of expectant mothers using its Pregnancy Risk Assessment Monitoring System (PRAMS). An analysis of PRAMS data revealed that between 2.9% to 5.7% of women reported being abused by their husbands or partners in the year before they gave birth. Author Jana L. Jasinski believed this estimate too low, since the PRAMS asks limited questions about domestic violence and asks about abuse rather than about particular behaviors ("Pregnancy and Domestic Violence: A Review of the Literature," Trauma, Violence, & Abuse, vol. 5, January 2004). Still, she argued, pregnancy does not appear to increase the risk of domestic violence, although more research into that question is needed.

Studies estimating higher rates of abuse of pregnant women—as many as 324,000 women per year and rates as high as 20% of pregnant women—have been reported (Julie Gazmararian et al., "Prevalence of Violence against Pregnant Women," Journal of the American Medical Association, vol. 275, no. 24, 1996; and Julie Gazmararian et al., "Violence and Reproductive Health: Current Knowledge and Future Research Directions," Maternal and Child Health Journal, vol. 4, no. 2, 2000). Higher abuse rates were reported later in pregnancy, with 7.4% to 20% of that violence occurring in the third trimester. The lowest rates were reported in a study of women with higher socioeconomic status who were treated in a private clinic. The assailants were mainly intimate or former intimate partners, parents, or other family members. Two studies that also examined violence in the period after birth found that violence was more prevalent after birth than during pregnancy.

Jana L. Jasinski's research, cited above, suggested that violence directed toward pregnant women is usually part of an ongoing pattern of domestic violence. Some factors, however, do seem to increase the risk of violence for pregnant women. Julie Gazmararian et al. found that women with unwanted pregnancies had 4.1 times the risk of experiencing physical violence by a husband or boyfriend during the months prior to delivery than did women with desired pregnancies. Researchers have found higher rates of violence during pregnancy for women who are young, have fewer than twelve years of education, are unmarried, are of low socioeconomic status, have postponed or foregone prenatal care, or have an unintended pregnancy. Other studies have found more reported violence when the partner is unhappy about the pregnancy.


Research demonstrates a relationship between having been a victim of violence and becoming violent in future relationships. In fact, a 1996 report prepared by the American Psychological Association Task Force on Violence and the Family concludes that children's exposure to their father abusing their mother is the single strongest risk factor for passing violence down from one generation to the next. A study of 352 male and 296 female undergraduate college students found that witnessing parental violence was the strongest predictor of perpetrating dating violence (Shelby A. Kaura and Craig M. Allen, "Power and Dating Violence Perpetration by Men and Women," Journal of Interpersonal Violence, vol. 19, May 2004). Straus, however, cautioned against jumping to the conclusion that once violence occurs in a family, it will inevitably or automatically be transmitted to the next generation. Not all men who grow up in violent families end up abusing their spouses, and not all abused children or abused wives will abuse others. Conversely, some violent individuals grow up in nonviolent families.

Richard Gelles and Murray Straus, in Intimate Violence (New York: Simon and Schuster, 1989), reported that the experience of children seeing their parents strike one another teaches three lessons:

  • Those who love you are also those who hit you and those you love are people you can hit.
  • Seeing and experiencing violence in your home establishes the moral rightness of hitting those you love.
  • If other means of getting your way, dealing with stress, or expressing yourself do not work, violence is permissible.

Gerald T. Hotaling and David B. Sugarman, in "An Analysis of Risk Markers in Husband-to-Wife Violence: The Current State of Knowledge" (Violence and Victims, vol. 2, no. 2, 1989), considered fifty-two studies of domestic violence for ninety-seven potential risk markers, defined as attributes or characteristics associated with an increased risk of either the use of husband-to-wife violence or of being victimized by husband-to-wife violence. They found only one consistent risk for the victims in the relevant literature: women who have experienced physical spousal abuse are more likely to have witnessed violence between parents or caregivers during their childhoods. Experiencing violence was a weaker predictor of severe husband-to-wife violence than is witnessing violence.


Donald Dutton et al. in "The Role of Shame and Guilt in the Intergenerational Transmission of Abusiveness" (Violence and Victims, vol. 10, no. 2, Summer 1995), tested their theory that being shamed in childhood leads to an assaultive adulthood. The distinction between shame and guilt is that shame produces disturbances in self-identity, while guilt produces bad feelings and remorse about the condemned behavior but not the self. Using a series of psychological tests with 130 battering men, researchers concluded that shaming experiences in childhood contribute to the formation of a borderline personality disorder, including identity disturbances, temporary psychotic experiences, and the use of defenses such as projecting blame on someone else or the "splitting" of an individual's personality. According to the authors, shaming experiences result in personality disturbances, while parental abuse contributes the model behavior for expressing anger.

