Navigation – Plan du site

AccueilNumérosVol. 58 - n° 4Dossier-débat : le gouvernement p...Cultural Dimensions of Power/Know...

Dossier-débat : le gouvernement par les indicateurs

Cultural Dimensions of Power/Knowledge: The Challenges of Measuring Violence against Women

Les dimensions culturelles du savoir/pouvoir. Les enjeux de la mesure de la violence envers les femmes
Sally Engle Merry
p. 370-380


La gouvernance internationale, comme tous les autres modes contemporains de gouvernement, s’appuie de plus en plus sur des mesures quantitatives. Des questions comme la corruption, l’état de droit, les performances scolaires, le respect des droits de l’homme, donnent lieu à des traductions en nombres et indicateurs. Dans la mesure où la quantification est devenue centrale dans ces formes de gouvernement, il est essentiel d’examiner les cadres théoriques culturels et sociaux au sein desquels les systèmes de mesure sont développés. Une comparaison de quatre approches culturellement distinctes de la mesure de la violence envers les femmes au niveau international met en évidence des différences significatives quant à ce qui est mis en visibilité et ce qui est occulté par chacune. Les organisations qui promeuvent ces différentes approches varient en outre significativement en termes de pouvoir et de ressources. Les mesures conçues au sein des organisations les mieux dotées tendent à dominer la définition d’un phénomène tel que la violence faite aux femmes et la façon dont il est compris. Finalement, ces processus façonnent la manière dont la question devient objet de politiques. Dans la mesure où la réglementation et le gouvernement dépendent de ce que les données quantitatives rendent visible, ces glissements ont des conséquences sur les pratiques de la gouvernance internationale.

Haut de page

Notes de la rédaction

Article received on 15 June 2016; accepted on 22 July 2016.

Notes de l’auteur

This research was supported by the National Science Foundation’s Law and Social Sciences Program and the Science, Technology, and Society Program, #SES-0921368. Two subsequent National Science Foundation grants to develop networks of scholars working on indicators, with Co–Principal Investigators Benedict Kingsbury and Kevin Davis, brought together scholars for conferences and edited book projects on issues of quantification, #SES-1023717 and #SES-1123290.

Texte intégral

1International governance, like all contemporary modes of governance, increasingly operates by means of quantitative measurements. Issues such as corruption, the rule of law, academic achievement, compliance with human rights norms, and accountability are generally translated into numbers and indicators. United Nations reform initiatives typically include requests for data on the problem being addressed. For example, a 2006 report by the Secretary-General on violence against women listed the need for more data as one of five critical dimensions of action on the problem. Evidence-based decision-making, experimentalism, audit mechanisms, results-based management, and new public management are emerging forms of governance that rely extensively on measurement and counting. All of these forms of governance require knowledge that is classified, categorized, and arranged into hierarchies. Indicators are part of a regime of power based on the collection and analysis of data and its representation.

2Rather than revealing truth, however, indicators create it. What is produced is not simply a fiction, but a particular way of dividing up and making known one reality among many possibilities. As indicators cross the gap from social science knowledge to that used by policy makers and the public, the drawbacks and complexities recognized by their creators, such as limited data, the use of proxies, and the uncertainty of flawed or missing data, are typically stripped away. The indicators are presented as unambiguous and objective, grounded in the certainty of numbers. In this form, they act to produce a truth of the world despite the pragmatic compromises inevitable in their creation. And this truth can be quite misleading, as Morten Jerven shows in his analysis of the lack of accurate information on African economies and the impact this has on development planning (Jerven, 2013).

3Despite these challenges, global development is increasingly organized around quantitative measures. In 2000 the United Nations launched the Millennium Development Goals (MDGs), with eight goals, 18 targets, and 48 technical indicators to structure development efforts for the next fifteen years. The MDGs are credited with having had an enormous impact on development planning. In 2016, the UN finalized a new set of goals for the next fifteen years, called the Sustainable Development Goals (SDGs). Experts from the United Nations, IMF, OECD and the World Bank developed goals, targets, and indicators after a two-year, open-ended process of consultation with civil society and states. Experts took the broad aspirations, divided them into more specific targets, and then developed specific indictors to assess compliance with these targets. As of February 2016, there were 17 goals, 169 targets, and 231 indicators. Many civil society groups complained that the indicators failed to cover the full range of aspirations in the goals and targets, despite the daunting measurement challenge posed by the number of indicators already accepted.

