WPS3618 Preventing and responding to gender-based violence in middle and low- income countries: a global review and analysis Sarah Bott bott.fahey@gte.net Andrew Morrison amorrison1@worldbank.org Mary Ellsberg mellsberg@path-dc.org World Bank Policy Research Working Paper 3618, June 2005 The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries they represent. Policy Research Working Papers are available online at http://econ.worldbank.org. Preventing and responding to gender-based violence in middle and low- income countries: a multi-sectoral literature review and analysis Table of contents: EXECUTIVE SUMMARY ..........................................................................................................................3 CHAPTER I: DEFINITIONS, CONSEQUENCES, COSTS, AND RISK FACTORS ..................................................8 DEFINITIONS AND RATIONALE ........................................................................................................8 MAGNITUDE AND DYNAMICS ..........................................................................................................9 HEALTH CONSEQUENCES OF GENDER-BASED VIOLENCE......................................................12 COSTS OF GENDER-BASED VIOLENCE .........................................................................................12 RISK FACTORS: THE ECOLOGICAL MODEL.................................................................................14 CHAPTER II: INITIATIVES TO PREVENT AND RESPOND TO GENDER-BASED VIOLENCE ........................16 INTRODUCTION ..................................................................................................................................16 JUSTICE.................................................................................................................................................17 HEALTH.................................................................................................................................................25 EDUCATION .........................................................................................................................................31 MULTI-SECTORAL: ECONOMIC DEVELOPMENT, SOCIAL SERVICES AND PUBLIC SAFETY ..................................................................................................................................35 CONCLUSIONS AND RECOMMENDATIONS .....................................................................................43 REFERENCES ...........................................................................................................................................47 ANNEX A...................................................................................................................................................60 ii EXECUTIVE SUMMARY Introduction In 1993, the United Nations General Assembly defined violence against women as "any act of gender- based violence that results in, or is likely to result in, physical, sexual or psychological harm or suffering to women" (United Nations, 1993). The United Nations referred to "gender-based" violence to acknowledge that such violence is rooted in gender inequality and is often tolerated and condoned by laws, institutions and community norms; it is not only a manifestation of gender-inequality, but often serves to enforce it (Heise, Ellsberg and Gottemoeller, 1999). Gender-based violence comes in many forms throughout the life cycle; this review focuses on the two most common types: physical intimate partner violence and sexual violence against women by any perpetrator. A growing body of international research documents the magnitude and patterns of gender-based violence. Worldwide, men experience higher levels of physical violence than women as a result of war, gang-related activity, street violence, and suicide, while women and girls are more likely to be assaulted or killed by someone they know, such as an intimate partner (Heise and Garcia Moreno, 2002). Population-based surveys have found that between 10-70% of women report being physically assaulted by an intimate male partner at some point in their lives (Heise, Ellsberg and Gottemoeller, 1999). The bonds of family, economic and emotional dependence that accompany intimate partner violence make prevention and protection particularly complex. In addition, a substantial proportion of girls and women experience child sexual abuse, rape and other forms of sexual coercion in virtually every setting of the world. Population-based studies report rates of forced sexual debut among sexually experienced young people ranging from 7% in New Zealand to 46% in the Caribbean (Heise and Garcia Moreno, 2002). Contrary to popular belief, most victims know the perpetrators, who often include intimate partners, stepfathers, fathers, other relatives, authority figures, teachers, and acquaintances (Heise, Ellsberg and Gottemoeller, 1999). Sexual violence by strangers is common in select settings, such as situations of armed conflict and displacement, for example, when rape is used as a weapon of war (Ward, 2002). This paper reviews what is known about more and less effective-or at least promising-approaches to prevent and respond to gender-based violence. Within the sections titled Justice, Health, Education, and Multi-Sectoral Approaches, the paper examines initiatives that have addressed laws and policies, institutional reforms, community mobilization, and individual behavior change strategies. The review also highlights cross-cutting lessons that have emerged from research and programs over the last 30 years. Unfortunately, the knowledge base about effective initiatives is relatively limited. Few approaches have been rigorously evaluated, even in high-income countries. For example, Chalk and King (1998) found that among several hundred relevant intervention studies, only 34 were deemed methodologically sound. Indeed, evaluating initiatives to prevent or respond to violence against women involves numerous methodological challenges. For example, prevention often appears to require multiple organizations, strategies and sectors - which makes it difficult to determine which specific strategies can be credited with any changes measured (Chalk and King, 1998). Measuring and interpreting changes in prevalence rates have also proven to be complex tasks (Campbell, 2000). As a result, evidence about effectiveness is often lacking, and the most that can be said about certain approaches is that they appear more or less promising. 3 Justice sector initiatives In many low and middle-income countries, penal and civil law codes fail to criminalize certain kinds of physical or sexual violence against women and include provisions that make convictions unlikely. Over the last 30 years, international human rights agreements combined with advocacy by women's groups have successfully convinced many governments to revise penal and civil legislation in regards to gender- based violence. These changes include criminalizing domestic violence and marital rape, eliminating provisions that allow perpetrators of rape to escape criminal sanctions by agreeing to marry the victims, and revising criminal procedures to make it easier to prosecute offenders. These reforms represent a significant symbolic achievement in the effort to strengthen women's rights and reduce violence against women. However, the overwhelming lesson from research on legislative reform in low and middle-income countries is that legislative reform is just the first step in a long and complex process. In many settings, law enforcement institutions are under-funded, inaccessible, incompetent or even corrupt, making it impossible for them to enforce criminal law more generally. Governments often fail to budget resources for implementing changes in law and policies. Police and judges are often unwilling or unable to enforce laws related to gender-based violence. And, in many settings, girls and women remain unaware of the law or face social and economic barriers that make it impossible for them to exercise their rights. To make the laws work more effectively, a number of initiatives show promise, including: · educating law enforcement and the public about new laws; · broad investment in strengthening the law enforcement response to gender-based violence; · reorganizing the police and the judiciary (e.g. special police cells, family courts, etc.); · comprehensive medico-legal system reform (e.g. introduction of forensic nursing); · building networks and alliances between legal, social and health organizations; and · reform of informal justice systems (e.g. traditional courts and councils). Meanwhile, some initiatives have produced negative unintended consequences, including policies and programs that require women seeking divorce to attempt reconciliation with abusive spouses or that create separate police stations for crimes against women without concurrently pursuing broader law enforcement reforms. Health sector initiatives Historically, the health sector was slow to recognize the public health implications of violence against women. For example, reproductive health programs for young people have often assumed that sexual activity was voluntary (Mensch, Bruce and Green, 1998), despite evidence that many young women experience sexual coercion (Jejeebhoy and Bott, 2003). Many universities do not prepare health professionals to recognize the health consequences of domestic violence, rape or sexual abuse; providers often view violence against women as a social issue rather than a health problem; and many organizations do not equip their staff to respond appropriately to girls and women who disclose that they have experienced violence (Heise, Ellsberg and Gottemoeller, 1999). In recent years, health care organizations around the world have tried to improve the care they provide to survivors of gender-based violence. In addition, public health programs, which have a long history of working to change sexual attitudes, practices, and behaviors, have begun applying those strategies to gender-based violence. Few have demonstrated an impact on levels of violence or measured quantitative outcomes among survivors, however, though a number of approaches appear promising, including: · national, regional and municipal policies that facilitate women's access to emergency contraception, 4 high quality forensic exams, prophylaxis for sexually transmitted infections / HIV and safe abortion; · government policies that clarify providers' roles and responsibilities regarding gender-based violence; · broad institutional reforms to improve the health care response to gender-based violence (e.g. protocols and policies, providers training, written resources, and alliances with referrals services); · efforts to integrate the issue of gender-based violence into the training of health care professionals; · networks and coalitions devoted to referrals, research, advocacy and education; · efforts to raise awareness of and reduce violence against women as a public health problem; · integrating attention to gender-based violence into reproductive heath / HIV education for youth; and · mass media `entertainment-education' programs, including prime time television soap operas that address domestic and sexual/or violence. Education sector initiatives Evidence suggests that high levels of education can reduce women's vulnerability to gender-based violence, though the link between education and violence is not necessarily linear (Jewkes, 2002c). In theory, schools could play a proactive role in prevention by promoting greater respect for women's human rights. Indeed, fueled largely by concern about the HIV/AIDS pandemic, schools in most regions of the world now offer some kind of reproductive health education, and many of these programs address gender issues, including sexual coercion (Birdthistle and Vince-Whitman, 1998). Meanwhile, researchers and advocates have called on universities to prepare the next generation of legal, medical, and social work professionals to address gender-based violence in their respective fields. Unfortunately, evidence from many middle and low-income countries suggests that schools and universities have a long way to go before they can play a positive role in preventing violence against women. For one thing, sexual harassment by educators and students appears to be widespread in many parts of the world (Mirsky, 2003; Wellesley Centers for Research on Women and DTS, 2003). Schools and universities cannot be positive agents for change as long as the school environment tolerates or condones discrimination and violence against girls. Moreover, the lack of school safety appears to reduce the enrolment of girls relative to boys in some settings. For example, parents' fears for their daughters' physical and sexual safety appears to be a major reason for withholding girls from school in South Asia and Sub-Saharan Africa (Mensch and Lloyd, 1998; Sathar and Lloyd, 1993; UNICEF, 2004). Fueled by evidence of sexual harassment in schools as well as of the role that sexual coercion plays in the spread of the HIV/AIDS pandemic, many governments, schools and universities have increasingly begun to address gender-based violence through policies, awareness campaigns and curriculum changes. Little is known about the impact of these efforts, but the following approaches seem to hold promise: · institutional reforms to reduce sexual harassment in schools, e.g. by educating staff about gender, human rights and nonviolence and by developing sexual harassment policies that include clear reporting mechanisms and sanctions for staff who violate such policies; · improving school infrastructure (e.g. building schools to reduce the distance that girls travel to school, providing safe latrines for girls, hiring more female teachers, and establishing single sex schools for girls); · school-based counseling and referrals; and · school-based programs for students that promote nonviolence, human rights and more equitable gender roles. 5 Multi-sectoral initiatives Multi-sectoral collaboration is important for most gender-based violence initiatives, regardless of sector, but those that aim to improve women's lives through social services, economic empowerment and infrastructure improvements require a multi-sectoral approach--almost by definition. Compelling evidence suggests that-in the long run-economic and social empowerment of women may reduce women's vulnerability to gender-based violence. This category includes a highly diverse group of approaches, objectives, and lessons learned. Similar to education, however, the relationship between empowerment and violence is not necessarily linear; some "successful" efforts to empower women may increase the risk of violence in the short run, by challenging traditional gender roles and increasing conflict in the household. Some key approaches have shown long-term promise however, including: · efforts to strengthen women's rights to property, inheritance, labor force participation, divorce, etc.; · expanding social services for women and children (e.g. counseling, legal aid, shelters, etc.) through public-private partnerships that include government ministries and non-governmental organizations; · integrating the issue of gender-based violence-particularly sexual violence-into policies and services that serve refugees and displaced populations; · micro-credit programs for women that explicitly address the implications of gender-based violence; · attention to women's needs and priorities within transport and infrastructure projects; and · community-based and mass media campaigns to reduce tolerance for violence against women. Conclusions and recommendations Bilateral donors and multilateral institutions can play an important role in addressing gender-based violence in developing countries by funding research on the health and socioeconomic costs of gender- based violence, by encouraging science-based evaluations of gender-based violence programs, by disseminating evaluation results across countries, by promoting investment in effective prevention and treatment initiatives, and by encouraging public-private partnerships. Cross-cutting lessons learned about the best way to prevent and respond to gender-based violence include the following: · Employ a multi-sectoral approach. One consistent finding from all sectors is the need for collaboration between law enforcement, legal aid, health care organizations, public health programs, educational institutions and agencies devoted to social services