Cover Image
close this bookViolence against Women (WB)
View the document(introduction...)
View the documentInformation
View the documentForeword
View the documentAcknowledgments
View the documentAbstract
View the document1. introduction
View the document2. The scope and evolution of the problem
View the document3. A primer on violence against women
View the document4. health consequences of gender-based violence
View the document5. Implications of gender violence for health and development
View the document6. Steps toward eliminating violence against women
View the document7. Research needs
View the document8. Conclusions
Open this folder and view contentsAppendix
View the documentNotes
View the documentBibliography

2. The scope and evolution of the problem

In the past decade violence against women has become increasingly recognized as deserving international concern and action. Women's organizations around the world embraced gender violence as a priority issue during the United Nations Decade for Women (1975 to 1985). The United Nations (UN) General Assembly passed its first resolution on violence against women in November 1985. Since then, the UN has sponsored several Expert Group Meetings on Violence against Women and pursued the issue through its Commission on the Status of Women, the Economic and Social Council, the UN Statistical Office, and its Committee on Crime Prevention and Control. Recently two new international instruments have been put forward that would recognize all gender-based violence as an abuse of human rights-the UN Declaration on Violence against Women and the farther-reaching Draft InterAmerican Convention to Prevent, Punish, and Eradicate Violence against Women (negotiated through the Organization of American States). Also, the Pan American Health Organization has recognized gender based violence as its priority theme for 1994 under its Women, Health, and Development Program, and the United Nations Fund for Women (UNIFEM) recently published a major document outlining the impact of gender violence on socioeconomic development (CarriIlo 1992).

The growing international recognition of the importance of gender-based abuse comes on the heels of almost two decades of organizing by women's groups to draw attention to the issue. Women have been saying in a multiplicity of ways that violence is a major concern for them (box 1). Recently more than 200 women's nongovernmental organizations (NGOs) combined forces to protest violence against women during Sixteen Days of Activism against Gender Violence (November 25 to December 10). During this annual event groups sponsor workshops, conferences, and street theater and organize media coverage to raise public awareness of gender violence and to demand a response from public officials.

Definition of violence against women

In September 1992 the United Nations Commission on the Status of Women convened a special working group to draft a declaration against violence against women This declaration, adopted by the General Assembly in the fall of 1993, offers for the first time an official UN definition of gender-based abuse. According to Article 1 of the declaration, violence against women includes: any act of gender-based violence that results in, or is likely to result in, physical, sexual or psychological harm or suffering to women including threats of such acts, coercion or arbitrary deprivations of liberty, whether occurring in public or private life. (Economic and Social Council 1992)

Article 2 of the declaration states that the definition should be understood to encompass, but not be limited to, physical sexual, and psychological violence occurring in the family and in the community, including battering, sexual abuse of female children dowry-related violence, marital rape, female genital mutilation and other traditional practices harmful to women, non-spousal violence, violence related to exploitation, sexual harassment, and intimidation at work, in educational institutions, and elsewhere, trafficking in women, forced prostitution, and violence perpetrated or condoned by the state.

Significantly, this definition recognizes both physical and psychological harm and threats of such harm in both the public and the private sphere. The definition also refers specifically to the gender-based roots of such violence. In this paper we use the United Nations definitions, but we also include a discussion of definitional issues in appendix B. The catalogue of abuses in the UN definition is not exhaustive, nor does it presume to be. For the sake of brevity, the paper will not address in depth certain forms of violence, including forced prostitution, sexual harassment, trafficking in women, or violence perpetrated by the state. This omission in no way reflects the relative importance of these issues. In fact, all forms of gender-based violence have common roots and can best be understood as points on a spectrum.