4The issue of violence against women offers an example of the way broad aspirations are narrowed in the conversion to indicators. Goal 5: Achieve Gender Equality and Empower all Women and Girls, is divided into six targets. One of them, Target 5.2, is “Eliminate all forms of violence against all women and girls in the public and private spheres, including trafficking and sexual and other types of exploitation.” It is measured by two indicators: 5.2.1. “Proportion of ever-partnered women and girls aged 15 years and older subjected to physical, sexual or psychological violence by a current or former intimate partner, in the last 12 months, by form of violence and age group,” and 5.2.2. “Proportion of women and girls aged 15 years and older subjected to sexual violence by persons other than an intimate partner, in the last 12 months, by age group and place of occurrence.” Thus, the very broad goal is subdivided into a more limited target which is measured by indicators that measure only a small part of the issues in the target. The indicators measure only frequency with reference to age, type of violence, and type of relationship with location added for non-intimate violence. Many important features are left out as is trafficking and other forms of exploitation.

5These indicators are designed for ease of counting and for the ability to travel across class, cultural, and national lines. But in the course of this translation into apparently measurable categories, much is lost about the social and cultural organization within which this violence takes place. The experience of violence, the trajectory of relationships, the cultural meanings of gender, marriage, and sexuality, are not measured, nor is the social context of the violence, the culturally prevailing forms of marriage and kinship, the general level of tolerance for gender-based violence, and the role of the state in engaging in or prosecuting gender-based violence. Any system of indicators represents a cultural choice of what to measure and how to count it. It is important to recognize that the sacrifice of context, history, and culture is inevitable in converting social life into numbers. Numbers provide a valuable way to assess problems, push governments to act, and create issues as targets for public debate and action. But they simultaneously obscure and neglect important dimensions of social life and experience.

6Clearly there has been a proliferation of indicators in the field of global governance. This raises critical questions about how numbers are developed and how the use of numbers affects governance. Since indicators contain implicit theoretical arguments about social change, about what is valued and what is not, and even about what is worth paying attention to, they have both knowledge effects and governance effects. Yet, the socially constructed and culturally informed nature of these systems of measurement is insufficiently recognized by those who use them to govern. It is therefore critical to ask how this technology shapes knowledge and how that knowledge is used in governance. This essay endeavors to raise these important questions and to show that the power to construct knowledge also shapes patterns of governance.

7Indicators are developed by states, international organizations, non-governmental organizations, and businesses. To understand the relationships of power and knowledge in the domain of quantification, it is necessary to consider the institutional context. Organizations that develop and use indicators differ significantly in their capacity to mobilize the resources, expertise, and legitimacy necessary to gather data, create and disseminate indicators, and shape decision-making. More powerful and wealthy organizations and states produce more influential indicators than smaller, poorer ones. My recent book analyzing indicators of violence against women, trafficking in persons, and human rights violations shows that the vast global inequalities in power and expertise structure indicator production and use (Merry, 2016). The relative capacity of organizations to produce indicators and data is an important dimension of the power/knowledge system (see further, Merry et al., 2015 and Rottenburg et al., 2015).

8Power relations also affect what gets measured. What is measured and counted by states and civil society organizations depends on which problems seem politically important. There is considerable interest in measuring human and drug trafficking, for example, but little on children’s right to express their opinion. Moreover, since politically important issues have been measured in the past, they are easier to measure again since data and templates already exist. For example, in our study of the introduction of a manual for measuring child rights in Tanzania, Summer Wood and I found that there was considerable data in the country on basic health issues such as the prevalence of breastfeeding, long a concern of governments and donors, but little on the obligation to establish a positive agenda for children’s rights or on children’s right to play (Merry and Wood, 2015). These issues were not as politically important as children’s health and survival. Deciding to expend scarce resources to count something depends on the importance of the issue. But what it is not counted is often not noticed.