and economic development-for the purposes of both prevention and ensuring an integrated response to survivors. · Work at different levels. Effective approaches generally require working at different levels (individual, community, institutional, and laws/policies). For example, simply changing the penal code may be ineffective if law enforcement institutions remain weak, if communities resist changes in women's legal rights, and if women remain unaware of the laws or unable to access services. · Create partnerships between government and nongovernmental agencies. This review highlights examples of benefits from collaboration between government and civil society. Both groups have a role to play and are unlikely to change levels of violence working alone. · Address norms, attitudes and beliefs at all levels of society. Attitudes that condone or tolerate violence against women and blame the victim are deeply entrenched throughout society in nearly all parts of the world-to varying degrees. Changing these attitudes and beliefs is a challenging, long- term process that requires a sustained effort and a commitment to a human rights perspective. · Target young people. Young people appear more open to changing their views about the 6 acceptability of violence than older adults. Thus, youth-oriented education programs represent an important strategy for preventing violence in the long run. Similarly, universities have an opportunity to prepare the next generation of professionals to respond adequately to gender-based violence. · Demonstrate the developmental impact of gender-based violence. Rigorous research documenting the public health and socio-economic consequences of gender-based violence is an important tool for promoting policy change and increased investment in prevention and response. · Build the knowledge base through rigorous evaluation. The dearth of evidence about effective ways to address gender-based violence limits policy makers' and program managers' ability to make informed decisions, and makes it difficult to argue for allocating increased resources for prevention and response. More rigorous evaluations are needed-particularly in the area of prevention. The best hope for reducing worldwide levels of violence against women may lie in mobilizing all levels of society-from international donors and national governments, to grassroots women's organizations, private firms and local governments. The challenge is not only to raise awareness of violence against women, but to maintain a long-run commitment by all these actors to address gender-based violence as an impediment to economic development, a public health problem, and a violation of human rights. 7 CHAPTER I: DEFINITIONS, CONSEQUENCES, COSTS, AND RISK FACTORS DEFINITIONS AND RATIONALE Definition of "gender-based" violence In 1993, the United Nations General Assembly defined violence against women as follows: "Violence against women" means 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. - 1993 UN Declaration on the Elimination of Violence Against Women The General Assembly went on to present a partial list of what it felt constituted gender-based violence, including: ·physical, sexual and psychological violence within the family ·child sexual abuse ·dowry-related violence ·marital rape ·female genital mutilation ·rape and sexual abuse ·sexual harassment in the workplace and educational institutions ·trafficking in women ·forced prostitution. The 1993 United Nations Declaration referred to "gender-based" to highlight the links between violence against women and women's subordinate status. International research provides compelling evidence that violence against women is rooted in gender inequities and is both tolerated and sometimes even condoned by laws, institutions and community norms that discriminate against women and girls (Heise, Ellsberg and Gottemoeller, 1999). Gender-based violence is thus not only a manifestation of gender inequality, but often serves to enforce it. Men often use violence to punish perceived transgressions of gender roles, to show authority, and to save honor. Violence against women is often considered normal and justified by the broader society rather than a criminal act, and victims instead of perpetrators are often blamed and stigmatized. Violence against women, therefore, cannot be understood in isolation from the gender norms, social structures and roles that influence women's vulnerability to violence. While men experience higher levels of overall violence than women as a result of war, gang-related activity, street violence, and suicide, women and girls are more likely to be assaulted or killed by someone they know, often by a family member or an intimate partner. For example, studies from Australia, Canada, Israel, South Africa and the United States have found that 40-70% of female murder victims are killed by their husbands or boyfriends, compared to only 4-9% of men; small-scale studies from developing countries report similar findings (Heise and Garcia Moreno, 2002). The bonds of family, economic and emotional dependence that accompany intimate partner violence make prevention and protection particularly complex. Women are also much more likely than men to be sexually assaulted as children, adolescents or adults, and the vast majority of perpetrators of sexual violence are male, as are virtually all perpetrators of rape (Heise, Ellsberg and Gottemoeller, 1999). Contrary to popular belief, most women know their attackers; studies typically report that between 60% and 80% of perpetrators of sexual assault are known to the victim (again see Heise, Ellsberg and Gottemoeller, 1999). In addition to intimate partners, perpetrators often include stepfathers, fathers, other relatives, authority figures, and acquaintances. Women and girls experience high levels of sexual violence by strangers in selected settings, notably situations of armed 8 conflict and displacement; for example, rape is often used as a weapon of war, and refugee settings often involve a high degree of violence and insecurity (Ward, 2002). To limit the scope of this review, this paper will focus on the two most common forms of gender-based violence: intimate partner violence (physical and sexual) and sexual violence by any perpetrator. As a starting point, the World Report on Violence and Health (Krug et al., 2002) offered the following definitions of intimate partner violence and sexual violence, respectively: Intimate partner violence: Any behavior within an intimate relationship that causes physical, psychological or sexual harm . . . Such behavior includes: ·Acts of physical aggression-such as slapping, hitting, kicking and beating ·Psychological abuse-such as intimidation, constant belittling and humiliating ·Forced intercourse and other forms of sexual coercion · Various controlling behaviors-such as isolating a person from their family and friends, monitoring their movement and restricting their access to information or assistance (Heise and Garcia Moreno, 2002, page 89). Sexual violence: Any sexual act, attempt to obtain a sexual act, unwanted sexual comments or advances . . . using coercion, by any person regardless of their relationship to the victim, in any setting, including but not limited to home and work...Sexual violence includes rape, defined as physically forced or otherwise coerced penetration...of the vulva or anus, using a penis or other body parts or an object. Sexual violence can include other forms of assault involving a sexual organ, including coerced contact between the mouth and penis, vulva or anus (Jewkes, Sen and Garcia Moreno, 2002, page 149). MAGNITUDE AND DYNAMICS Measuring levels of violence against women poses many challenges. Estimates vary depending upon how researchers define violence, the questions they ask, the timeframes they explore, and the sample characteristics. Heterogeneous approaches often make it difficult to compare research findings. Another challenge is that surveys are best at measuring discrete incidents of physical violence; rather than the pattern of controlling, violent behavior often referred to as "battering", which is typically characterized by multiple forms of abuse--physical, emotional, sexual, etc. (Ellsberg et al., 2001a). Such violence often escalates with time and involves a high level of fear among women for their safety and lives. Sexual violence poses its own definition and measurement challenges. For one thing, sexual violence encompasses a large range of manifestations (from verbal harassment, to unwanted touch, molestation, assault and penetration), settings, and perpetrators. Jewkes (2002) developed a framework for understanding different types of sexual violence against women, ranging from the most visible, namely fatal sexual assault and rape reported to the police, to the most common form and least visible forms of sexual violence that occur within marriage, dating relationships and families. Theoretical and methodological approaches to research on sexual abuse/violence are even more diverse than those focused on intimate partner violence. Typically, researchers use different study designs to measure levels and patterns of child sexual abuse, forced sexual initiation during adolescence, rape among the general population of women, forced sex within marriage, and sexual harassment / abuse within schools. A serious challenge for all types of research on the prevalence of gender-based violence is that most estimates probably reflect a certain degree of under-reporting. Women are often reluctant to disclose experiences of physical or sexual violence due to shame or fear of reprisals (Koss, 1993). Underreporting appears to be particularly problematic in large-scale surveys designed primarily for other purposes, such 9 as the Demographic and Health Surveys (DHS), though researchers have reduced under-reporting by providing special training to interviewers, placing greater emphasis on respondents' privacy and safety, and allowing women multiple opportunities to disclose their experiences (Ellsberg et al., 2001a; Garcia Moreno et al., 2003; Ellsberg et al., forthcoming). Estimates of the magnitude of the problem Population-based surveys have found that between 10-70% of women report being physically assaulted by an intimate male partner at some point in their lives (Heise, Ellsberg and Gottemoeller, 1999). See Annex A for estimates from many recent population based studies (Ellsberg et al., forthcoming). Findings from a multi-country study on domestic violence and women's health carried out by the World Health Organization in fifteen sites and ten countries found that between 13-62% of women had experienced physical violence by a partner over the course of their lifetime, and between 3-29% of women reported violence within the past year (Figure 1.1). Figure 1.1. Percentage of women who reported sexual violence by an intimate partner (ever), physical violence by an intimate partner (ever), and physical violence by an intimate partner in the past 12 months. 37% Bangladesh (Urban) 19% 40% Bangladesh (Province) 50% 42% 10% 16% Brazil (Urban) 27% 8% 14% Brazil (Province) 34% 13% Ethiopia (Province) 59% 49% 6% 29% Japan (Urban) 13% 3% 17% Namibia (Urban) 31% 16% 23% Peru (Urban) 50% 17% 47% Peru (Province) 62% 25% 30% Thailand (Urban) 8% 23% 29% Thailand (Province) 34% 13% 23% Tanzania (Urban) 33% 15% 31% Tanzania (Province) 47% 19% 6% Serbia 23% sexual violence ever 3% 20% physical violence ever Samoa 41% 18% physical violence past 12 months Source: Unpublished data from the WHO Multi-Country Study on Women's Health and Domestic Violence Against Women. The final published comparative report is forthcoming. Cited with permission. Prevalence data on sexual violence is even more limited than physical violence. However, evidence suggests that a substantial proportion of girls and women have experienced child sexual abuse, forced sex and other forms of sexual coercion in virtually every setting of the world. For example, population-based studies have asked about "forced" sexual debut among sexually experienced young people and found rates from 7% (New Zealand), to 46% (in the Caribbean) (Heise and Garcia Moreno, 2002). 10 Figure 1.2. Female adolescents reporting forced sexual initiation, as a percent of those reporting having had sex (Population-Based Surveys 1993-1999) New Zealand 7% United States 9% Mozambique 19% Ghana 21% South Africa 28% United Republic of Tanzania 29% Cameroon 37% Peru 40% Caribbean 48% Source: Population-Based Surveys 1993-1999. All figures cited in Jewkes, Sen, and Garcia Moreno, 2002. Sexual violence within marriage is also common, with approximately 10-13 percent of women reporting having been forced by a partner to have sex against their will at some point (Heise, Ellsberg, and Gottemoeller, 1999). Sexual violence often accompanies physical battery by intimate partners, and in many settings, women and men believe that men have the right to beat their partners for refusing sex (Figure 1.3). Women who experience physical violence by intimate partners may be less able to negotiate when and how they have sex. Nonetheless, large variations exist in patterns and prevalence levels between and within countries. For example, a study from León, Nicaragua found that nearly all women who reported sexual violence had also experienced physical violence (Ellsberg et al, 2000). In contrast, research from Indonesia suggests that sexual violence often occurs outside the context of physical violence and may be even more common than physical violence (Hakimi et al., 2001). Figure 1.3. Percentage of men and women who say that men have the right to beat their wives when they refuse to have sex. 81% 57% Men 43% Women 33% 28% 10% 5% Egypt (rural) Egypt (urban) Ghana West bank / Gaza Nicaragua (rural) Singapore strip Source: Various population-based surveys, cited in Heise, Ellsberg, and Gottemoeller, 1999. 11 HEALTH CONSEQUENCES OF GENDER-BASED VIOLENCE A growing body of epidemiological evidence documents the consequences of gender-based violence for women's health and wellbeing, ranging from fatal outcomes, such as homicide, suicide and AIDS-related deaths to non-fatal outcomes such as physical injuries, chronic pain syndromes and gastrointestinal disorders (Heise, Ellsberg, Gottemoeller, 1999). Physical and sexual violence has consequences for women's mental health, such as post-traumatic stress syndrome, depression, anxiety, and low self-esteem, as well as behavioral outcomes such as alcohol and drug abuse, sexual risk-taking and a higher risk of subsequent victimization. Table 1.1. Health consequences of intimate partner violence and sexual violence by any perpetrator Fatal outcomes Nonfatal outcomes Physical injuries and chronic Sexual and reproductive Psychological and behavioral conditions sequelae outcomes Femicide Fractures Gynecological disorders Depression and anxiety Suicide Abdominal/thoracic injuries Pelvic Inflammatory disease Eating and sleep disorders AIDS-related mortality Chronic pain syndromes Sexually-transmitted Drug and alcohol abuse Maternal mortality Fibromyalgia infections, including HIV Phobias and panel disorder Permanent disability Unwanted pregnancy Poor self-esteem Gastrointestinal disorders Pregnancy complications Post-traumatic stress disorder Irritable bowel syndrome Miscarriage / low birth weight Psychosomatic disorders Lacerations and abrasions Sexual dysfunction Self harm Ocular damage Unsafe abortion Unsafe sexual behavior Sources: Adapted from Heise and Garcia Moreno, 2002; and Heise, Ellsberg and Gottemoeller, 1999. Intimate partner violence and sexual coercion have particularly serious consequences for women's sexual and reproductive health. Sexual violence may lead to gynecological problems, unwanted pregnancy, chronic pelvic pain, unsafe abortion, and sexual dysfunction. Intimate partner violence has been linked to similar outcomes, as well as complications during pregnancy, miscarriage and low birth-weight (Campbell, 2002). Sexual abuse in childhood and adolescence has been linked to a higher risk of subsequent victimization, early sexual activity, substance abuse, and multiple sexual partners (Felitti et al., 1998; Heise, Ellsberg, and Gottemoeller, 1999; Walker et al., 1999). Some evidence suggests that women who suffer violence are less able to negotiate family planning or condom use (Garcia-Moreno, 2002); consequently, victims of gender-based violence may experience higher rates of unintended pregnancies (Gazmararian et al., 1995) as well as increased vulnerability to