Beaten Child, Beaten Wife

Ronald Simons et al. in "Explaining Women's Double Jeopardy: Factors that Mediate the Association between Harsh Treatment as a Child and Violence by a Husband" (Journal of Marriage and the Family, vol. 55, 1993), examined the link between women who received harsh treatment in childhood and later married abusive husbands. Researchers asserted that women who were abused in childhood marry abusive husbands—not because they have learned that violence is permissible, but because they are apt to marry men from similar backgrounds. Children raised in violent environments are often noncompliant, defiant, aggressive, and perform poorly in school. Researchers found that the abusive behavior by adults who were abused in childhood was part of a long-standing pattern of interpersonal difficulties and antisocial behavior.

The researchers did not find a connection between abuse and traditional gender beliefs that men are supposed to be dominant. Nor did they find a connection between the level of control the women felt they had in their lives and the incidence of abuse. On the contrary, the women in abusive marriages were not submissive; they tended to have a history of aggressive, deviant behavior. Simons et al. theorized that rebellious girls are more likely to date and marry equally antisocial, rebellious young men and end up in abusive relationships.


Straus and Gelles have found that selected variables, including employment status, income, and number of children, are often associated with domestic violence in a given family. Families with the lowest incomes, the most children, and the lowest level of employment of the husband tend to be at greater risk for spouse abuse.

In Behind Closed Doors: Violence in the American Family (Garden City, NY: Anchor Books, 1980), Murray Straus, Richard Gelles, and Suzanne Steinmetz constructed a scale to measure overall family stress and applied it to National Family Violence Surveys data. In addition to income, number of children, and employment, they included other stressors: illness or death in the family, arrest or conviction of a family member, relocation, sexual difficulties, and problems with in-laws. They found a strong correlation between the number of stressful events experienced during the past year and the rates of family violence and abuse. The greater the stress, the greater the likelihood of abuse.

Other studies have also found a relationship between domestic abuse and stress. "Frequency and Correlates of Intimate Partner Violence by Type: Physical, Sexual, and Psychological Battering" (American Journal of Public Health, 2000) revealed that stressors, such as a male partner's unemployment and alcohol and/or drug use, were associated with an increased risk for physical, sexual, and emotional abuse.


Jill Suitor, Karl Pillemer, and Murray Straus in "Marital Violence in a Life Course Perspective" (Physical Violence in American Families, [Piscataway, NJ: Transaction, 1990]), found that both marital conflict and verbal aggression consistently decline with age over every ten-year period. Analysis of the National Family Violence Surveys data from 1975 and 1985 revealed that the rate of violence in the eighteen- to twenty-nine-year-old group dropped when its members entered the thirty- to thirty-nine-yearold group. The rate dropped even further when the older group became the forty- to forty-nine-year-old group. The consistent decline applied to both men and women in all age groups between eighteen and sixty-five.

Suitor, Pillemer, and Straus concluded that marital conflict and verbal aggression decrease with age. They considered several different possible explanations for this observation, including greater pressure to conform (perhaps because of a greater stake in society), the greater cost of deviating from accepted patterns—having "more to lose"—and greater expectations.

Subsequent studies confirm that intimate partner violence declines with advancing age. Analyzing National Crime Victimization Survey data, Rennison found married women aged twenty to twenty-four had eight victimizations per one thousand women, compared to just one per one thousand among married women aged fifty or older (Intimate Partner Violence and Age of Victim, 1993–99, [Washington, DC: Bureau of Justice Statistics, 2001]). Callie Rennison and Michael Rand also found lower rates of intimate partner violence in women over age fifty-four in their study "Nonlethal Intimate Partner Violence against Women" (Violence against Women, vol. 9, December 2003). They believed lower rates might be due to several factors, such as homicides of younger women, earlier divorces from abusive partners, or the turning of older perpetrators to other forms of victimization, such as psychological abuse or economic domination.

Spousal Abuse among Older Adults

Spouse abuse is a known form of elder abuse, but there is little known about its precise causes or frequency. The rate of spouse abuse among older adults is estimated to be less than 20% of all elder abuse reported by the National Elder Abuse Incidence Study (Washington, DC: National Center on Elder Abuse, 1998). Researchers speculate that an abusive relationship between older adults may simply be a continuation of abuse that began earlier in a marriage or may begin in response to age-related stresses, such as retirement, failing health, caregiver burdens, or increased dependency. Historically, social and support services for abused older adults have been largely health related and there is scant help available for elders trapped in abusive intimate partner relationships.

Sarah Harris, in "For Better or for Worse: Spouse Abuse Grown Old" (Journal of Elder Abuse and Neglect, vol. 8, no. 1, 1996), used the 1985 National Family Violence Survey data to compare respondents under sixty years old to those over age sixty. Although the incidence of spouse abuse in older couples was significantly less than that of younger couples, many of the risk factors for violence were the same. Not surprisingly, abuse occurred most often in situations where there was a high degree of conflict.

The factors associated with older partner violence included lower education levels, lower family income, verbal aggression, drug abuse, depression, perceived stress, low use of reasoning tactics, and marital conflict. Racial/ethnic group affiliation also played a part: blacks and Hispanics in the younger group and blacks in the oversixty group were more likely to experience couple violence. When violence was reported for more than the twelve months preceding the survey, intergenerational violence and poor physical health were also found to be significant.