1. Measuring Violence against Women

9The global project to measure violence against women provides a useful example of the impact of institutional power on measurement processes. As violence against women has emerged as a global problem, there have been a variety of efforts to measure it. But these measurement systems confront difficult questions. What is violence against women? How should we count it? These apparently simple questions pose enormous challenges to measurement, particularly at the global level where conceptions of gender-based violence and of the relationships and social structures within which it occurs are highly variable. What constitutes violence against women is highly contested, ranging from narrow definitions of rape and intimate partner violence to broad conceptions that include sexual slavery in wartime, female genital cutting, sex trafficking, child marriage, and violence in police custody. Here I discuss four different approaches to defining the phenomenon and the institutional support behind each one. As this case study shows, each approach defines the scope of the problem and the topics to be measured differently and each is embedded in a different institutional framework with its own resources and expertise. This comparison dramatically illustrates the role that institutional power plays in defining and counting problems and shaping approaches to governance. More powerful institutions are likely to establish the dominant measurement system and therefore the prevailing approach to governance.

10My inquiry began from my ethnographic research on a project of the UN Statistical Commission (UNSC) to develop a set of indicators and guidelines for collecting data on violence against women that could be used by countries around the world. The UN General Assembly requested the indicators and guidelines in 2006. Developing guidelines and a proposed survey instrument meant determining how to classify all forms of violence universally, a daunting project since violence is a culturally constituted category that takes many different forms and has a wide array of cultural meanings depending on context, relationships, and ideas of kinship and discipline. As the UNSC decided what to measure, how to categorize it, and how to gather data, it, in effect, defined what violence against women is.

11However, the definition was not uncontested. Although the statisticians opted for a narrow definition of violence against women, feminists and human rights organizations pressured the statisticians to take a broader, more socially grounded conception. The feminist approach emphasized gender equality and structural violence while human rights groups focused on state responsibility. The UNSC zeroed in on interpersonal relationships. In order to show the way measurement systems are competing interpretive projects, this article compares the UNSC’s statistical capacity approach with three others. One is based on gender equality, one on human rights, and one on criminal justice. These four approaches display significant cultural and interpretive differences in the way they conceive, categorize, and count violence against women. Each is a product of its historical origins, the institutions that promote it, the disciplinary expertise of those who develop it, and the audience for which it is intended. And each contains a theory about how to govern violence against women even though these theories are rarely made explicit. One framework assumes that increasing gender equality is necessary to reduce violence against women while another emphasizes better policing, for example. Ultimately, how violence against women is understood at the global level depends on which framework prevails. And which prevails depends largely on institutional resources and power.

12The four frameworks take different approaches to gathering and analyzing information and constructing indicators. The differences are institutional as well as theoretical. Each comes from a different UN organization and academic discipline. They have signifycantly different resources available to them for gathering and analyzing data and creating indicators. Both criminal justice and national statistical offices are better funded to collect and analyze data than advocates of gender equality or human rights. Consequently, it is these conceptions that provide the bulk of information available on violence against women. It seems likely that their narrower, more individualistic definitions of violence against women will prevail over the broader, more structurally embedded ones of the latter.

1.1. Gender Equality

13The underlying theoretical framework for many of the feminist advocates focusing on violence against women is that it is linked to inequality between men and women. In the 1970s, feminists saw violence as a key mechanism for maintaining gender subordination. This perspective was developed through a series of global conferences on women from 1975 until 1995, culminating in the Beijing Platform for Action of 1995 which identified violence against women as one of twelve major issues for women. Feminists working on violence against women tend to advocate gender equality as a solution. Gender equality models see violence against women as the product of entrenched patterns of gender inequality and social tolerance of violence (see Merry, 2009). This contrasts with the UNSC view that violence against women depends largely on the nature of their interpersonal relationships, especially intimate ones. Surveys produced within this framework address social tolerance of violence against women by asking about attitudes toward this violence as well as a wide range of other dimensions of violence, such as stalking, controlling behaviors, intimidation, and isolation. The implication for governance is the need for a broad program to achieve economic, political, and social equality for women and to prevent gender-based discrimination in many spheres of social life.