sexually-transmitted infections, including HIV/AIDS. Health consequences for children of women who experience gender-based violence Researchers have also documented negative outcomes among children of women who experience violence. For example, researchers in Nicaragua found that children of women who were physically and sexually abused by their partners were six times more likely than other children to die before the age of five, with one third of all child deaths in this setting being attributed to partner violence (Åsling-Monemi et al., 2003). Some evidence suggests that children who witness violence in the home may be at increased risk for emotional and behavioral problems, such as anxiety, depression and violence towards their peers (Jaffe and Sudermann, 1995); and for perpetrating intimate partner violence and/or sexual violence as adults (Straus, Gelles and Smith, 1990; Kishor and Johnson, 2004). COSTS OF GENDER-BASED VIOLENCE Gender-based violence poses significant costs for the economies of developing countries, including lower worker productivity and incomes, lower rates of accumulation of human and social capital, and the generation of other forms of violence both now and in the future. Cost estimates for all the different types 12 of gender-based violence are beyond the scope of this paper, but this section will present a summary classification of the types of costs and the methodologies that can be used to measure them. Regardless of the relative socioeconomic costs of gender-based violence, gender-based based violence remains a women's health issue and a violation of human rights, but cost estimates can facilitate the "dimensioning" of the issue: how does gender-based violence compare to the panoply of pressing development issues? What are the impacts of gender-based violence on a country's socioeconomic development? The most common approach used to calculate the costs of gender-based violence has been an "accounting methodology", in which costs are calculated for specific categories, and total cost to society is simply the sum of all distinct categories of costs. Typical of this approach is CDC (2003), which specifies two types of costs: · Direct costs are actual expenditures related to gender-based violence, including health care services, judicial services and social services. · Indirect costs represent the value of lost productivity from both paid work and unpaid work, as well as the foregone value of lifetime earnings for women who have died. Recent estimates of the health care costs of intimate partner violence against adult women in the United States found costs of over 4 billion dollars in 1995, including both mental health and medical care costs (CDC, 2003). Similar methodologies have been employed to generate estimates for Canada (Greaves et al., 1995), British Columbia (Kerr and McLean, 1996), Holland (Korf et al., 1997), Northern Territory (Office of Women´s Policy, 1996), Queensland (Blumel et al., 1993), Switzerland (Godenzi and Yodanis, 1998) and the U.K. (Stanko et al., 1998). To the best of our knowledge, there are only two such direct cost studies for gender-based violence in developing countries. Mansingh and Ramphal (1993) estimated that the direct costs of treating victims of intimate partner violence in Kingston Public Hospital (Jamaica) totaled U.S. $454,000 in 1991 (in 2001 dollars). Sanchez et. al. (2004) found that the Colombian national government spent approximately 184 billion pesos (U.S. $73.7 million) in 2003 to prevent, detect and offer services to survivors of family violence-an amount equal to approximately 0.6% of the total national budget. Direct cost estimates are problematic in developing country contexts, where the lack of services (or serious under-funding of services) generally means that direct cost estimates associated with gender-based violence will be low, giving a mistaken impression that that the problem is not important. In such settings, indirect cost estimates may be more useful than direct cost estimates. Indirect cost estimates have typically focused on foregone earnings due to death and lost productivity (CDC, 2003), job loss, lost productivity of women victims, lost productivity of abusers due to incarceration, and mortality (Laurence and Spalter-Roth, 1995); loss of tax revenues due to death and incarceration (Greaves et al., 1995), and decreases in women's earnings (Morrison and Orlando, 1999; Sánchez et al., 2004). For example, Morrison and Orlando (1999) estimated that lost wages due to family violence amounted to 1.6 and 2.0% of GDP in Nicaragua and Chile, respectively, by estimating earnings equations of the determinants of women's earnings. Using a non-parametric matching methodology to analyze DHS data from 1995, Sánchez et al. (2004) found that Colombian women who suffered physical violence had 14% lower earnings than women who did not suffer violence. Using more recent data from 2003, they estimated that the wage loss due to family violence was equivalent to 0.85% of GDP in 2003. One weakness of the accounting approach is that any list of categories is essentially arbitrary, and alternative categorizations can always be devised (Buvinic and Morrison, 1999). An even more serious weakness is that key categories of costs can be inadvertently left out of the calculations. For example, the indirect cost category mentioned above focuses on productivity and earnings losses of female survivors, but does not discuss the impact that witnessing or being a victim of family violence may have on children. These impacts may include: poorer performance in school (Larraín et al., 1997); increased probability of delinquency, both as a juvenile and as an adult (Windom, 1989; Dahlberg, 1998; Thornberry et al., 2001); 13 children leaving abusive homes to live on the street (Hernández Rosete, 1998); substance abuse (Molnar et al., 2001); attempted suicide (Dube et al., 2001); a higher probability of committing family violence as an adult (Strauss et al., 1980); and child mortality (Åsling-Monemi et al., 2003). Nor does the indirect cost category necessarily include the erosion of social capital that results from the isolation that survivors of domestic violence often experience. Another option for estimating the socioeconomic costs associated with gender-based violence-and one that is frequently employed by economists to establish the market value of non-market goods-is to estimate the willingness of individuals (and by extension of society) to pay for lives free of such violence. This approach has only been used infrequently to gauge the welfare loss occasioned by gender-based violence (see Sorenson, 2003, for one of the few examples), presumably for two reasons. First, international human rights agreements have recognized that freedom from violence is a fundamental human right; therefore, estimating the willingness to pay for a fundamental right while demonstrating the importance society attaches to an issue, may in itself generate controversy. A second unattractive feature of willingness-to-pay estimates is that they are sensitive to income levels and income distribution. Disability-adjusted life years A third option for estimating the socioeconomic costs of gender-based violence is to use the metric of disability-adjusted life years (DALYs) lost to gender-based violence. DALYs have the advantage of including years lost due to both premature mortality and to disability. DALYs can be used to measure not only death and injuries directly attributable to violence, but also the proportion of other health problems (such as depression, sexually transmitted infections, etc.) to which violence contributes indirectly as a risk factor. DALY calculations are, however, methodologically complex and data intensive. Heise, Pitanguy and Germain (1994) developed the first attempt to estimate DALYs lost due to gender-based violence in 1994. More recently, Lozano (1999) estimated that rape and family violence against women were the third most important cause of DALYs lost in Mexico City-behind diabetes and perinatal conditions, but ahead of auto accidents, congenital anomalies, rheumatoid and osteo-arthritis, cardiovascular disease, stroke and pneumonia. DALYs give less weight to years of lost health and life among women older than 25, discounting the value of health in most of the reproductive years and they use a fairly narrow definition of disability. However, they can be useful for comparing the relative burden of death and disability caused by different public health problems. In sum, there is no perfect methodology with which to gauge the socioeconomic costs of gender-based violence. All methodologies have strengths and weaknesses, and the challenge is to choose the appropriate methodology given both data constraints and the potential consumers of the estimates. RISK FACTORS: THE ECOLOGICAL MODEL Because gender-based violence is a complex phenomenon, shaped by forces that operate at different levels, many researchers have used an ecological framework that combines individual level risk factors with community and society level factors as a way to examine the combination of risk factors that increase the likelihood of violence against women in a particular setting. Although the ecological framework has gained broad acceptance for conceptualizing violence, there have been few attempts to explore how individual and community level risk factors relate to each other and ultimately influence women's vulnerability to violence. One exception was a study in Bangladesh, which found that some aspects of women's status could either increase or decrease a woman's risk of being beaten, depending on the socio-cultural conditions of the community in which she lives (Koenig et al., 2003). That study found that in one site that was characterized by more conservative norms regarding women's roles and status, women with greater personal autonomy and those who participated for a short time in savings and credit groups experienced more violence than women with less autonomy. The 14 opposite was true in the less conservative setting where women had higher overall status. In the less conservative site, individual measures of autonomy and participation in credit schemes had no impact on the risk of violence, rather, living in a community where more women participated in credit groups, and where women had a higher status overall had a protective effect. These findings suggest that the same condition (mobility or participating in a credit group) may have completely different effects on a woman's risk of violence, according to whether the activity is seen as acceptable or not by community norms. These findings underscore the complexity of these issues and the dangers of applying knowledge gained from one site to another without understanding the broader cultural context. Figure 1.4. Risk factors often associated with violence against women: an ecological model Individual level: biological and personal history factors among both victims and perpetrators Relationship level: proximal social relationships, most importantly those between intimate partners and within families. Community level: the community context in which social relationships are embedded, including peer groups, schools, workplaces and neighborhoods. Societal level: larger societal factors that "create an acceptable climate for violence, reduce inhibitions against violence." (Adapted from Krug et al., 2002). Societal Community Relationship Individual · Traditional gender norms that · Weak community sanctions · Marital conflict · A history of violence in the give men economic and against GBV · Family dysfunction perpetrator's or victim's family decision-making power in the · Lack of shelters or other forms of · Male dominance in the family of origin (including intimate household assistance/sanctuary · Economic stress partner violence and child abuse) · Social norms that justify violence · Poverty · Early age at marriage · Male alcohol use against women · Women's traditional gender roles · Large number of children · Male personality disorders · Women's lack of legal rights in transition · Friction over women's (particularly in low prevalence (including access to divorce) · Normative use of violence to empowerment settings) · Lack of criminal sanctions settle all types of disputes · Family honor considered more · Young age (both women and against perpetrators of GBV · Social norms that restrict important than the health and men) (impunity) women's public visibility safety of the victim · High levels of crime · The safety of public spaces · Armed conflict Source: Adapted from Heise and Garcia Moreno, 2002; and Jewkes, Sen and Garcia Moreno, 2002. 15 CHAPTER II: INITIATIVES TO PREVENT AND RESPOND TO GENDER-BASED VIOLENCE INTRODUCTION Over the past 20 years, many initiatives have sought to address violence against women at different levels and within different sectors. Unfortunately, relatively few have been rigorously evaluated. Evidence about effective approaches is particularly sparse in middle and low-income countries, but even in high- income countries, a review by Chalk and King (1998) found that among several hundred relevant intervention studies, only 34 were deemed methodologically sound. In general, evaluations have been characterized by a number of weaknesses such as: · Exclusive reliance on "process" or "output indicators" · Failure to measure (or even specify) the outcomes that the interventions were expected to achieve · Lack of baseline data because evaluations did not begin until after programs were fully implemented · Lack of control groups (or communities) · Short follow-up periods or no follow-up at all · Small sample sizes Moreover, evaluating initiatives to prevent or respond to violence against women involves a host of methodological challenges. First, preventing violence often appears to require multiple organizations, strategies and sectors, making it difficult to determine which specific strategies can be credited with any changes measured (Chalk and King, 1998). Second, defining, measuring and interpreting levels of violence against women have not proven to be straightforward tasks (Campbell, 2000). Third, "successful" programs may appear to increase levels of reported violence. In some cases, this increase reflects a greater willingness to report experiences of violence without necessarily indicating that the underlying prevalence has changed (although such increases can have real consequences for police, service providers, and others). In other cases, programs that "successfully" empower women may increase conflict between men and women and thereby cause the incidence of violence to rise--at least in the short run (Sullivan and Bybee, 1999). Finally, few programs have measured outcomes over long periods of time. As a result, evaluations that measure only short-term effects cannot determine whether attitudinal or behavior changes are sustained in the long run. The structure of this chapter The rest of this chapter reviews what is known about more and less effective ways to prevent and respond to gender-based violence. The review is divided into four different sections organized by sector, namely: Justice, Health, Education, and Multi-sectoral approaches (including social services, economic development, and infrastructure). In keeping with the ecological model, each sectoral chapter is subdivided into four sections, according to the level at which programs have operated, namely: · Law and policies · Institutional reforms · Community level interventions · Individual behavior change strategies This review emphasizes efforts to prevent violence against women rather than initiatives aimed solely at assisting survivors with recovery. Unfortunately, even less is known about how to prevent violence than about how to care for survivors once it has occurred. This review will explore some initiatives that address the needs of survivors, but will not include a comprehensive discussion of survivor services. Finally, as mentioned earlier, this review will focus primarily on the two most common types of gender- 16 based violence: intimate partner violence (physical and sexual) and sexual violence by any perpetrator. As a result, this review will not provide a comprehensive discussion of interventions that address other types of gender-based violence, such as female genital mutilation, dowry deaths, or rape used as a weapon of war. Methods used to compile this review This review draws from many published and unpublished sources, using databases such as