Dimensions of the problem

Accurately estimating the global health burden of violence against women is hampered by lack of data on the incidence and the health impact of abuse. Crime statistics are virtually useless in estimating the incidence of gender based abuse because of gross underreporting. According to recent victimization surveys in the United States, only 2 percent of intrafamilial child sexual abuse, 6 percent of extrafamilial sexual abuse, and 5 to 8 percent of adult sexual assault cases are reported to the police. By comparison, 61.5 percent of robberies and 82.5 percent of burglaries are reported (Koss 1990). Nonetheless, significant progress has been made in recent years in estimating the prevalence of wife abuse in both industrial and developing countries. Because of the stigma associated with sexual violation, data on rape and sexual abuse are less easily collected, although large-scale epidemiologic surveys of sexual assault are beginning to emerge in industrial countries.

The following section reviews the data available on the different types of violence covered in this report: abuse of women, sexual assault, sexual abuse of children, neglect of girl children, and culture-bound practices that are harmful to women. Box 2 presents an overview of violence as it occurs through the life cycle. A life-cycle approach to gender-based victimization provides important insights into the immediate as well as the cumulative effects of violence on the lives of women and girls. Violence can occur during any phase of a woman's life; many women experience multiple episodes of violence throughout their lives. A life-cycle perspective also reveals that violence experienced in one phase can have long-term effects that predispose the victim to severe secondary health risks, such as suicide, depression, and substance abuse. Evidence suggests that the earlier in a woman's life violence occurs especially sexual violence-the deeper and more enduring are its effects (Burnam and others 1988).

Prevalence of abuse by intimate partners

The most endemic form of violence against women is wife abuse-or, more accurately, abuse of women by intimate male partners. The 35 studies from a wide variety of countries summarized in table 1 straw that in many countries one quarter to more than half of women report having been physically abused by a present or former partner. An even larger percentage have been subjected to ongoing emotional and psychological abuse, a form of violence that many battered women consider worse than physical abuse. Although some of these studies are based on convenience samples, most are based on probability samples with a large number of respondents (Colombia, Kenya, Mexico, the United States).

Box 1 Evidence of women's interest in gender-based violence

When MATCH International, a Canadian funding organization, surveyed women's groups in developing countries for suggestions on future funding priorities, violence against women was the number one reply (Carrillo 1992).

At a recent 12-country workshop held in China on women's non-formal education, participants were asked to name the worst aspect of being female. The almost unanimous answer was fear of men's violence (Bradley 1990).

At a accent meeting on women's reproductive health in Asia, sponsored by the Population Council and the Indonesian Epidemiology Network violence was identified as one of two priority areas (Population Council 1991).

At the November 1991 biannual conference of the Association of Women in Development women as one of their five priority health concerns. More abstracts were submitted on issues relating to violence than on any other single theme (Marcelo 1992).

The National Black Women's Health Project has identified violence as the number one health issue for African American women in the United States (Avery 1990).

Women around the world collected more than 400,000 signatures (representing 124 countries) on a petition demanding that the United Nations Human Rights Conference held in June 1993 recognize violence against women as an abuse of women's human rights (Center for Women's Global Leadership 1993).

When the National Council for Women's Rights in Brazil sponsored a meeting in 1986 to develop a list of women's demands for the upcoming constitution writing process, women demanded a new constitutional clause recognizing violence against women (Pitanguy, personal communication, 1993).

Each of the studies is individually valid, but they are not directly comparable because each uses a different set of questions to probe for abuse. Most of the studies ask the respondents whether they have been "abused,. "beaten," or "involved in a violent relationship. " A subset (the studies from Barbados, Chile, and the United States) makes this determination using a list of "acts" that a woman may or may not have been subjected to during her lifetime (being hit with an object or fist, being bitten)." Clinical and research experience suggests that question formats that require a woman to self-identify as abused generally underestimate the physical and psychological violence in intimate relationships. In many cultures women are socialized to accept physical and emotional chastisement as a husband's marital prerogative, limiting the range of behavior they consider abuse. Moreover, women are sometimes reluctant to report abuse out of shame or out of fear of incriminating other family members. Both factors suggest that the prevalence rates in table 1 likely underestimate the abuse of women.