14The gender equality approach to measuring violence against women is institutionally supported within the UN system by Commission on the Status of Women, the Convention on the Elimination of all forms of Discrimination against Women (CEDAW), the women’s secretariat UN DAW, the Special Rapporteur on Violence against Women, and the European (UNECE) and Latin American and Caribbean regional commissions (UNECLAC), among others. Since the 1990s, the CEDAW committee has queried countries about their work on violence against women and advocated gender equality. Europe has already developed considerable statistical capacity in the area of measuring violence against women and the UNECE is a leader in gender statistics. The 2006 Secretary-General’s in-depth report on violence against women took a gender equality approach as well. Thus, this theoretical framework is widely supported by UN institutions and some regional ones, as well as many feminist organizations. However, none of these organizations has the resources or expertise to carry out significant global data collection and analysis.

1.2. Human Rights

15The human rights system focuses on individual rights and the protection of the body as well as the plight of vulnerable populations such as women and children. Human rights work takes a broad and inclusive definition of violence against women in a variety of contexts that includes both state and non-state actors’ violence. Violations include a wide range of behavior encompassing female genital cutting, killings in the name of honor, rape in wartime, sexual violence against women by prisons and police, and domestic violence. It advocates dealing with apparently disparate issues such as domestic/family violence and trafficking through an integrated approach (Erturk, 2008b, p. 15).

16The human rights perspective is similar to that of gender equality, although it begins from the premise that violence against women violates women’s fundamental freedoms and human rights and that it is the responsibility of the state to protect these rights. It seeks to hold states accountable for protecting and fulfilling rights. While the gender equality perspective views victims of violence as embedded in unequal social structures, the human rights approach sees victims as entitled to state protection from the discrimination and violence that produces inequality. States are obligated to exercise due diligence to prevent violations, even by non-state actors.

17The Secretary-General’s 2006 report takes this position, as do the Special Rapporteurs on Violence against Women appointed by the Human Rights Council. Yakin Erturk, who served from 2003-2009, presented a major report on indicators of violence against women to the Council in 2008 that included measures of state action such as ratification of relevant international and regional laws, the criminalization of perpetrators, programs on awareness raising and prevention, an index of support services, and action plans on access to justice, reporting, and training (Erturk, 2008a; Erturk, 2008b, p. 90).

18In a project to develop indicators for major human rights, the Office of the High Commissioner for Human Rights (OHCHR) developed a set of indicators for violence against women that includes sexual and reproductive health and harmful traditional practices, violence at work, forced labor and trafficking, community violence and abuse by law enforcement officials; and violence and (post) conflict and emergency situations (UNSD, 2013). These indicators are designed to be used by countries when they report to the committees that monitor compliance with human rights treaties such as CEDAW. Thus, the institutional support for the human rights perspective includes the Human Rights Council, human rights treaties and treaty bodies, the Special Rapporteurs on Violence against women, and the OHCHR. This approach defines violence against women broadly and presents it in a social and political context that emphasizes the responsibility of the state to ameliorate the violence. Again, none of these organizations has the bureaucratic capacity to carry out surveys and produce data.

1.3. Criminal Justice

19Measuring gender-based violence through crime victimization surveys builds on a history of surveys since the 1960s when they became popular as a way of assessing the extent of crime and providing a victim’s perspective. National surveys were redesigned to collect information on women’s experiences of sexual and domestic violence by the early 1990s. A 1992 redesign in the US surveys resulted in significant increases in data on the prevalence of sexual and domestic violence (Johnson et al., 2008, p. 12). The British Crime Survey was remodeled in 1996 to incorporate domestic violence. Statistics Canada was the first national statistical agency to overhaul its approach by adding a dedicated survey of women’s experiences of violent victimization. It used the basic crime victimization survey as a starting point (Johnson et al., 2008, pp. 11-13).

20One of the important institutional supports for this approach is the UN Office of Drugs and Crime (UNODC). UN Surveys on Crime Trends and the Operation of Criminal Justice have been carried out regularly since 1984. For these surveys, the UNODC sends questionnaires to member states which gather the information. In 1989, the International Crime Victimization Survey (ICVS) began to collect data every few years on eleven forms of crime in over 60 countries. However, this general victimization survey did not initially measure crimes of violence against women. A series of UN conferences and resolutions beginning in the mid-1980s have addressed violence against women as a crime (Johnson et al., 2008, p. 7). By the end of the 1990s, the problem had moved into the mainstream of criminal justice concerns and statistics.