POPLINE, PUBMED and CURRENT CONTENTS. Many program evaluations from middle and low-income countries appear only in the grey or unpublished literature, so this review relies heavily on unpublished sources, often cited secondhand. Many national and international organizations have produced reviews of gender-based violence initiatives in recent years. These reviews often focus on a single sector, such as health, education or communication, and they sometimes place a greater emphasis on describing program strategies than on synthesizing what is known about effective prevention. This paper benefited greatly from their work, especially from reviews of unpublished evaluations by USAID (Guedes, 2004; White, Green and Murphy, 2003), the Panos Institute (Mirsky, 2003), the World Health Organization (Krug et al., 2002), and the Inter-American Development Bank (Morrison and Biehl, 1999). JUSTICE Background Legal systems throughout the world have historically encoded discrimination against women into penal and civil law. In many low and middle-income countries, penal codes do not criminalize certain kinds of physical or sexual violence against women; consider sexual violence to be a private offense against "family honor" rather than a criminal offense against the personal integrity of a woman; allow perpetrators of rape to evade criminal responsibility by agreeing to marry their victim; involve criminal procedures that make conviction unlikely; and sometimes punish victims of sexual violence by prosecuting them for adultery or abortion (Mehotra, 1998; Center for Reproductive Law and Policy, 2002). Even when strong legislation exists, law enforcement institutions often fail to enforce the law, and/or inflict additional trauma on survivors of violence through bias and mistreatment (Human Rights Watch, 1997). In many countries, women have limited rights to voluntary marriage, divorce, child custody, and child support, either by law or in practice-legal tools that can allow women to avoid or escape abusive situations and may be essential before they can consider seeking help from law enforcement. Research suggests several ways in which justice sector reform could contribute to gender-based violence prevention, for example by sanctioning those who perpetrate crimes against women; by increasing awareness throughout society that physical or sexual violence against women is a crime; by increasing women's access to the legal system; and by reducing mistreatment of women by law enforcement institutions themselves. Empirical evidence of a causal link between criminal justice and prevention is somewhat weak, however. While Counts, Brown, and Campbell (1992) found that the lowest rates of violence against women occurred in societies that consistently imposed sanctions (including legal sanctions) on perpetrators among 16 societies studied (both developed and developing), some researchers question whether criminal justice approaches can deter violence (Gillegan, 2000). Even in countries such as the United States and Australia that have measured declines in rates of intimate partner and sexual violence over the past decade (Rennison, 2003; Dunne et al., 2003), researchers have not been able to determine whether these declines are due to legal reforms, demographic shifts, social and cultural changes, or other factors-assuming the trends themselves are correctly identified. Criminal justice approaches to violence against women have inherent limitations. Even within strong legal systems, prosecution is often lengthy, complicated, expensive, traumatic for survivors, and is not 17 guaranteed to result in conviction. Criminal justice is typically focused on punishing perpetrators rather than restoring the safety and wellbeing of women. Criminal sanctions are not necessarily appropriate for women who want physical or sexual violence to end, but who do not want to break up their family (Larraín, 1999). In some settings, women and children have no means of economic survival if a violent partner or sexually abusive parent receives jail time (Parenzee, 2001). For all these reasons, legal aid programs often find that women express more interest in assistance with divorce, division of marital property, child custody, and child support-the legal tools that make it possible to leave a violent partner- rather than in criminal prosecution (Guedes et al., 2002). Nonetheless, most researchers consider criminal justice reform to be a crucial component of reducing violence against women (National Research Council, 2004). If nothing else, they argue, failure to sanction offenders sends a message that society condones violence (Larraín, 1999). Overview of justice sector initiatives Table 2.1 presents a selection of justice sector initiatives over the last 20 years. The following section then explores in depth the initiatives and the available evidence about which approaches have been more and less effective. Table 2.1. Examples of objectives and strategies of legal reform initiatives Level Objectives: Examples of Specific Initiatives: Laws and To improve laws and policies related to gender-based · Nationalandinternationaladvocacy campaigns policies violence · Reducediscriminationagainstwomeninthelaw · Ratificationofinternationalhuman rightsagreements · Strengthencriminalsanctionsforperpetrators · Internationalcourts · Improve criminal law procedures · Revisionof thecivilandpenalcode · Strengthenlegaltoolsforprotectingsurvivors · Specificlegislationonfamily,domesticorsexualviolence · Legaltoolssuchasprotectionorders, alternativesentencing Institutional To strengthen the response of key legal institutions · Policies,proceduresandprotocolstoimprovetheresponseof reform (police, judiciary, medico-legal) police, judges, forensic doctors, and other professionals · Improvecapacity to enforcelaws · Investmentinresourcesandequipment · Reducebiasand mistreatment · Sensitizationandtrainingofjusticesystemprofessionals · Increase women's accesstothelegal system · Monitoring mechanisms suchashumanrightsombudsmen · Improvethequalityandcomprehensivenessof · Replacementofforensicdoctorswith forensicnursesorfemale survivor services doctors nominated by women's groups · Improvelegalprotectionforwomenindanger · Women'spolicestationsorcells;FamilyCourts · Increase coordinationwithothersectorsthat · Court-appointedadvocates provide services or work on violence prevention Community To increase community mobilization in defense of · Duluthcoordinatedcommunityresponse model mobilization women's legal rights · Strengthencommunitysupportforwomen'scivil · Legalliteracytrainingforkeygroupsandstakeholders rights and access to justice · NGOlegal aidservices · Strengthennetworksoflegalrightsservices · Effortsto monitor thejustice systematthecommunitylevel · Increase communityactiontobringperpetratorsto · Humanrightspromotersand defensoraspopulares justice · Informal /traditionalarbitration mechanisms Individual To improve individual's knowledge, attitudes and · Mass mediacampaignsonlawsandrights behavior practices among key groups/the broader population changes · Increasepublicawareness/supportforgender- · Legalliteracytrainingforgroupssuch aswomenand youth based violence laws and women's rights · Legalaidreferrals andservices · Increase women's abilitytoexercise rights 18 Reforming laws and policies International rights agreements Over the past 30 years, countries around the world have signed international rights agreements that make specific reference to violence against women (see Box 1). Box 1. Key Human Rights Declarations and Legal Instruments that address gender based violence · Convention on the Elimination of All Forms of Discrimination against Women (CEDAW) (Entry in force, 1981) · Convention on the Rights of the Child (Entry in force, 1990) · Vienna Declaration and Program of Action (Adopted by the World Conference on Human Rights, 1993) · The Declaration on the Elimination of Violence against Women (Adopted by the UN General Assembly in1993) · Inter-American Convention on the Prevention, Punishment and Eradication of Violence against Women, "Convention of Belem do Para" (1994) · Beijing Declaration and Platform for Action (Adopted by the Fourth World Conference on Women, 1995) Source: World Health Organization, 1999. The Convention on the Elimination of All Forms of Discrimination against Women (CEDAW)-one of the most influential--requires signatory governments to implement specific reforms and periodically report on their progress to the United Nations. The Optional Protocol to CEDAW (adopted in 1999) enables individuals and organizations from states that have ratified the Protocol to file claims directly with the committee in charge of overseeing the Convention's implementation. By June 2004, 177 countries had ratified CEDAW, and 75 countries had signed the optional protocol (United Nations, n.d.). A study of CEDAW's global impact noted numerous examples of court rulings and legal reforms-including those that dealt with violence against women--that made explicit reference to the CEDAW Convention (McPhedran, 2000). Unfortunately, many countries have signed or ratified CEDAW "with reservations" about certain articles, have yet to bring their legislation into compliance, or simply have failed to enforce their laws. Advocacy for legislative change International donors, United Nations monitoring bodies, and women's rights organizations have lobbied governments around the world to revise their civil and penal legislation, and these efforts have achieved progress in many countries (Center for Reproductive Law and Policy, 2002; Mehotra, 1998). Their efforts have included research and dissemination of findings, educating parliamentarians and journalists, mass media campaigns to build awareness of the problem, and targeted lobbying for legal reform. The Nicaraguan Network of Women against Violence (composed of over 100 women's organizations) drafted legislation on violence against women and successfully campaigned for its passage (Ellsberg, Liljestrand, and Winkvist, 1997). In South Africa, Soul City carried out a similar campaign in collaboration with the National Network on Violence against Women (Usdin et al., 2000). These efforts continue in middle and low-income countries, since much legislation remains flawed. At the same time, new legislation sometimes raises new concerns (e.g. mandated reporting of violence against adult women, a problematic policy that is discussed below in the health section) or provokes a conservative backlash, requiring advocates to defend the laws after they have been passed. Legislative reform in low and middle-income countries Recent criminal law reforms include revision of penal codes to strengthen criminal sanctions against perpetrators of family, domestic and sexual violence; specific legislation to address domestic or sexual violence; reform of laws and policies that regulate criminal justice procedures, and stronger regulation of the obligations of public and private service providers. Much progress has also been made in the reform of civil legislation. For example, countries in Africa have passed laws raising the minimum age of marriage to 18 for women and men; granting women in customary marriages the same privileges and rights as women in formal marriages (including rights to property and child support); and introducing new 19 enforcement policies to ensure compliance with maintenance and child support orders (Ghana, Kenya and South Africa) (Center for Reproductive Law and Policy, 2001). In Latin America, there has been a long and continuing trend to improve women's rights to divorce, control over marital property, and child support (Center for Reproductive Law and Policy, 2000). Box 2. Examples of recent legislative reforms that address gender-based violence · Laws that criminalize and/or strengthen sanctions against perpetrators of family or domestic violence (South Africa) · Revision of the penal code to allow public prosecution of sexual offenses (Peru) · Introduction of marital rape as a criminal offense (Federal District, Mexico) · Eliminating provisions that allow rapists to escape criminal sanctions by agreeing to marry the victim (Argentina, Peru) · Laws that require police to inform rape victims about the possibility of legal abortion (Brazil) Symbolically, these reforms represent a significant achievement in the effort to strengthen women's rights and reduce violence against women. Some research has documented a positive impact on intermediate outcomes, such as an increase in the number of women who report cases to police (Villanueva, 1999). The overwhelming lesson from research on legislative reform in low and middle-income countries is, however, that changing the law is only the first step in a long process. Women often face social, economic or procedural barriers to accessing the justice system. Much legislation has been implemented poorly if at all. Governments often fail to budget resources for implementing new legislation. Police often fail to investigate cases or protect women in danger, and the judiciary is often unable or unwilling to enforce the laws. Widespread failures to enforce existing laws have been documented by Human Rights Watch (1997; 1999; 2000; 2001; 2002; 2003; 2003) in countries as diverse as Afghanistan, Brazil, Iraq, Jordan, Nepal, Pakistan, Peru, South Africa, Uganda, Uzbekistan, and Zambia. For example, South Africa passed new legislation in 1998 that strengthened criminal sanctions for violence by intimate partners and other types of perpetrators. It imposed obligations on police to arrange shelter and medical treatment for survivors as well as to provide information about their rights, and included specific sanctions for noncompliance. The law was a significant milestone, but there were delays and difficulties in implementation, and a 2001 evaluation found that the government's failure to allocate sufficient resources to police, courts and support services had undermined the effectiveness of the act (Parenzee, 2001). Box 3. The challenges of legal reform in low-income countries: Zambia The experience of Zambia exemplifies the problems that confront low-income countries. During the past decade, Zambia revised its civil and penal codes and implemented a nationwide system of "Victim Support Units" to address violence against women. By 2001, these units had trained officers (many of whom were women) in nearly every police station in the country. Nonetheless, an evaluation by Human Rights Watch (2002) noted the following problems: Women and children face serious social and cultural barriers to legal redress. Women are often reluctant to use legal remedies because they don't believe that they are entitled to protection, because they are afraid of additional violence from the perpetrator or their families, because they are pressured to avoid bringing "shame" upon their family, or because jailing the perpetrator would cut off the family's economic support. Support for new laws has often been low among police, the judiciary and the general public, especially when laws counter long-standing traditions of discrimination against women. Law enforcement institutions often simply refuse to enforce the laws. Police face a severe lack of resources for investigating and prosecuting cases of violence, for adequate office space, training, equipment, transportation, forensic tests, etc. As a United States State Department report described, "the VSU (Victim Support Unit) has no transportation, no phone--so how can you report to police?" (cited in Human Rights Watch 2002, page 71). Finally, the VSU system has been undermined by bias and corruption in other parts of law enforcement. For example, despite the best efforts of specialized units, the rest of the police force is 20 known for violating human rights; law enforcement cannot be effective in addressing violence against women as long as the broader police force has a reputation for abuse. Institutional reform of police, the judiciary, forensic medicine, and legal aid The example of Zambia highlights the need to go beyond legislative reform by strengthening key law enforcement institutions, including police, the judiciary, the forensic medical system, and legal aid. Even in countries with strong legal systems, law enforcement institutions have historically responded to intimate partner and sexual violence with bias, neglect and mistreatment. For example, in Britain, researchers noted that police and judges often minimized the criminal nature of gender-based violence and expressed bias against victims of sexual