Prevalence of rape and sexual assault

Statistics from around the world suggest that sexual coercion is common in the lives of women and girls. Six well-designed studies from the United States, for example, suggest that between one in five and one in seven U.S, women will be the victim of a completed rape in her lifetime (Koss 1993; Kilpatrick, Edmunds, and Seymour 1992).3 The U.S. data are consistent with studies of rape in other parts of the world. Studies of rape among college-age women in Canada, New Zealand, the United Kingdom, and the United States reveal remarkably similar rates of completed rape across countries (DeKeseredy and Kelly l993; Gavey 1991; Beattie 1992; Koss, Gidycz, and Wisniewski 1987; table 2). A study among adult women (many of them college students) in Seoul, Korea, yielded a slightly lower rate of completed rape, but an equally high rate of attempted rape (Shirn 1992). All of these studies used adaptations of the same survey instrument, based on the Sexual Experiences Survey (SES) by Koss and Oros (1982).

The estimates in table 2 are based on legal definitions of rape in the United States, which recognize as rape the penetration of any orifice by physical force or threat of force, or because a woman is incapacitated due to drugs or alcohol. (For comparison, the U.K. data include forced anal and oral penetration, although the legal definition of rape in the United Kingdom includes only vaginal penetration by a penis.)

Box 2 Gender violence throughout the life cycle


Type of violence present


Sex-Selective abortion (China, India, Republic of Korea); battering during pregnancy (emotional and physical effects on the. woman; effects on birth outcome); coerced pregnancy (for example, mass rape in war).


Female infanticide; emotional and physical abuse; differential access to food and medical care for girl infants.


Child marriage; genital mutilation; sexual abuse by family members and strangers; differential access to food and medical care; child prostitution


Dating and courtship violence (for example, acid throwing in Bangladesh, date rape in the United States); economically coerced sex (African secondary school girls having to take up with "sugar daddies" to afford school fees); sexual abuse in the workplace; rape; sexual harassment; forced prostitution; trafficking in women.


Abuse of women by intimate male partners; marital rape; dowry abuse and age murders; partner homicide; psychological abuse; sexual abuse in the workplace; sexual harassment; rape; abuse of women with disabilities


Abuse of widows; elder abuse (in the United States, the only country where data are now available, elder abuse affects mostly of women).

Table 1.1 Prevalence of wife abuse, selected countries

Table 1.2 Prevalence of wife abuse, selected countries

Table 1.3 Prevalence of wife abuse, selected countries

Table 1.4 Prevalence of wife abuse, selected countries

Table 2 Prevalence of rape among college-age woman selected countries

Women have also been subjected, throughout history, to repeated and especially brutal rape as part of war. In recent years mass rape in war has been documented in Bosnia, Cambodia, Liberia, Peru, Somalia, and Uganda (Swiss and Giller 1993). A European Community fact finding team estimates that more than 20,000 Muslim women have been raped in Bosnia since the fighting began in April 1992. Many have been held in "rape camps". where they have been raped repeatedly and forced to bear Serbian children against their will (Post 1993). These examples notwithstanding, rape in war is neither a new phenomenon nor one limited to developing countries.

Prevalence of child and adolescent sexual abuse

Because the sexual abuse of children is such a sensitive issue, there are few population-based studies from which its prevalence can be estimated. The few studies that do exist and ample indirect evidence suggest that sexual abuse of children and adolescents is widespread In the United States, for example, studies show that 27 to 62 percent of women recall at least one incident of sexual abuse that occured before they were 18 (Peters, Wyatt, and Finkelhor 1986).' An anonymous, island-wide probability survey in Barbados revealed that one woman in three and one to two men in 100 report haying been subject to behavior constituting childhood or adolescent sexual abuse (Handwerker 1993a). In Canada a government estimated that one in four female children and one in 10 male children are sexually assaulted before age 17 by 1984).