21However, criminal justice surveys confront questions about the definition of violence against women. Crime victimization surveys are designed to survey the incidence of crimes, but violence against women is not “self-defined” as a criminal act. Some commonly referenced forms of violence against women such as sexual harassment, sex trafficking, and marital rape are not crimes everywhere. Some forms of violence are condoned or perpetrated by states, such as forced sterilization or the violence of police or prison guards. Thus, deciding what counts as violence against women for a global criminal justice survey is difficult. At a 2007 UNSC expert group meeting to develop violence against women indicators, the UNODC suggested using the following categories, based on a review of international crime victimization surveys and specialized surveys on violence against women: homicide, rape, major assault, assault, sexual assault, harassment, female genital mutilation/cutting, and trafficking in persons (Malby et al., 2007, p. 10). Clearly, this effort to produce a generic set of acts of violence against women which are crimes misses many acts considered violence against women by feminists and human rights advocates. At the same time, even this limited list includes acts that are not crimes in some countries.

22Crime victimization surveys typically focus not only on the incidence of crimes but also on the adequacy of police responses to crimes. Thus, their impact on governance focuses on improving police and legal services rather than reducing discrimination or mitigating gender inequality in work and education.

1.4. National Statistics

23The national statistical approach encourages national statistical offices to develop regular mechanisms for measuring violence against women. It offers training and support to national statistical offices and develops guidelines for these offices to use in carrying out a survey. The major concern is persuading states to carry out a survey on violence against women on a regular basis, to develop administrative statistics, and to foster as much comparability among nations in their measurement strategies as possible so that countries can assess their policies and their accomplishments.

24In order to support and facilitate work done by national statistical offices and to foster comparability among national statistical surveys and administrative data collection, the United Nations Statistical Division (UNSD), the secretariat of the UNSC, creates guidelines for data collection and analysis. The UNSD develops a common definition across countries and seeks national buy-in so national offices will carry out the data collection. It is primarily interested in providing technical assistance to national statistical bureaus through developing guidelines and making recommendations to national statistical offices. It is concerned that the indicators it proposes are technically workable, affordable, and acceptable so that governments will use them. The guidelines produced by the UNSD in 2013 proposed a list of core indicators that categorize violence against women by age, type of relationship, type of violence, severity of violence, and frequency (UNSD, 2013; see Annex 1). They measure the incidence of physical and sexual violence in intimate and non-intimate relationships plus female genital cutting and psychological and economic violence but leave out state violence by police and the military, harassment and threats, stalking, female foeticide, violence against men, isolation, intimidation, stalking and many other forms of violence. The complexity of relationships, their histories, the interplay of love and fear, the role of kinship and residence patterns, disappear. Instead, a series of relatively discrete and apparently objective categories are used to measure this phenomenon across class, national, and cultural divides.

25This approach to measurement is far more influenced by government concerns and interests than the other three. The UNSC needs governments to adopt and use these guidelines. Its narrow definition of violence against women that focuses on interpersonal relationships avoids national concerns about measuring “sensitive” and “cultural” issues. In contrast to the other initiatives, these indicators are the product of governments working together in the UNSC. That also means that they are reduced to the least common denominator on which most can agree. They also respond to statisticians’ concerns about focusing on phenomena that are measurable, such as the nature of the violence and the relationship with the perpetrator, rather than those that are not, such as fear or stalking. The UNSD secretariat sees its role as supporting government efforts by providing technical assistance for whatever governments choose to do.

26Because of this state involvement, it seems more likely that the UNSC survey will take place than that the surveys proposed by other organizations will. In interviews with staff from the UNSD, I was told that governments are eager to find ways of measuring violence against women in response to considerable international pressure to provide this data. The cost of surveys falls on governments, while the UNSD provides expertise.

2. Conclusion

27Clearly, this massive effort to quantify violence against women has brought greater visibility to a problem long hidden and unspoken. Counting is important to raise awareness and to discourage the idea that violence is only a problem for the unruly few. These efforts at quantification clearly help to show the extent of the problem, whatever their interpretive frameworks.