abuse until serious reform began in the 1970s (Morley and Mullender, 1994). In middle and low-income countries, law enforcement institutions often face a severe lack of resources for personnel, equipment, training, and transportation. Many legal systems are plagued by police corruption, expensive, slow and inaccessible courts, and an unaccountable judiciary (Human Rights Watch, 1997, 1999, 2000, 2001, 2003a, and 2003b). A growing number of low and middle-income countries have tried to improve the law enforcement response to gender-based violence by training professionals, reorganizing police and courts, and trying to provide a more comprehensive response to survivors. Evidence of effectiveness is relatively limited; most well-evaluated initiatives come from high-income countries, and the lessons learned may not be applicable to developing countries. Evaluations of law enforcement reforms in low and middle income countries have typically been limited to case study approaches drawing from police records (notorious for under-reporting), qualitative perspectives from key informant interviews, intermediate outcomes such as changes in attitudes and knowledge among police and judges, and interviews with small numbers of women who have sought legal redress. Population-based data collection, control groups, or follow-up among more than a handful of survivors are rare. Nonetheless, the following initiatives illustrate the types of efforts that have produced important lessons learned. Training personnel in the police and judiciary and other parts of the justice system Throughout the world, organizations have launched efforts to improve the knowledge, attitudes, and practices of justice sector personnel regarding gender-based violence. Some law enforcement institutions organize training internally, as did South Africa following passage of the 1998 Domestic Violence Act (Usdin et al., 2000). In other settings, NGOs such as Rozan in Pakistan (Rashid, 2001), Profamilia in the Dominican Republic (Guedes et al., 2002), and the Musasa Project in Zimbabwe have trained law enforcement personnel on issues related to gender-based violence. Elsewhere, governments have collaborated with the United Nations to provide training and support for the police and judiciary. For example, ILANUD is a joint institute of the government of Costa Rica and the United Nations that works with governmental agencies throughout Latin America to improve the work of prosecutors, judges, lawyers, police and other professionals in criminal justice generally, and gender-based violence specifically (Villanueva, 1999; ILANUD, n.d.). Most of these initiatives have been evaluated using key informant interviews and pre and post questionnaires before and after training-if they have been evaluated at all. Nonetheless, training appears to be both constructive and urgently needed (Rashid, 2001; Villanueva, 1999). Other lessons learned include the finding that changing attitudes of law enforcement is a challenging, long-term process. The quality of the trainings' content and the skills of the trainer are essential. Training appears to be most effective when all levels of personnel (especially high-level officials) participate, and when training is backed up with changes throughout the institution, such as policies, procedures, adequate resources, and continual monitoring and evaluation. Special police stations or cells for crimes against women All-women police stations began in Brazil and were later tried in other countries in Latin America and Asia. As of 2003, for example, Nicaragua had 17 police stations for women and children (called 21 Comisarías) operating throughout the country with the help of funding from international donors (Velzeboer et al., 2003). Alternatively, countries such as Zambia have experimented with special police "cells" for women and children, composed of one or more police officers working in a regular station but dedicated to cases of family and/or sexual violence. Much research has evaluated women's police stations, primarily by gathering data on the number, nature and outcomes of cases reported, and by exploring perspectives of police, judges, NGO staff, lawyers, prosecutors, judges, detectives, and survivors (Thomas, 1994; Mesquita da Rocha, 1999; Brown, 2001; Jubb and Izumino, 2002; Krug et al., 2002; Human Rights Watch, 2002). Special police stations generally appear to increase reporting as well as the likelihood that women will receive services such as forensic exams, counseling, emergency contraception and sexually-transmitted infection (STI) prophylaxis. On the other hand, evaluations have documented numerous problems. Women officers have not necessarily demonstrated better attitudes towards victims of violence simply by virtue of their sex, and special stations are often under-funded and lack equipment, transportation, training, staff or referral services. Even when stations work well, their efforts are often undermined by law enforcement personnel (forensic doctors, prosecutors and judges) who are unwilling or unable to enforce the law. As a result, conviction rates sometimes remain unchanged. Finally, the creation of women's police stations may encourage regular police stations to abdicate responsibility for crimes against women, which is especially problematic in rural areas where women must travel long distances to reach police of any kind. Hence, organizations such as Human Rights Watch (2002) argue against separate stations for women as a long-term strategy, in favor of integrating a better response to violence throughout law enforcement. Protection orders Orders of protection have been introduced in countries such as South Africa, to increase the safety of women who are in danger of ongoing violence. Protection orders can take a great many forms, including prohibiting someone from entering a residence or from approaching a woman. A 1998 review of the U.S. literature found many studies that evaluated the effectiveness of protection orders, but no randomized controlled studies on their risks and benefits (Chalk and King, 1999). Some studies suggest that they can increase women's safety, but even in settings with strong legal systems, men routinely violate such orders, often without serious consequences. In low and middle-income countries, most research on protection orders highlights the challenge of this type of tool in resource-poor settings. Procedural barriers make them difficult to obtain in many settings, police often lack the resources to enforce them, and police are often willing to accept the perpetrator's assurances that violence will stop (Human Rights Watch, 2000). Judicial reforms: special courts, procedural reform, and court services A host of countries have enacted judicial reforms to address gender-based violence (Center for Reproductive Law and Policy, 2001 and 2002). Special courts for crimes against women and children and/or for family matters have been introduced in Asia, Africa, and Latin America. Other reforms include policies that allow closed court hearings for victims of sexual offenses (Tanzania); closed-circuit television testimony and separate waiting areas for vulnerable witnesses or victims (Zimbabwe and South Africa); and special courts for sexual offenses (South Africa, Ghana). Many countries have revised the rules regulating evidence used for prosecution. For example, Tanzania eliminated the requirement of testimonial corroboration in rape cases in 1998. Unfortunately, much remains to be done, few reform measures have been rigorously evaluated, and little information exists about their effectiveness. Evaluations often highlight challenges rather than successes. For example, judicial reforms often are poorly implemented, simply because governments lack the resources to educate judicial personnel about the changes (Usdin et al., 2000). In other cases, reform produces unintended consequences or other problems, such as the heavy emphasis on couple reconciliation in Latin American and Indian family courts that often leads judges to pressure women to reconcile with abusive partners in order to preserve the family (Center for Reproductive Law and Policy, 2000; Mitra, 1998). 22 Increasing multi-sectoral collaboration with the justice system Evidence suggests several reasons why inter-sectoral collaboration is an essential element of law enforcement reform. First, many women cannot access the justice system unless they first obtain basic information about their rights, about how to report cases to the police, and/or about how to find legal aid services. In response, non-governmental organizations throughout the world have integrated basic referral services and legal services for survivors of violence into community-based health programs, social services, and economic development programs (BRAC, n.d.; Inter-American Development Bank, 2002). Second, increasing women's access to social services may make it easier for police and courts to enforce the laws. For example, an evaluation of the 1998 Domestic Violence legislation in South Africa found that police had difficulty enforcing orders of protection, prosecuting cases, and imposing jail time on perpetrators without adequate community services to assist women and children with emergency shelter, long-term housing and economic support (Parenzee, 2001). The Coordinated Community Response model (first piloted in Duluth, Minnesota, United States) is perhaps the best known and most rigorously evaluated model of ways to increase collaboration between law enforcement, health, social services, schools, and religious institutions. Studies have demonstrated that this approach significantly improved law enforcement outcomes, such as the numbers of cases reported, the numbers of arrests, prosecution and conviction rates (Pence, 1995; Shepard and Pence, 1999). Similar approaches have been tried in middle-income countries such as Costa Rica and low- income countries such as Nicaragua. Beginning with a National Plan in 1994, Costa Rica aimed to ensure a comprehensive response to survivors by mobilizing collaboration between law enforcement and sectors such as health, counseling, housing, employment, and child welfare services (Villanueva, 1999). In Central America, the Pan American Health Organization (PAHO) collaborated with governments and NGOs throughout the region to link law enforcement with service providers. Qualitative case studies suggest that this is a promising model of law enforcement reform for developing countries. Reform of the medico-legal system Many countries restrict the collection of forensic evidence to specially certified doctors, such as district surgeons (South Africa) or forensic physicians (most countries in Latin America). These professionals are often public sector employees and are often notorious for poor access, poor treatment of survivors, and unwillingness or inability to provide urgent medical care including emergency contraception and prophylaxis for STIs (Human Rights Watch, 1997). In some settings, such as Pakistan, Peru and Turkey, forensic exams are often used (unreliably) to establish whether and when a girl or woman lost her virginity--evidence that may be used against her in court. Under the Hudood Ordinances in Pakistan, a woman who alleges rape and then fails to "prove" the case can even be sentenced to prison for the criminal offence of illicit sex (Human Rights Watch, 1999; Brown, 2001). Some countries have attempted to reform their systems; for example, after researchers and advocates documented barriers to access and lack of emergency medical care for survivors of sexual assault in South Africa, the government issued regulations to reform the system, including: a) procedures to ensure that survivors receive emergency medical services including STI prophylaxis; b) introduction of a system of forensic nurses; c) policies to reduce barriers to reporting; and d) more coordination between police, prosecutors, judges and service providers (Kim, 2000; Brown, 2001; Kim, Martin and Denny, 2003). In theory, these policies have the potential to improve the system, but future evaluations are needed to assess their impact. Legal aid services Many NGOs and a few publicly-funded programs provide legal aid services to women who experience gender-based violence. The effectiveness and impact of these services have rarely been evaluated. One exception was the Projusticia project, launched in Ecuador in 1994 (Owen and Portillo, 2003). As part of a 10 million dollar justice sector reform project, the World Bank contracted three NGOs to provide expanded legal aid services for poor women who needed help in dealing with domestic violence and lack of child support (among other issues). The two are closely linked, since nearly half the women in Ecuador 23 who seek child support cite "severe physical violence" by the father. Evaluators collected qualitative and quantitative data among 180 clients and 182 matched controls to examine whether or not the services increased the incomes of clients, reduced domestic violence, or improved the quality of children's education. The evaluation found that 38% of clients obtained some child support; in half of those cases, child support made a substantial contribution to their standard of living (though none of the mothers felt that the amount was sufficient or fair). Levels of domestic violence were similar among clients and non- clients before seeking services, but lower among clients after receiving services. However, researchers acknowledged that women who feared future harm from their former partner may have been less likely to seek services at legal aid clinics in the first place, pointing to the difficulty in assessing evaluation outcomes when the clients accessing the services often represent a more empowered group of women than the control group. From a cost-effectiveness point of view, researchers suggested that the legal aid clinics had missed an opportunity to "capitalize on [a] spillover effect" by publicizing women's rights and the effectiveness of clinics in representing women's interests. They argued that this type of campaign might make it easer for other women to achieve out-of-court settlements without having to go through the risks and difficulties of formal court proceedings. Community mobilization Reforming community-based non-formal systems of justice In certain parts of the world, the formal justice system has a limited presence. For example, Human Rights Watch has written about the role of semi-official traditional councils, mahallas, in Uzbekistan, which arbitrate local disputes and act as gatekeepers to the formal justice system. Unfortunately, they frequently protect perpetrators of gender-based violence, force women to live with ongoing abuse, and block women's attempts to access the police and the courts, sometimes threatening to testify against women in court (Human Rights Watch, 2003a). Human Rights Watch argues that reforming these traditional councils in Central and South Asia is urgently needed to improve women's access to justice and human rights. Such an approach appears to have achieved some success in settings such as rural Gujarat, Uttar Pradesh and West Bengal, India (International Center for Research on Women, 2002). In West Bengal, rural women's groups known as the Samity adapted the traditional village dispute resolution system--shalishi-- to address violence against women (Sadasivam, 2000; International Center for Research on Women, 2002; Shramajibee Mahila Samity, 2003). Cases include physical violence by husbands and in-laws, forced sex by husbands, dowry disputes, and other forms of violence against women. Facilitators of the shalishis conduct an inquiry, hold a public hearing (depending on the sensitivity of the case), try to negotiate a resolution, and then call on the community to enforce the decision. Police are asked to intervene only if the shalishi fails. By 2002, caseloads had grown to over 400 a year. Key informant interviews, focus groups and a survey of 151 women suggested that the shalishi was vastly more accessible than the formal legal system. While the formal legal system focuses almost exclusively on punishing the perpetrator, the shalishi aimed to restore women's safety while keeping the family intact. Some criticize the emphasis on preserving the family, but researchers note that rural women often have no means of survival outside the family in this setting. The shalishi appeared to be effective in stopping most cases of physical violence and most women were satisfied with the