The indirect evidence available elsewhere also suggests cause for concern. Two studies from Nigeria documenting sexually transmitted diseases (STDs) in very young children suggest that sexual abuse is at least present in Nigerian society. A 1988 study in Zaria, Nigeria, found that 16 percent of female patients seeking treatment for STDs were children under age five and another 6 percent children between ages six and 15 (Kisekka and Otesanya 1988). An older study in Ibadan found that 22 percent of female patients attending one STD clinic were children under age 10 (Sogbetun, Alausa, and Osoba l977). In Peru a study conducted in the Maternity Hospital of Lima revealed that 90 percent of the young mothers age 12 to 16 had been raped-the vast majority by their father, their stepfather, or another close relative (Roses 1992). An organization for adolescent mothers in Costa Rica reports similar findings: 95 percent of its pregnant clients under 15 are victims of incest (Treguear and Carro 1991).

This indirect evidence is consistent with cross-cultural data from rape crisis centers which reveal that 40 to 58 percent of sexual assaults are perpetrated against girls age 15 and under, including girls younger than 10 or 11. Most of these rapes are committed by family members or other persons whom the victim knows. In fact, justice statistics and data from rape crisis centers show that in more than 60 percent of all rape cases the victim knows the perpetrator (table 3). The common perception of rape as a "stranger" crime is sorely misguided.

A final indication of the prevalence of sexual abuse comes from the observations of children themselves. In 1991, when Centro de Informaci Servicios de Asesorin la Salud (CISAS), a Nicaraguan health NGO, held a national conference for the children involved in their CHILD to Child program (a project that trains youngsters age 8 to 15 to be better child care providers for their siblings), participants identified physical and sexual abuse as the most important health problem that young people in their country faced. Since then, CISAS has helped initiate a national campaign to educate the public about sexual abuse of children ("Rompiendo el Silencio," 1992).

Prevalence of discriminatory treatment and infanticide

The preference for sons common in many cultures can have serious consequences for the health and lives of females. Studies show that where this preference is strong and resources are scarce, girls receive less food, education, and medical care than boys. In rural Bangladesh malnutrition was found to be almost three times more common among girls than among boys (Bhatia 1985; Chen, Huq and D'Souza 1981). Even more striking are differences in access to health care. In Matlab, Bangladesh, boys outnumber girls at diarrhea! Treatment centers by 66 percent, even though both sexes get diarrhea with equal frequency (Bhatia 1985).

Not surprisingly, this discriminatory treatment shows up in mortality statistics for girls and women. In 45 developing countries for which recent data are available, the mortality rate among girls age one to four is higher than that among boys in this age group in all but two countries (UNICEF 198 6). Discrimination against girl children is so strong in the state of Punjab in India, for example, that girls age two to four die at twice the rate that boys in this age group do (Das Gupta 1987). According to a World Bank report, "deaths of young girls in India exceed those of young boys by almost one third of a million every year. Every sixth infant death is specifically due to gender discrimination" (Chatterjee 1990).

In some parts of the world the preference for male children is so strong that parents eliminate girl children through infanticide or selective abortion. A 1987 census

Table 3 Statistics on Sexual crimes selected countries

Percentage of perpetrators known to victim

Percentage of victims age 15 and under

Percentage of victims age 10 and under

Lima, Peru








Mexico City




Panama City




Papua New Guinea




Santiago Chile




United States




- Not available.

Note: Except for the U.S. data, which cover only completed rapes, data include rape and sexual assaults, such as attempted rape and molestation.

a. Percentage of survivors age nine and younger.
b. Percentage of survivors age six and younger.
c. Percentage of survivors age seven and younger.
d. Based on five-year averages derived from crimes reported to the Legal Medicine Service 1987-91, a. published in Anuano Estadistico del Servicio Medico Legal de Chile (as cited in Avendand Vergara 1992).