28However, the nature of the problem itself and what is to be counted differs among these frameworks. Although the four approaches claim to be measuring the same thing, they are actually using different categories and counting different things. Some features of violence against women are foregrounded in each approach while others are ignored. The gender equality and human rights frameworks, which take a broader definition of violence against women, are likely to produce higher numbers of victims than the criminal justice or national statistics approaches. They include a wider range of situations and activities and are more concerned with maximizing disclosure of experiences of violence. The criminal justice framework focuses more narrowly on forms of violence defined as crimes for which there should be a police response, primarily domestic violence and rape. It does not count the wider array of forms of violence incorporated in the human rights model or the gender equality framework, such as police violence, sexual harassment in the workplace, sex trafficking, sexual slavery in armed conflict, or fear of walking at night, to list a few. Social tolerance for violence against women, including in the family, is not part of this system of enumeration. The fourth approach, focusing on building statistical capacity, emphasizes acts of violence that are measurable rather than the experience of the victim. Fear is hard to measure and not counted. It also avoids issues that governments might resist. Honor killings, religious prostitution, early marriage, state-sponsored sterilization, and an array of other practices of violence not defined as criminal in a particular society or viewed as too “sensitive” will not be counted.

29Thus, the choice of the framework used to measure violence against women has a major effect on how it is understood and governed. The gender equality approach shows that violence against women is the product of larger patterns of inequality and discrimination which need to be addressed. The human rights approach targets states for failure to protect women who are victims of a wide range of forms of violence, including state and police violence and advocates diminishing discrimination and promoting equality. The criminal justice approach focuses on flaws in the provision of criminal justice interventions. The national statistics model seeks to describe only those forms of violence that are discrete acts between two people. The larger social context of these acts is not considered.

30Clearly, each of these frameworks requires different strategies to tackle the problem. Each approach to defining and measuring the problem directs attention to a particular solution. Of course, the creators of each framework already have a predetermined set of solutions at hand. The solutions available to the institutional sponsors determine how the problem is conceived and measured in the first place. Whether the problem is seen as one of male/female relationships in the private sphere, as the statistical approach does, or as a problem of state indifference and social tolerance of violence against women, as the gender equality and human rights approaches do, depends on whether the data collection and analysis are promoted and carried out by national statistical offices, criminal justice agencies, human rights commissions, or feminist NGOs. These differences in cultural frameworks and research design determine whether the problem appears widespread and severe or not. They also influence policy decisions about how to handle the problem, so that each organization’s measurement scheme ends up reinforcing the wisdom of its own solution to the problem.

31Which of these approaches will come to dominate the field of international violence against women? This is hard to predict, but these approaches vary significantly in their institutional and financial support. The gender equality and human rights constituencies lack the funding for major survey research. Moreover, their broad categories and complex measures are not readily amenable to quantification. Both the criminal justice and national statistics approaches have more resources. Criminal victimization surveys often have international support in developing countries and in wealthier countries are domestically financed. Countries can produce administrative data or carry out their own surveys. Thus, it seems likely that the criminal justice and national statistics approaches will prevail in defining violence against women, while the gender equality and human rights ones will not. It is noteworthy that the definition of violence against women used in the SDGs, as discussed above, comes from the statistical approach since it uses the indicators developed by the UNSD.

32What does this mean for global governance? Violence against women will be defined narrowly, focusing almost exclusively on physical and sexual violence within intimate and stranger relationships. Fewer victims will be counted. Structural violence, state violence, and social attitudes that tolerate or even support violence as well as state responsibility and the impact of social attitudes toward violence will not be measured. In sum, this example shows that there are significant differences in the cultural frameworks used to measure social phenomena, and that inequalities in the power of those who measure social phenomena have enormous consequences for the way those phenomena are understood and governed. Differential political and institutional support for various cultural frameworks for measurement is a key dimension of the contemporary processes of global governance.

Haut de page


Erturk, Y., 2008a. Promotion and Protection of all Human Rights, Civil, Political, Economic, Social and Cultural, including the Right to Development: Report of the Special Rapporteur on Violence against Women, its Causes and Consequences, Yakin Erturk: Indicators on Violence against Women and State Response. Human Rights Council, seventh session. UN Docs A/HRC/7/6, 29 January 2008.