process. Community mobilization to influence and monitor policy reforms An important role of community mobilization is to pressure policy-makers to change laws, implement reform, or monitor the performance of law enforcement institutions. For example, in 1999, after the South African Domestic Violence Act was passed, Soul City launched a campaign in partnership with the National Network on Violence against Women to motivate policy makers to speed up the date that the act would go into effect, to develop a strategy for implementing the law, to allocate resources for training police and the judiciary, and to develop plans for monitoring and evaluating enforcement (Usdin et al., 24 2000). They used mass media, lobbying efforts, and social mobilization to pressure Ministers of Justice and Safety, members of those departments, the Task Team in charge of implementing the legislation, the police, and the judiciary. Based largely on key informant interviews, Usdin et al. (2000) concluded that the campaign contributed to more timely and effective implementation of the act. Individual behavior change strategies Campaigns to increase public awareness of and support for women's legal rights An essential component of effective legislative reform is to raise awareness of and build support for new laws among the general population. Increasing access to justice requires that women know their rights and feel that they can exercise them. Thus, many NGOs have launched media and legal literacy campaigns to raise awareness of new legislation and encourage women to exercise their rights. Well- evaluated initiatives include Soul City (South Africa) and the Nicaraguan Network of Women against Violence. Both have demonstrated an impact on knowledge of and attitudes towards women's rights through large population-based surveys, including raising awareness that violence against women is a crime, and knowledge of specific legislation (Scheepers, 2001; Scheepers and Christophides, 2001; Muirhead, Kumaranayake, and Watts, 2001; Ellsberg, Liljestrand and Winkvist, 1997; Velzeboer et al., 2003). HEALTH Background Historically, the health sector was slow to recognize the public health implications of violence against women. Even today, many medical and nursing schools do not adequately prepare professionals to recognize or respond to the health consequences of domestic violence, rape or sexual abuse; providers often view violence against women as a social problem and hesitate to discuss physical or sexual violence with patients; and many organizations do not equip personnel to respond appropriately to girls and women who disclose violence (Heise, Ellsberg and Gottemoeller, 1999). Mensch, Bruce and Greene (1998) argue that most reproductive health initiatives for young people have assumed that sexual activity is voluntary and therefore have aimed to help young people "make better and more responsible decisions." This approach ignores evidence that large numbers of girls and young women experience forced sex and/or cannot negotiate sex, condom use or other contraceptives without fear of physical violence (Jejeebhoy and Bott, 2003). Overlooking the health implications of gender-based violence is not just a missed opportunity. Women sometimes disclose intimate partner violence, rape or sexual abuse to heath care providers, and providers who respond by blaming the victim may inflict severe emotional trauma. Providers who view violence as a social rather than a health issue may fail to provide holistic care, to recognize women in danger, or to provide necessary, even life-saving care, such as STI prophylaxis. Moreover, health systems that do not protect patient confidentiality may put women at risk of additional violence from partners or other family members. Growing evidence suggests that public health policies, institutions and programs must pay more attention to gender-based violence not only as a public health problem in and of itself, but also as a key component of the HIV/AIDS pandemic. Overview of health sector initiatives In recent years, the health sector has launched many gender-based violence initiatives (see Table 2.2). Countless health care organizations have begun to improve the quality of care they provide to survivors of gender-based violence. Public health programs have a long history of working to change sexual attitudes, norms and behaviors, and they have begun to apply those strategies to violence prevention. Despite the 25 proliferation of prevention and intervention initiatives, research on their effectiveness is still at an early stage. Few have been able to demonstrate an impact on levels of violence or benefits to survivors, though many appear promising. Table 2.2. Examples of objectives and initiatives from the health sector Level Objectives: Specific Initiatives: Laws and To improve laws and policies · Reformsoflaws andpoliciesregulatingthe medico-legal policies · Clarifyproviders' legalresponsibilities system (e.g. introduction of forensic nurses) · Encourageabetterhealthsector responsetoGBV · Reformoflawsandpoliciesregulatinghealthcareprovider's through national, regional, and municipal policies obligations vis-à-vis victims of gender-based violence regarding screening, referral, documentation and · Mandatory reportinglaws/policies counseling for victims of violence · Ensuresurvivors'rightstoservices(e.g.emergency · Nationalhealthpoliciesandprotocols contraception, STI prophylaxis, etc.) · Laws/policiesgoverningforensic medicine;provider obligations, abortion, EC and patient confidentiality Institutional To strengthen the response of health care and public · Policies,proceduresandprotocolstoimprovethehealth care reform health institutions to gender based violence response · Raiseawarenessof thelinksbetween violenceand · Sensitizationandtrainingofhealthprofessionals health among service providers, managers, and public health policy makers · Routinescreeningandreferralsystems · Improvethequalityofcareforsurvivorsofviolence, · Developmentofinformationsystems suchasepidemiological including identification, treatment, documentation, surveillance, and morbidity statistics on violence information referrals and follow-up · Specializedsurvivorservices (counseling,supportgroups) · Increase coordinationwithothersectorsthatprovide · Improvedcoordinationandreferralsto NGOsandother sectors services or work on violence prevention · Curricularchanges intrainingofnursesand medicalpersonnel Community To increase community mobilization to address gender- · Coalitionsforpublichealthresearch andadvocacy mobilization based violence as a public health problem · Strengthencommunitysupportforsurvivorservices · Communitylevel preventionand mobilizationinitiatives · Strengthencoalitionsandnetworks · Community-based awarenesscampaignsaimedat mobilizing journalists, policy makers, and opinion leaders · Improveattitudes, norms,practicesandresourcesatthe community level Individual To improve knowledge, attitudes and practices of key · Clinicandcommunity-basededucationefforts(theatre, videos, behavior groups and the broader population pamphlets, talks, etc.) change · Promotegender-equitable,nonviolentsexual · Massand multi-mediabehaviorchangecampaigns,such as partnerships edutainment programs (e.g. Soul city and Sexto Sentido) · Increase women's abilityto makedecisionsaboutthe · Programsfor men aimedatpromotinggenderequitable timing and nature of sexual relationships relationships and changing norms, attitudes and behaviors · Decrease tolerance forviolencebyraisingawarenessof · Gender-basedviolencepreventionwithinHIV/AIDSand gender-based violence as a public health problem adolescent reproductive health programs · Encouragevictims ofabusetoseekhelpandtodisclose violence to service providers Reform of laws and policies Many governments, particularly in Latin America, have enacted legislation and policies specifying the obligation of the health sector to address violence against women, either through ministerial decrees, or as part of national family violence legislation. These policies are often fairly general; sometimes limited to recognizing violence as a public health problem and outlining basic principles for ensuring an integrated service response to survivors. While these laws and policies may have the potential to raise awareness of gender-based violence among health care providers, a review of Central American countries in 2001-2003 found that the policies had not been widely disseminated, and most health providers were either not aware of the policies, or did not know their contents (Velzeboer et al., 2003). In some cases, national legislation regarding health care and gender-based violence has occasioned unforeseen problems. For example, several countries have passed laws requiring health providers to report suspected cases of domestic violence to legal authorities. This puts providers in the position of violating the privacy and confidentiality of clients and may reduce women's willingness to disclose violence. Providers may also be more reluctant to ask women about violence for fear of becoming 26 involved in legal cases. Another example of problematic legislation are laws mandating the health sector to provide certain services, such as specialized services for victims or treatment for batterers, without allocating resources enabling the ministry to provide these services. Another area of legal and policy reform is access to safe abortion for survivors of sexual violence. Many countries that severely restrict women's access to legal abortion allow exceptions in cases of rape or incest; in other settings, advocates are working to make these exceptions part of the law. Nonetheless, even in countries where this right exists by law, safe abortion often remains inaccessible to survivors of sexual violence. In a few settings, such as Brazil and Mexico, governments and NGOs have begun to address this issue as a legal and a policy matter (Billings et al., 2002). Institutional reform The "systems approach" to improving the health care response to violence against women Over the last 20 years, many organizations have tried to improve the health service response to gender- based violence in high, low and middle-income countries. Organizations sometimes take small steps, such as providing a single training session for staff. Unfortunately, evidence suggests that without system-wide reforms and support, single training sessions or even routine screening policies rarely produce long-term changes in the quality of care for survivors (McLeer et al., 1989; Heise, Ellsberg, and Gottemoeller, 1999). Instead, Heise and colleagues argue that the most effective way to improve the health care response is to use a "systems approach" involving reforms throughout the organization. Typically, these initiatives include changes in norms, policies and protocols, infrastructure upgrades to ensure private consultations, training all staff (including managers), ensuring that providers have adequate resources such as referral networks and directories, and strengthening the ability of staff to provide emergency services such as danger assessment, safety planning, emotional support, STI prophylaxis, and emergency contraception. In settings where adequate referral services do not exist, health programs sometimes offer specialized services such as counseling, legal aid and women's support groups. The International Planned Parenthood Federation, Western Hemisphere Region (IPPF/WHR) carried out an initiative illustrating the "systems approach" in four member associations in Latin America, namely: Profamilia (the Dominican Republic), INPPARES (Peru), and PLAFAM (Venezuela), with some participation from BEMFAM (Brazil). Quantitative and qualitative baseline, midterm and follow-up studies concluded that the initiative improved provider attitudes and practices; strengthened patient privacy and confidentiality; increased detection of women who experienced physical and sexual abuse; improved the overall quality of women's health care; and benefited survivors through the provision of specialized services such as legal aid, counseling and support groups (Guedes, Bott, and Cuca, 2002; Guedes et al., 2002; Bott, Guedes, and Guezmes, forthcoming). This initiative benefited from generous funding from international donors, and it might be difficult for other organizations to replicate the project in its entirety; however, IPPF/WHR has disseminated a large body of recommendations and tools designed to help organizations in low-income settings build on their experiences. Routine screening (also called routine enquiry) Research indicates that without routine screening, providers typically identify only a fraction of women requiring assistance with physical or sexual abuse. Routine screening for violence has increasingly been considered the standard of care within women's health services in the United States and other industrialized countries (American Medical Association, 1992; Buel, 2001). However, a vigorous debate has erupted over the benefits and risks of routine screening, particularly in resource-poor settings (Ramsay et al., 2002; Garcia Moreno, 2002). Some argue that routine screening may harm women in settings where providers are unprepared to respond appropriately, where privacy and confidentiality cannot be ensured, and where adequate referral services do not exist. In many settings, providers blame victims of gender-based violence-without an appreciation of gender issues or human rights-and may 27 inflict additional pain on women who disclose violence in response to screening questions (e.g. Kim and Motsei, 2002). Moreover, some question the feasibility of universal screening in low and even middle- income countries given the scarcity of time and resources. Researchers and advocates agree, however, that women's health care programs have an ethical obligation to ensure that staff are prepared to respond to disclosures of violence against women, whether or not they implement routine screening. Community-based networks and multi-sectoral collaboration Many health care organizations have joined community-based networks with other governmental and non-governmental institutions, such as legal aid, criminal justice institutions, social welfare, education, and social services. Evidence suggests that these networks enhance the range and quality of services accessible to survivors, and help to ensure that women do not fall through the cracks when they have to interact with several different institutions. For example, the Pan American Health Organization (PAHO) and the Inter-American Development Bank have pioneered multi-sectoral approaches in Latin America. PAHO has worked with 10 countries in Central America and the Andes to promote an integrated health sector approach to gender-based violence (Velzeboer et al., 2003). The PAHO program aimed to improve policies and legislation related to violence, to increase access to services and to forge multi-sectoral networks at a community level for violence prevention. Similar pilot initiatives have taken place in South Africa (Jacobs and Jewkes, 2001), Zimbabwe (Watts and et al., 1997) Bangladesh, (Haque and Clarke, 2001), Brazil (d'Oliveira and Schraiber, 2001), Thailand (Grisurapong, 2002) and elsewhere. Although evaluations of these approaches have not necessarily measured their impact in quantitative terms, case study assessments suggest that they are promising. Community mobilization Coalitions for public health research and advocacy Health sector coalitions can play an important role in advocating for public policy and institutional reform. For example, the South African Gender-Based Violence and Health Initiative (SAGBVHI) consists of 15 partner organizations and individuals in South Africa. They conduct research, build research capacity, disseminate research findings, and use research to advocate for policy reforms. SAGBVHI works closely with the Gender and Women's Health Directorates of the National Department of Health. The impact such networks produce are difficult to measure, but informal assessments suggest that SAGBVHI has achieved important results. For example, they convinced two medical schools to increase their emphasis on gender-based violence within their curriculum, helped to develop a one-week module on rape for medical students, and contributed to new national policies on gender-based violence and health. Their work exemplifies how researchers can collaborate with government to translate research findings into policy (Medical Research Council, 2003; Guedes, 