Percentage of survivors age 17 and younger. Source: Data for Malaysia from Consumer's Association 1988; data for Panama City from Perez 1990; data for Peru from Portugal 1988; data for Papua New Guinea from Riley, Wohitahrt, and Carred 1985, as cited in Bradley 1990; data for Mexico City from COVAC 1990 and Procurador de Justicia del Distrito Federal de Mexico 1990; data for Chile cited in Avandand Vargara 1992; data for United States from Kilpatrick, Edmunds, and Seymour 1992 survey by the State Statistical Bureau of Chins showed that were half a million fewer female than one would predict given the expected biological ratio of male to female births. A detailed analysis of the census data, published in the Population awl' Development Review, shows that the ratio of males to females has been rising since 1982 (Hull 1990). The missing female infants represent mainly second- and third-order births in rural areas. Both these observations strongly suggest that the one child policy implemented in 1979 in China has led to either increased female infanticide or the selective abortion of female fetuses.

Other studies confirm that in China, India, and the Republic of Korea access to amniocentesis and ultrasound is sufficiently widespread that feticide could in fact be skewing the male-female sex ratios (Coale and Banister 1992). Genetic testing for sex selection has become a Flourishing business in India, especially in the north Until protests from women's organizations stopped blatant promotion of the testing, Indian sex detection clinics advertised that it was better to spend $38 now on terminating a female fetus than $3,800 later on her dowry. One study of amniocentesis in a large Bombay hospital found that go 5 percent of fetuses identified as female were aborted, compared with only a small percentage of male fetuses (Ramanamma 1990).

Studies in remote regions of southern India also confirm the persistence of female infanticide. In a prospective study of all births within a 12-village region of Tamil Nadu, nutritionist Sabu George of Cornell University found that 58 percent of deaths (19 of 33) among female infants were due to infanticide. The girls were most often killed within seven days of birth; the two most common methods were to feed them the poisonous sap of a plant and to choke them by lodging rice hulls soaked in milk in their throats. There were no male infanticides during the two-and-half-year study (George, Abel, and Miller 1992).

An analysis first advanced by Harvard economist Amartya Sen and later refined by demographer Ansley Coale offers a simple but powerful illustration of the cumulative impact of these factors on women's survival prospects. Sen and Coale compared the sex ratios in countries where both sexes receive similar care, such as countries in Europe and North America, with those in countries where females are severely discriminated against, such as China and India. The female-male ratio in the first group of countries is about 1.05 or 1.06, reflecting women's biological advantage. But in South Asia, West Asia, North Africa, and China the ratio is typically 0.94 or lower. If these regions had the sex ratio typical of countries in which there is less gender discrimination (including some Sub-Saharan African countries), there would be more than 60 million more females alive today. The "missing women arc victims of feticide, infanticide, selective malnourishment, lack of investment in women's health, and various forms of gender violence (Sen 1990; Coale 1991)

Prevalence of culture-bound practices hark to women

A variety of other practices can reasonably fall under the rubric of gender violence if that term is defined to link physical and mental harm to women with male powering consul. In this section we advance the term "culture-bound practices harmful to women" as a replacement for the phrase commonly used in the United Nations system, untraditional practices harmful to women."
In UN parlance traditional practices refers primarily to such practices as genital mutilation and child marriage. Although these practices are known to have negative consequences for women's health and well-being, they are undertaken, often with women's collusion, to males girls more acceptable marriage partners.

We suggest broadening this concept to include harmful practices and behaviors common in industrial societies which are likewise motivated by a desire to make women into acceptable, attractive sexual partners for men. These include pathological dieting (anorexia and bulimia) and high-risk cosmetic surgery. There are, of course, significant differences among these practices, not the least of which is that genital mutilation is generally performed on children, with or without their consent. Nonetheless, we feel that genital mutilation should be viewed as one point on a continuum of harmful practices motivated by women's desire to conform to socially prescribed Words of beauty and marriageability.