Rennison and Rand found that spouse abuse among their oldest cohorts was similar to that found in younger cohorts: most incidents (85%) were committed without a weapon, a substantial portion of the abusers used alcohol or drugs, and the likelihood of injuries across all age groups was similar. However, mature women were less likely to report the violence to police than younger women.


Predictive Characteristics and Risk Markers

Can a woman expect to see certain signs of potential violence in a man she is dating or living with before she becomes a victim of abuse? The National Coalition against Domestic Violence published a checklist of predictive behaviors in men that signal violence. (See Table 2.11 in Chapter 2.) Along with the predictors described in the checklist, there are other indicators, known as risk markers, which may indicate an increased propensity for violence. These include:

  • an unemployed male
  • a male who uses illegal drugs
  • males and females with different religious backgrounds
  • a male who saw his father hit his mother
  • male and female unmarried cohabitants
  • males with blue-collar occupations
  • males who did not graduate from high school
  • males between eighteen and thirty years of age
  • males or females who use severe violence toward children in the home
  • total family income below the poverty level

In "Men Who Batter: The Risk-Markers," Richard Gelles, Regina Lackner, and Glen Wolfner found that in families where two risk markers were present, there was twice as much violence as those with none (Violence Update, vol. 4, no. 12, 1994). In homes with seven or more of those factors, the violence rate was a staggering forty times higher.

A separate analysis by Hotaling and Sugarman surveyed risk markers present in more than four hundred studies. In "A Risk Marker Analysis of Assaulted Wives" (Journal of Family Violence, vol. 5, no. 1, 1990) and in "Prevention of Wife Assault" (Treatment of Family Violence [New York: Wiley, 1990]), the researchers found that abused women do not differ in specific personality traits from women who are not abused in terms of age, educational level, race, occupational status, length of time in the relationship, or number of children. Furthermore, a woman's poor self-esteem did not appear to be a risk factor but rather a consequence of abuse.

Contrary to other researchers, Hotaling and Sugarman found that being an abused child or seeing a parent being abused does not necessarily mean a woman has a greater chance of becoming a battered wife. Instead, severe male batterers can be distinguished from nonassaultive and verbally abusive men, as well as men who commit minor abuse, by the greater likelihood of having witnessed violence between their parents. In addition, men who engage in minor physical aggression are more likely to have experienced violence in the past than are men who were verbally abusive.

Hotaling and Sugarman also concluded that the only factor that differentiates abused wives from wives who are not abused is the level of marital conflict. Obviously, if a marriage has very little or no conflict, there is no reason or provocation for violence. The researchers maintained that while there is some level of conflict in every relationship, it does not necessarily result in violence. Individuals in adequately functioning relationships negotiate their way through disagreements, while individuals in violent relationships lack these skills and resort to violence.

Hotaling and Sugarman concentrated on four factors most often associated with abuse: marital conflict, the frequency of the husband's drinking, expectations about the division of labor in the relationship, and a measure of educational incompatibility. Hotaling and Sugarman concluded that to understand wife abuse, researchers are better served by consideration of the perpetrators' behavior rather than the characteristics of the victims.

risk markers and a continuum of aggression. Sugarman et al. used the analysis of risk markers to test the theory that there is a continuum of aggression in husband-to-wife violence that is linked to some of the risk markers. The risk markers considered were marital conflict, depressive symptoms, alcohol use, attitude toward interpersonal violence, violence in the family in which the individuals grew up, nonfamily violence level, and socioeconomic status. The researchers found that an increase in the severity of husband-to-wife violence was associated with an increase in depressive symptoms in the husband, along with his greater acceptance of marital violence and a higher likelihood that he experienced and witnessed violence in his family as a child. In addition, greater alcohol use and higher levels of nonfamily violence by the couple were linked to more severe violence.

Consistent with other research, Sugarman et al. concluded that persons who engage in minor violence do not necessarily progress to severe violence, but those who use severe violence almost always began with minor violence. The most important implication of this finding is that early intervention may prevent more severe abuse. Most treatment programs are only initiated after a woman has suffered severe battering. Prevention programs that emphasize the importance of seeking treatment for lowlevel abuse before it escalates to serious violence might encourage women to escape abuse before it claims their health or their lives.

A study by Jacquelyn C. Campbell et al. titled "Assessing Risk Factors for Intimate Partner Homicide," evaluated the risk factors among abused women for being killed by their intimate partners (National Institute of Justice Journal, no. 250, November 2003). The researchers found that abused women whose abusers owned guns and who had threatened to kill them were at high risk of being killed by their intimate partners. (See Figure 3.2.) Other


high risk factors for homicide included extreme jealousy, attempts to choke, and marital rape. The study authors hope that the "danger assessment" tool they used may assist women and advocates for battered women to better assess the level of risk in abusive relationships.

About this article

The Causes of Wife Abuse

Updated About content Print Article