2004). Community-level initiatives to reduce gender-based violence Many NGOs have launched programs to promote community-wide changes in attitudes and practices related to gender norms and violence against women--often as a component of HIV/AIDS prevention or reproductive health programs. The few that have been well evaluated suggest that community level approaches can be effective in changing violence-related attitudes and behaviors. The following three programs exemplify this approach: The Stepping Stones (The Gambia, Ghana, Kenya, the Philippines, South Africa, Tanzania, Uganda, Zambia) is a community "training package" focused on STI/HIV prevention, gender, and reproductive health, originally developed in Uganda. The aim is to encourage communities to question and rectify the gender inequalities that contribute to HIV/AIDS, gender-based violence and other health issues. The program uses workshops, community-wide meetings, drama, peer group discussions and other strategies and has been adapted to sites in the Gambia, Ghana, Kenya, the Philippines, South Africa, Tanzania, and Zambia. Numerous 28 evaluations have been conducted (Renton et al., 2000; Shaw, 2000; Shaw, 2002a; Shaw, 2002b; Paine et al., 2002; White, Greene and Murphy, 2003; Interagency working Group, 2003). For example, the Medical Research Council (from the United Kingdom) conducted an external evaluation of program effectiveness in the Gambia using key informant interviews, focus groups, and a KAP (knowledge, attitudes and practices) survey in four villages, including two program sites and two controls. That study found that the Gambia program improved self-reported attitudes and behaviors related to violence against women. Specifically, the program reduced the social acceptability of wife-beating at the community level and appeared to produce a corresponding drop in that behavior. Qualitative findings from other Stepping Stones sites suggest similar benefits. Program H (Bolivia, Brazil, Colombia, Jamaica, Mexico and Peru) is being carried out by four NGOs. It aims to change gender norms and sexual behaviors in Bolivia, Brazil, Colombia, Jamaica, Mexico and Peru (Barker, 2003; White, Green and Murphy, 2003; Guedes, 2004). The initiative includes four components, namely: a) training professionals to work with young men in the area of health and gender-equity using a set of manuals and videos; b) social marketing of condoms; c) promoting health services; and d) evaluating changes in gender norms. In 2002, PROMUNDO and Horizons began a 2-year evaluation to measure the effectiveness of two different approaches, compared to a control site. Researchers have developed a "Gender-Equitable Men" (Leichert) scale with 24 items for measuring attitudes. Methods include pre and post-tests as well as a six-month follow-up community-based survey. In addition, they are gathering qualitative information among men and their female partners. Preliminary results suggest that the program has been successful at increasing gender equitable norms and reducing behavior that puts men at increased risk of HIV/AIDS. ReproSalud (Peru): Manuela Ramos launched ReproSalud in 1995 as a USAID-funded rural reproductive health program. ReproSalud used participatory rural appraisal (PLA) to help women's groups identify women's reproductive health needs and to organize community meetings to design strategies to address those needs. Domestic violence and forced sex within marriage emerged as important problems in those communities. In response, ReproSalud organized workshops for women and men on gender issues, carried out community awareness campaigns and established a microcredit program for women. By 2002, ReproSalud had reached over 123,000 women and 66,000 men. Qualitative and quantitative evaluation data suggest that the community-based PLA approach had a positive impact on attitudes and behaviors related to gender based violence (Rogow and Bruce 2000; Ferrando, Serrano, and Pure, 2002, cited in Boender et al., 2004). The quantitative evaluation (using community based surveys) was complicated by the fact that the project coincided with a period of strong investment by the Ministry of Health, which made it difficult to isolate the project's impact. Gender-equitable attitudes and practices increased significantly in both intervention and control communities, though improvements in intervention sites were slightly higher. The qualitative data suggested a much greater difference in intervention and control sites and gathered evidence of dramatic changes in social relations and men's behavior. Respondents spoke at length about decreased alcohol consumption, domestic violence, and forced sex in all intervention villages studied. In the words of one 35 year-old woman, "Before, they brutally forced sex. They hit, especially when they were drunk. Now, no more" (Rogow and Bruce, 2000, page 20). Individual behavior change strategies Many other programs have attempted to produce individual (rather than community-level) behavior change by working with individual men and boys. White, Greene and Murphy (2003) reviewed the literature on such programs aimed at men. That review suggests that less information is available on the effectiveness of individual behavior change strategies compared to community-level approaches. Some 29 programs report a positive impact on men's self-reported attitudes and behaviors, but most information is still preliminary or based on evaluations without control groups or baseline data; nonetheless, the following examples illustrate promising work being done in this area. Community-based workshops for men "Men as partners" (MAP) is a program by EngenderHealth in South Africa (with expansion beginning in Kenya). This program involves community-based workshops with men and mixed sex audiences, in settings such as workplaces, trade unions, prisons and faith-based institutions, among others (Levack, 2001; EngenderHealth, 2003; White, Greene and Murphy, 2003). Based on the premise that gender inequity contributes to both AIDS and violence against women in South Africa, MAP promotes discussions of gender issues, power dynamics and gender stereotypes. A preliminary evaluation suggested that-compared to control groups--a higher percentage of participants believed that women and men should have the same rights and that wife-beating was wrong. Adolescent boys appeared more open to changing their view of masculinity than older men. A large, multi-year prospective study is planned in collaboration with Frontiers / Population Council for the near future, which will include baseline and follow-up data collection in Soweto, Durban, Cape Town, and Umtata. This future research may be able to shed more light on the effectiveness of this approach. Similar work has been carried out by CANTERA (Population Education and Communication Center) in Latin America, specifically in Nicaragua, Costa Rica, El Salvador, and Guatemala. CANTERA runs workshops for men on masculinities, gender, power, and violence. A 1997 evaluation suggested that as a result of the project, men expressed less "macho" attitudes, assumed more responsibility for household chores and demonstrated more solidarity with women. However, the evaluation lacked a control group or baseline data, and the participants appeared to be a highly self-selected group. Violence prevention within reproductive health, HIV prevention and life-skills programs for youth Many NGOs have tried to change attitudes and behaviors relative to gender-based violence among youth, often in the context of reproductive health programs, HIV/AIDS prevention, "life skills" and/or peer education. Many are school-based and will be discussed in the section devoted to the Education Sector. Research on the effectiveness of these initiatives is still preliminary; nonetheless, these initiatives appear to be promising-both for violence prevention and for prevention of HIV/AIDS, unwanted pregnancy, and unsafe abortion. For example, the New Visions Program is a community-based "life-skills" program for males aged 10-20 in rural Egypt implemented through youth councils and NGOs. It aims to increase respect and support for girls and women, while helping males to improve their own lives. By mid 2004, it will have reached approximately 13,895 beneficiaries through 720 classes run by 633 facilitators. Preliminary evaluation results suggest that the program raised awareness of gender issues and improved some attitudes towards gender roles, equity and violence in the intervention villages compared to controls. However, the final results have not yet been published, and it remains to be seen if the program has changed behavior (White, Green and Murphy, 2003). Better Life Options for Boys (India) is a similar project implemented by 10 NGOs involving over 8,000 boys. Some evidence suggests that the project has had a positive impact on gender equitable attitudes and behaviors, but the evaluation had some limitations, including the lack of a control group (White, Green and Murphy). Mass media `Entertainment-education' (`edutainment') programs "Entertainment-education"-the use of radio and television to promote health and social change was pioneered by Televisa, the largest broadcast network in Mexico, which used television soap operas to model behaviors, promote values, and demonstrate the consequences of choices. This strategy has been used in Asia, Africa, and Latin America and has demonstrated effectiveness in changing behaviors related to reproductive health, AIDS education and the status of women (CDC, 2004). NGOs have recently begun to use radio and television `edutainment' to address violence against women. Whether it can be effective in changing violent behavior has not yet been demonstrated, but its success in other areas suggests promise for violence prevention. 30 The most well known examples of using `edutainment' for gender-based violence prevention are Sexto Sentido, carried out by Puntos de Encuentro in Nicaragua, and Soul City, carried out by the Institute for Health and Development Community in South Africa and eight other Sub-Saharan countries in Africa. Both have used prime time television soap operas, radio programs, school-based work, and other media to address violence against women, gender, sexuality and rights. Some research suggests that Sexto Sentido has had a positive impact (Abaunza, 2002; UNFPA, 2002; Berliner, 2002), but little quantitative evidence of effectiveness has been published or disseminated. However, an ambitious qualitative and quantitative evaluation is underway involving a three-year panel/cohort study among more than 4500 young respondents and at least three rounds of data collection--in 2003, 2004, and 2005 (Guedes, 2004). This study will measure the program's success in changing attitudes and behavior around gender, stigma, HIV/AIDS and gender-based violence. Soul City has already conducted extensive population-based impact studies of Series 4, which addressed gender-based violence for the first time in 1999 (Scheepers, 2001; Scheepers and Cristophides, 2001; Soul City, 2001; UNFPA, 2002; Singhal et al., 2004). That research found that the series reached 82% of the population in South Africa and had a positive impact on both awareness and knowledge about violence against women. The impact on attitudes appeared to be mixed. Certain attitudes about violence improved slightly (such as whether violence is a private affair), but some stayed the same or deteriorated. For example, there were no changes in the levels of those who believed that men have the right to beat their wives, or in the belief that such beatings are socially acceptable. The evaluation focused on intermediate outcomes such as knowledge, attitudes, and intentions, but not on how these related to outcomes such as levels of actual violence. The program measured a greater impact on HIV/AIDS attitudes and behaviors compared to violence, but researchers pointed out that Soul City had addressed HIV/AIDS in multiple series over several years, while Series 4 was the first series to address violence against women. EDUCATION Background Education and violence are linked in multiple and complex ways. Researchers have found an inverted U- shaped relationship between levels of education and levels of intimate partner violence in the United States and South Africa (Jewkes, 2002c), where women with the most and least education appeared to be less vulnerable to violence compared to women with some education. Jewkes argues that small amounts of education may encourage women to challenge existing gender roles and provoke conflict in the household, without conferring the type of protection seen from higher educational attainment. Conversely, violence itself-particularly sexual harassment and abuse-appears to contribute to lower levels of education among girls relative to boys in many settings. For example, parents' fears for their daughters' physical and sexual safety appears to be a major reason for withholding girls from school in South Asia and Sub-Saharan Africa (Mensch and Lloyd, 1998; Sathar and Lloyd, 1993; UNICEF, 2004). In theory, educational reform could prevent gender-based violence by increasing school safety, by empowering women through education, and by promoting better attitudes and practices among students with regard to women's human rights. Unfortunately, a growing body of evidence suggests that sexual harassment is widespread in educational settings in many parts of the world (Mirsky, 2003; Leach et al., 2003; Wellesley Centers for Research on Women and DTS. 2003). Moreover, the response of schools to sexual violence and harassment is often notoriously poor. In a scathing report on sexual violence in South African and Zambian schools, Human Rights Watch (2001) documented a persistent pattern: " . . whereby schools discounted [girls'] . . . reports of sexual violence and harassment and failed to respond with any degree of seriousness. Girls were discouraged from reporting abuse to school officials for a variety of reasons, not least of which was the hostile and indifferent responses they received from their school communities." 