Best estimates indicate that 85 million to 114 million girls and women in the world have undergone genital mutilation. Most of these girls and women live in Africa, a few live in Asia, and an increasing number live in Europe, Canada, and the United States, as immigrant and refugee families import the practice (Toubia 1993). Genital mutilation-frequently called female circumcision-is a broad term applied to a range of practices involving the removal of all or part of the clitoris and other external genitalia. In its most severe form,, letdown as infibulation, the clitoris and both labia are removed and the two sides of the vulva are sewn together, leaving only a small opening to allow urine and Dual blood to pass. In its less exam form, all or part of the clitoris is removed (ditoridectomy) orate clitoris and inner lips are removed (excision). About 85 percent of women who are mutilated undergo one of the two less severe operations.
Many observers trace the origins of genital mutilation to a desire to control female sexuality and to preserve the viscidity of young girls until marriage Medley and Dorkenoo 1992). A host of superstition help perpetuate the practice, but the core belief driving the tradition men will not marry uncircumcised women, believing them to be promiscuous, unclean, and sexually untrustworthy (Mohamud 1991).

In other societies in which virginity at marriage is considered absolutely essential, girls are often married off at extremely young ages, frequently to men many years their senior. Recent Demographic and Health Surveys from the developing world indicate that, al though child marriage is declining, a large percentage of young girls are still married off before their fifteenth birthday (table 4). Not uncommonly, these child brides are traumatized by adult sex and forced to bear children before their bodies are fully mature. A common side effect of too early childbearing is vesico-vaginal or recto-vaginal fistula-a tearing of the walls between the vagina and the bladder or rectum-due to prolonged obstructed labor and lack of access to health care (birth control, prenatal care, or surgical intervention). Women with unrepaired fistulae constantly drip urine and feces, making them social outcasts and likely candidates for divorce or abandonment.

Elsewhere young girls and women themselves put their health at risk-by severely restricting their dietary intake. In North America and Europe surveys of adolescent women suggest that roughly one of every 100 to 200 young women suffer from anorexia. This psychological disorder, restricted almost exclusively to girls and women in western industrial societies end Japan, is characterized by extreme weight loss induced trough gradual self starvation (Gordon 1990). As dieting transforms into fasting and finally into willful starvation, the anorexic typically withdraws from ordinary activities and relationships and intensifies an already excessive exercise routine. She becomes obsessed with dieting and counting calories and with the sight of her own image in the mirror.

Bulimia -a different but related disorder- is characterized by binge eating accompanied by purging through self-induced vomiting or ingestion of laxatives. Estimates using the most stringent criteria (binging and purging on a weekly basis, and a preoccupation with shape and body weight) put the prevalence of bulimia among U.S. high school and college females at between 3 and 5 percent, suggesting that several million young American women have a clinically significant problem with bulimia.

Table 4 Women age 20 to 24 today who were married before age 15, selected countries



Year of report




























Dominican Republic






Trinidad and Tobago



Source Selected Demographic and Health Surveys

The incidence of bulimia, like that of anorexia nervosa, has been increasing throughout the westerns world, a trend corresponding to the growing emphasis on thinness cultural and sexual ideal (Gordon 1990; Brumberg 1988). The same cultural obsession, with thinness and socially defined notions of beauty that promote eating disorders are prompting women in some western societies to seek cosmetic surgery, often at considerable risk to their physical health. An estimated 2 million U.S. women have undergone breast enlargement surgery, at an average cost of about $4,000 per operation. Amid much controversy, the U.S. Food and Drug Administration recently prohibited the use of silicone breast implants for cosmetic purposes, fearing that the products of silicone breakdown could be carcinogenic (implants may still be used for breast reconstruction after mastectomies). There is also concern that silicone implants complicate detection of breast cancer and may be linked to certain auto-immune diseases.