Erturk, Y., 2008b. Promotion and Protection of all Human Rights, Civil, Political, Economic, Social and Cultural, including the Right to Development: Report of the Special Rapporteur on Violence against Women, its Causes and Consequences, The Next Step: Developing Transnational Indicators on Violence against Women. UN Docs A/HRC/7/6/Add.5, 25 February 2008.

Jerven, M., 2013. Poor Numbers: How We Are Misled by African Development Statistics and What to Do about It. Cornell University Press, Ithaca, NY.

Johnson, H., Ollus, N., Nevala, S., 2008. Violence Against Women: An International Perspective. Springer Science, New York, NY.

Malby, S., Alvazzi del Frate, A., 2007. Indicators, Crime, and Violence against Women: Supporting Paper 3. Paper submitted by UNODC to the Expert Group Meeting on Indicators to Measure Violence against Women, Co-organized by UNDAW, UNECE, and UNSD, Geneva, Switzerland, 8-10 October 2007.

Merry, S.E., 2009. Gender Violence: A Cultural Introduction. Blackwell, London.

Merry, S.E., 2016. The Seductions of Quantification: Measuring Human Rights, Gender Violence, and Sex Trafficking. University of Chicago Press, Chicago.

Merry, S.E., Davis, K., Kingsbury, B. (Eds.), 2015. The Quiet Power of Indicators: Measuring Governance, Corruption, and Rule of Law. Cambridge University Press, Cambridge.

Merry, S.E., Wood, S.J., 2015. Quantification and the Paradox of Measurement: Translating Child Rights in Tanzania. Current Anthropology 56 (2), 205–229.

Rottenburg, R., Merry, S.E., Park, S.-J., Mugler, J. (Eds.), 2015. A World of Indicators: The Making of Governmental Knowledge through Quantification. Cambridge University Press, Cambridge.

UNSD, 2013. Guidelines for Producing Statistics on Violence against Women: Statistical Surveys. United Nations Statistics Division, September 9. ST/ESA/STAT/SER.F/110. Online: (accessed 8 January 2014).

Haut de page


The list of core indicators

i. Total and age specific rate of women subjected to physical violence in the last 12 months by severity of violence, relationship to the perpetrator and frequency;

ii. Total and age specific rate of women subjected to physical violence during lifetime by severity of violence, relationship to the perpetrator and frequency;

iii. Total and age specific rate of women subjected to sexual violence in the last 12 months by severity of violence, relationship to the perpetrator and frequency;

iv. Total and age specific rate of women subjected to sexual violence during lifetime by severity of violence, relationship to the perpetrator and frequency;

v. Total and age specific rate of ever-partnered women subjected to sexual and/or physical violence by current or former intimate partner in the last 12 months by frequency;

vi. Total and age specific rate of ever-partnered women subjected to sexual and/or physical violence by current or former intimate partner during lifetime by frequency;

vii. Total and age specific rate of women subjected to psychological violence in the past 12 months by the intimate partner;

viii. Total and age specific rate of women subjected to economic violence in the past 12 months by the intimate partner;

ix. Total and age specific rate of women subjected to female genital mutilation.

Haut de page

Pour citer cet article

Référence papier

Sally Engle Merry, « Cultural Dimensions of Power/Knowledge: The Challenges of Measuring Violence against Women », Sociologie du travail, Vol. 58 - n° 4 | 2016, 370-380.

Référence électronique

Sally Engle Merry, « Cultural Dimensions of Power/Knowledge: The Challenges of Measuring Violence against Women », Sociologie du travail [En ligne], Vol. 58 - n° 4 | Octobre-Décembre 2016, mis en ligne le 01 décembre 2016, consulté le 05 décembre 2020. URL : ; DOI :

Haut de page


Sally Engle Merry

Department of Anthropology, New York University, Rufus D. Smith Hall, 25, Waverly Place, New York, NY 10003, United States

Haut de page
  • Logo Association pour le développement de la sociologie du travail
  • Logo CNRS
  • Logo La revue Sociologie du travail mobilisée
  • OpenEdition Journals
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search