From gender issues to gender statistics on physical and sexual violence against women: illustrative examples
||Sources of data
|How widespread is physical or sexual violence against women? What groups of women have a higher prevalence of this type of violence?
||Women who have experienced physical and/or sexual violence in the past 12 months and during their lifetime. As much as possible, data should be disaggregated by age, relationship to perpetrator, educational attainment, ethnicity, migration status, wealth status of the household, urban/rural areas and geographical areas.
||Violence against women surveys.
Module on violence against women attached to household surveys, such as DHS or other health or social-related survey.
|How often do victims of violence against women report the violence incident or access other services of support?
||Women who have experienced violence in the past 12 months by type of violence, severity of violence and type of services accessed.
||Violence against women surveys.
|What are women’s and men’s attitudes toward wife-beating? In their view, is wife-beating in certain circumstances justified?
||Persons agreeing with a statement that wife-beating is justified in certain circumstances by sex.
||Household surveys, such as DHS.
Violence against women surveys.
- + Gender issues
- Physical and sexual violence against women is widespread, with prevalence levels varying considerably across countries (Garcia-Moreno and others, 2005; United Nations, 2010). The majority of women physically or sexually abused are abused by their own intimate partners. Many women who are victims of physical or sexual violence by their own intimate partners experience acts of violence more than once. Lifetime experience of sexual violence is reported as lower than physical violence in some countries, while higher in others.
Non-partner physical or sexual violence against women also varies widely among countries; however, it is less frequent than partner violence. For physical violence, commonly mentioned perpetrators include fathers and other male or female family members; in some countries, teachers are also frequently mentioned (Garcia-Moreno and others, 2005). For sexual violence, the perpetrators usually include strangers, male family members other than fathers, or male friends of the family. Early sexual abuse (before the age of 15) is often linked to a male family member other than the father or the stepfather (Garcia-Moreno and others, 2005).
Sexual harassment, taking place in the workplace or in other public spaces, tends to be less often reported and less often captured in statistics. Similarly, psychological and economic violence are also not reported as often as physical and sexual violence, but affect the well-being of many women. Women with physically and sexually abusive partners, for instance, may experience controlling behaviour such as the restriction of contact with her family and friends, insistence on knowing where she is at all times and the control of her access to health care.
Some groups of women may be particularly vulnerable to violence. The type of groups of women most at risk may vary from one country to another. For example, in some countries, a woman’s younger age at marriage is associated with higher prevalence of violence. Forced first sex is often associated with early age at first sex and, in some countries, is linked to sexual initiation in the context of early marriage (Garcia-Moreno and others, 2005). Lower educational attainment, in some countries, is also associated with higher prevalence of violence. Other groups of women are also considered to be vulnerable to violence, although the available data is limited. Indigenous women, women from ethnic minorities, poor women, migrant women, older women and women with disabilities, for example, may, in some countries, be subjected to violence more often.
Violence against women that occurs within the family or within the home is tolerated in many contexts. Attitudes towards wife-beating vary across countries (Hindin, Kishor and Ansara, 2008; United Nations, 2010). In some countries, women who justify wife-beating are a small minority, while in other countries they are the majority. Acceptance of wife-beating tends to be higher among women who have experienced abuse than among those who have not (Hindin, Kishor and Ansara, 2008). Also, in some countries, less educated women are more likely to embrace such attitudes than women with a higher level of education (Uthman, Lawoko and Moradi, 2009).
Nevertheless, men are also vulnerable to violence, although they are less likely than women to be victims of violence, especially victims of domestic violence. Over the life course, men’s vulnerability to domestic violence may be more pronounced in childhood, adolescence and at older ages; vulnerability to violence by non-family members may be more pronounced for young men.
- + Data needed
- Several types of data can be used to analyse physical and sexual violence against women. They are:
(a) Statistics on prevalence of violence against women, such as:
(i) Victims of physical violence in the past 12 months and victims of physical violence during lifetime;
(ii) Victims of sexual violence in the past 12 months and victims of sexual violence during lifetime;
(iii) Victims of physical or sexual violence in the past 12 months and victims of physical or sexual violence during lifetime;
The data above should be further disaggregated by sex, age, sex of the perpetrator, relationship with the perpetrator, frequency and/or severity of violence and types of injuries. Violence during lifetime may be collected separately for two age groups: 15 and above, and 15 and under.
(iv) Victims of psychological violence in the past 12 months by sex, age, sex of the perpetrator and relationship with the perpetrator;
(v) Victims of economic violence in the past 12 months by sex, age, sex of the perpetrator and relationship with the perpetrator;
(vi) Victims of sexual harassment in the past 12 months by sex, age, sex of the perpetrator, relationship with the perpetrator and place of harassment.
Additional breakdowns should be considered for statistics on prevalence of violence against women, such as urban/rural areas, geographical areas, ethnicity and migration status. Indicators of women’s empowerment, such as educational attainment, property ownership and paid employment, should also be used.
(b) Statistics on use of services by victims of violence, such as:
(i) Victims of violence by sex and age, type of violence, severity of violence, and type of services accessed (health, police, women’s NGOs, social services). Data on satisfaction with the services accessed should be included, if possible. Additional breakdowns should be considered, such as urban/rural areas, geographical areas, ethnicity and migration status. Indicators of women’s empowerment, such as educational attainment, property ownership and paid employment, should also be used;
(c) Statistics on reported incidents of violence, such as:
(i) Physical or sexual violence reported to the police by sex and age of the victim, type of violence, sex of the perpetrator and relationship with the perpetrator;
(ii) Victims reporting violence and accessing medical treatment in health facilities by sex and age of the victim and type of violence;
(d) Statistic on attitudes towards violence against women, such as:
(i) Persons considering that wife-beating for specific reasons is justified by sex and age. Specific reasons that have been used in surveys such as DHS include burning the food, arguing with the husband, refusing to have sex, going out without telling the husband and neglecting the children. Additional breakdowns should be considered, such as urban/rural areas, geographical areas, ethnicity and migration status. Indicators of women’s empowerment, such as educational attainment, property ownership and paid employment, should also be used.
- + List II.34
Examples of indicators derived from gender statistics on physical and sexual violence against women
Proportion of women aged 15 to 49 subjected to physical or sexual violence in the past 12 months by an intimate partner
Proportion of women aged 15 to 49 subjected to physical or sexual violence in the past 12 months by persons other than an intimate partner
Proportion of women who consider that wife-beating is justified for reasons such as going out without telling the husband
Note: See United Nations, Economic and Social Council (2009) and United Nations (2013) for a complete list of indicators related to violence against women.
- + Sources of data
- Dedicated surveys on violence against women are the preferred method of collection of data on violence against women. Although these specialized surveys can be relatively expensive, they provide an opportunity to collect more detailed data by exploring the topics in depth. They also suffer less from underreporting, owing to the careful selection and focused training of the interviewers. These surveys can provide information on the prevalence of various forms of violence against women and girls, including those occurring in the family or within the general community, characteristics of the victims and their households and characteristics of the perpetrators and their relationship to the victim. Information on violence against women related to intimate partner is the most widely collected. In some countries, surveys may collect data not only on violence against women, but also on violence against men (see also the section entitled “Surveys on violence against women” in chapter 3).
Modules on violence against women in health-related surveys, such as DHS or selected multipurpose surveys focused on social issues, can also be used to collect data on violence against women. Health surveys are a good vehicle for violence against women modules because they usually cover other similarly sensitive topics. In any surveys, however, attention should be paid to ethical issues and confidentiality, as well as to the safety of the respondents and interviewers, ensuring that the special features of the violence against women surveys are considered (such as the special training of interviewers and support to victims).
Administrative police and court records can provide information on reported incidents of violence. Although these sources have a limited value in estimating the prevalence of violence against women (a large proportion of incidents usually remain underreported), they are a valuable source of data on the use of services and the capacity of the system to respond to the problem.
Health administrative records may provide information on some forms of violence that required victims to seek treatment in hospital emergency rooms, family clinics or other health-care providers. They are an important source of information on the use of health services by victims of violence.
Administrative records from public, private or non-governmental agencies that provide support services to women victims of violence may provide data on the use of their services by victims of violence. The agencies usually included are emergency shelters, crisis centers, sexual assault /domestic violence phone lines, women’s NGOs and legal counsel and legal aid services.
- + Conceptual and measurement issues
- Surveys on violence against women may not adequately cover women who are particularly vulnerable to violence. Women belonging to minority groups, indigenous women, refugees, women migrants and older women are a relatively small proportion of the population and tend to be harder to reach. As a result, those groups of women are often not present in the sample in big enough numbers to allow for calculations and analysis of their specific levels of prevalence of violence.
Use of specific and detailed questions on various forms of physical, sexual, economic or psychological violence and focused training of interviewers increase the accuracy of statistics on violence against women. The use of a set of questions in the surveys instead of one general question reduces the chances of underreporting of violence and increases the comparability of statistics over time and by various groups of population. One of the concerns related to the use of modules on violence against women in existing household surveys is the increased likelihood of underreporting of violence. It is important that the modules used have an adequate number of well-tested questions rather than just a few added questions. In addition, it is important that interviewers are provided with additional training focused on violence against women. For more details on improving data collection in violence against women surveys, see the section entitled “Violence against women surveys” in chapter 3 of this manual, as well as United Nations (2014).
Police and court statistics usually tend to underestimate the prevalence of violence against women. Police and court records are necessarily based on the law, such as the penal or criminal code and family violence law or domestic violence law. If there is no law that qualifies or specifies acts against women as a crime, then there is no legal basis for filing a complaint. Even where a law exists, it is widely recognized that only a small proportion of crimes of violence against women makes their way into the administrative records of criminal justice systems. Violence against women occurring within the family may not be reported either because it is considered normal and is therefore tolerated or for such reasons as fear of reprisal, stigma, distrust of the police or legal system or lack of knowledge of legal rights. Violence against women in the general community is also likely to be underreported. Many women victims of rape, physical or sexual assault or sexual harassment do not report those crimes to the criminal justice system because of fear of reprisal, stigma or fear of not being believed or even of being blamed.
Some police and court statistics may not be disaggregated by basic characteristics such as sex, age and relationship between the victim and the perpetrator.
Statistics based on health records may underestimate violence against women because the victims may not identify the violence as the underlying cause of their injury or because not all health-care providers may record this type of information.
Surveying populations under the age of 15 on violence poses safety and ethical issues, including challenges of laws requiring mandatory reporting if minors report abuse. Consent forms and ethical guidelines may need to be developed (see Reza and others (2007) for examples).