31 At the same time, schools and universities have increasingly been asked to incorporate gender, rights and violence prevention in their curriculum. Fueled largely by concern about the HIV/AIDS pandemic, primary and secondary schools in most regions of the world now offer some kind of reproductive health education, often in the form of "life-skills" or "family life education." These courses often address gender issues if not gender-based violence directly (Birdthistle and Vince-Whitman, 1998). Meanwhile, researchers and advocates have called on universities to prepare the next generation of legal, medical, and social work professionals to address gender-based violence in their respective fields. Overview of education sector initiatives The challenge for the education sector is two-fold: to reduce discrimination and violence within the school setting and to strengthen the capacity of schools to promote nonviolence (see Table 2.3). Table 2.3. Examples of objectives and initiatives from the education sector Level Objectives Specific Initiatives Laws and To improve laws and policies · Nationalpoliciesandprograms toincreasefemaleschool policies · Increasepublicpolicycommitment to girls' enrolment and enrolment in schools and universities retention · Sexualharassmentlawsandpolicies foreducators,including · Strengthenlegalandadministrativesanctionsagainst enforcement mechanisms sexual harassment in schools · Inclusionofgenderandrightsinnationalschoolcurriculum · Increase attentiontogenderwithinnationaleducational policies curriculum Institutional To strengthen the institutional response of schools, · Institutionalsafety measures(e.g.all-femaleschools; more reform universities, school districts, ministries of education, female teachers, reducing the distances girls must travel to teachers unions, etc. school; separate latrines and/or canteens) · Improvethecapacityofschoolstopreventsexual · Inclusionof sexualharassmenttraininginteachertrainingand violence/harassment in school buildings and grounds certification requirements · Improvetheinstitutionalresponsetosexual · Institutionalcodesofconduct;administrativeenforcement harassment/violence when it occurs mechanism · Improvethecapacityof staff toteachviolenceprevention · In-servicesensitizationandtrainingofschoolpersonnel content, including gender and human rights · Inclusionofgender,rights,andviolencewithinthehealthand family life education and life-skills curriculum · School-basedcounselingandreferral services Community To increase community mobilization in support of girls' · Campaignstoaddresscommunityconcernsaboutgirls'safety Mobilization safety and rights and increase support for girls' education · Increaseparticipationof parentsin monitoringsafetyof · School-basedworkshops,theatre, andothereventstoraise school environment awareness of sexual harassment/violence among the broader · Increase collaborationbetweenschoolsandcommunity community services related to violence · Effortstostrengthenthecapacityofparentorganizationsto · Increase awarenessof sexualharassmentandviolence at monitor the performance of schools the community level · AlliancesandcollaborationbetweenschoolsandNGOs Individual To improve knowledge, attitudes and behaviors in · Inclusionofgender,rightsandviolencepreventioncurriculum behavior change support of girls' safety and rights in school or university coursework · Improvestudents' andparents'knowledge,attitudesand · Otherschool-basedviolencepreventionprograms (e.g. behaviors with regard to gender, rights and violence awareness campaigns, peer educators, role-playing exercises, theatre Reforming laws and policies of the education sector In many settings, the legal and policy framework for addressing sexual harassment/violence in schools is notoriously weak. Few low and middle-income countries have strong comprehensive, national, state or provincial laws and policies to sanction educators who violate codes of conduct. In almost every setting, only the most egregious cases result in criminal prosecution. Some countries have tried to address this issue. For example, after researchers documented widespread sexual harassment and rape in schools, the South African Department of Education established a task force to study the problem and make policy 32 recommendations (Jewkes, 2000). Their efforts contributed to the Employment of Educators Act and new Department of Education guidelines, both of which were introduced in 2000. These regulations mandate dismissal of educators found guilty of sexual or physical assault, or of having a sexual relationship with a student. They also define penalties for failing to report abuse. It remains to be seen whether these measures will have the intended impact. After the act was passed, Human Rights Watch (2001) suggested that the South African government needed to do more to increase awareness of the law among school principals and to strengthen enforcement. Institutional reform Efforts to improve the institutional response to gender-based violence range from sensitization and training of staff, sexual harassment policies, curriculum reform, school-wide anti-violence awareness campaigns, counseling and referrals, and broader efforts to reduce discrimination against girls and improve school safety. Initiatives to increase female enrolment by improving girls' safety at and on the way to school As mentioned earlier, parental concerns about girls' safety in school appears to lower female school enrolment in settings such as South Asia, Africa and the Middle East. Some initiatives have addressed these concerns by establishing single sex schools, hiring more female teachers, building separate latrines or canteens for girls, reducing the distance that girls must travel in order to receive an education, and/or providing in-service gender sensitivity training to teachers, principals and inspectors (UNICEF, 2004). For example, the UNICEF African Girls Education Initiative (AGEI) used a combination of these approaches, (along with other strategies) to boost girls' enrolment in 34 African countries (UNICEF, 2003a). Evidence of this project's effectiveness was limited in many sites, largely due to limitations in the evaluation design. While some demonstrated significant enrolment increases (for example, 15% in Guinea, 12% in Senegal, and 9% in Benin) in relatively short periods of time, the extent to which this was due to the project impact was not clear. Overall, however, the experience of this project suggests that addressing concerns about girls' safety and reducing the risk of sexual harassment and violence in schools is not only a high priority for parents, but also a potentially promising way to improve girls' access to education in selected settings (UNICEF, 2003b). Improving attitudes, knowledge, skills and practices of educators Many initiatives have aimed to improve educators' attitudes, knowledge and practices in regards to gender discrimination, sexual violence and sexual harassment. Few have been well documented or evaluated. A comparative study of HIV/AIDS education in three African countries found evidence that awareness and responses to sexual harassment in the Uganda sites was markedly better than those in Botswana and Malawi; researchers attributed this to the Ugandan government's efforts to curb sexual harassment in schools (Bennell, Hyde, and Swainson, 2002). The South African National Department of Education (in collaboration with international organizations) has developed a training module for educators (South African National Department of Education, 2001). Composed of eight interactive workshops and other materials, the module aims to increase educators' awareness of sexual harassment and gender violence, highlight the links between violence and HIV/AIDS and increase the safety of the school environment. The module is a professional development tool, rather than a part of the national curriculum. It has been field tested in some sites, and according to some reports is being rolled out nationwide. In other settings, schools have trained educators to teach courses promoting gender-equitable norms and nonviolence among students. For example, a consortium of researchers and advocates field-tested the "Gender and conflict" Model Curriculum in South Africa (Dreyer et al., 2001; Guedes, 2004) to compare a "whole school" approach (which trained the entire primary school staff, including principals and auxiliary staff) with a "trainer of trainers" approach (which trained two teachers from each school and 33 expected them to transfer their training to their colleagues). The evaluation produced two key findings virtually identical to those from other sectors. First, training all staff-the "whole school" approach was more effective than the "trainer of trainers" approach. In fact, none of the teachers in schools that used a "trainer of trainers" approach had transferred training to their colleagues due to time constraints. Second, effective training must address teachers' own experiences of abuse and perpetration. Almost half the female teachers reported having been physically abused by a male partner at some point in their lives; almost one-third reported experiences of sexual abuse; and several male teachers admitted to physically or sexually abusing their partners or colleagues. Expanding and improving school-based counseling and referral services A number of schools have tried to improve their response to sexual violence and harassment by providing counseling and referral services to students. For example, the TANESA "Guardian Project" aimed to improve girls' safety by designating one teacher from each of 185 primary schools as a "guardian" or mlezi (Mgalla et al., 1998; Mirsky, 2003; Guedes, 2004). Mlezi's were trained to counsel girls who experienced sexual violence or harassment, or who needed advice about other sexual and reproductive health issues. Evaluators assessed the results of the program through interviews with teachers, `guardians' and 1219 students in 40 schools with a guardian and 22 schools without. In the participating schools, over 61% of girls consulted the guardians during the first year. In control schools, not a single girl said that she would ever report sexual harassment by a teacher, compared to 52% of girls in schools with a mlezi. The program had less success in other areas; for example, it was not able to change teachers' negative attitudes towards girls who became pregnant or who contracted an STI. Community mobilization Mobilizing parents and community members to monitor school safety In theory, parent organizations and other community groups could pressure schools to ensure students' safety and enforce sexual harassment policies. Leach et al. (2003) investigated sexual abuse and harassment in schools in Zimbabwe, Ghana, and Malawi. They followed up their research with pilot interventions such as seminars and workshops with students, parents and teachers (in all three countries); workshops with government officials and NGOs (in Zimbabwe and Malawi); and community theatre in the context of a Durbar--a traditional meeting of community stakeholders that aims to raise awareness of a problem and find solutions (in Ghana). Researchers evaluated the results of these efforts using informal, qualitative methods and concluded that community workshops and theatre raised awareness of abuse, increased willingness of parents to report abuse, and allowed the community to confront the problem of abuse without putting individual girls at risk of retaliation. In one site, Leach and colleagues report that the initiative resulted in sanctions against the head of a school who had repeatedly sexually abused students. Individual behavior change strategies School-based sexual abuse prevention programs Beginning in the 1970s, schools and universities throughout the United States launched programs designed to prevent child sexual abuse, "dating violence," or other forms of sexual coercion among students. Typically these programs were based on the theory that children and young women could be taught to protect themselves by recognizing abuse, reporting threatening situations, and adopting safety precautions. Few low and middle-income countries have adopted this model, but the number of programs and the breadth of research on these initiatives make them worthy of mention. By 1992, these programs were so widespread in the United States that a national survey of young people (aged 10-16, n=2000) found that approximately two-thirds had participated in a sexual abuse prevention program through their school (Finkelhor and Dziuba-Leatherman, 1995). A large number of studies have evaluated the effectiveness of these programs (see reviews by Finkelhor and Strapko, 1992; Meyer and Stein 2000; 34 Chalk and King, 1998), some with rigorous scientific designs, control groups, extensive quantitative outcome measures, and long-term follow-up. This research indicates that while programs can improve participants' knowledge and willingness to report abuse, they do not appear to reduce victimization among participants compared to controls. Thus, many researchers and advocates argue against initiatives that focus primarily on increasing girls' ability to protect themselves and instead advocate for those initiatives that aim to change male norms and behaviors and to promote positive models of forming relationships (Chalk and King, 1998; Bolen, 2003). Violence prevention in the context of sexual and reproductive health education In middle and low-income countries, most gender-based violence prevention in schools has occurred in the context of reproductive health education (HIV/AIDS prevention, "family life education" or "life- skills" programs). These initiatives range from formal, stand-alone classes to material integrated into the broad school curriculum, peer education programs, and school-based awareness campaigns. They vary greatly in the extent to which they address gender-based violence directly; some merely discuss gender issues or sexual negotiation more generally. While school-based programs have successfully improved students' knowledge about HIVAIDS and other health issues in many settings, they have had a more limited impact on sexual behavior (Leach et al., 2003). In many cases, researchers suggest that weak programming may be to blame; school-based programs often suffer from poor content, lack of teacher training, resistance by teachers, an overemphasis on anatomy, cultural barriers, and parental objections to sex education. Little is known about the ability of these programs to improve violence-related attitudes or behaviors. One quasi-experimental study of life skills education among 9th graders in South Africa found that the program had only a minor impact on attitudes towards sexual coercion, in part because levels of reported attitudes condemning such behavior were already high at baseline. Nor is it clear that there is a link between reported attitudes and behavior (Magnani et al., 2003). Researchers have presented compelling evidence to suggest that HIV/AIDS prevention programs need to do a better job of addressing gender power imbalances to be effective in many settings. For example, researchers from the London School of Economics conducted a longitudinal, qualitative evaluation of a school-based peer education program in Zimbabwe where HIV prevalence was several times higher among girls than boys (Campbell and MacPhail, 2002). The evaluation concluded that the program's failure to address gender power imbalances had entrenched them further. Male peer educators bullied and harassed female counterparts and encouraged other boys to pressure girls for early and frequent sex. Researchers concluded that the program had failed to address sexual coercion, despite the evidence that it plays a substantial role in the spread of HIV in this area. MULTI-SECTORAL: ECONOMIC DEVELOPMENT, SOCIAL SERVICES AND PUBLIC SAFETY Background Throughout this review, we have emphasized the value of collaboration between justice, health, education and other sectors to ensure an effective and comprehensive response to gender-based violence. However, this section will address key areas that-almost by definition-require a multi-sectoral approach, namely, economic empowerment of women, social services for girls and women who experience gender-based violence, and infrastructure projects. Research highlights the links between social and economic factors and intimate partner violence against women. In many settings, women with high economic and social status appear to be somewhat protected, but the relationship between socio-economic status and violence is not necessarily linear (Jewkes, 2002b). Evidence suggests that violence against women may actually rise as women initially gain greater access 35 to social and economic opportunities and resources, and in some settings, women in the poorest households may also be somewhat protected from violence. Jewkes argues that social and economic empowerment appears to increase women's risk of violence in some settings by challenging traditional gender roles and increasing conflict in the household until, she writes, "a high enough level [of empowerment] has been reached for the protective effects to predominate." The World Health Organization World Report on Violence came to the same conclusion, arguing: Where women have a very low status, violence is not "needed" to enforce male authority . . . Partner violence is thus usually highest at the point where women begin to assume nontraditional roles or enter the workforce (Heise and Garcia Moreno, 2002, page 99). Short-term increases in violence against women do not necessarily mean that a program will fail to reduce violence in the long run. Nor should an increase in violence be used as an excuse to deny women access to social and economic resources. Under some conditions, however, increases in violence against women may be a temporary, but unavoidable byproduct of challenging traditional gender norms. Ideally, social and economic programs should find ways to mitigate a violent male response. While most sexual violence is perpetrated by persons known to the girl or woman, Jewkes, Sen and Garcia Moreno (2002) outline factors that can increase women's vulnerability to sexual violence outside of intimate partnerships, including poverty, the safety of the physical environment in the community, social norms about men's entitlement to sex, norms that place heavy responsibility on women for maintaining family "honor", norms about women's right to travel alone or work outside the home free of harassment, and armed conflict in which rape is sometimes used as a weapon of war. Researchers have increasingly begun to evaluate infrastructure projects--such as transportation, urban upgrading, water and sanitation-from a gender perspective, but little is known about the links between these projects and gender-based violence, and most information is anecdotal if not speculative. Overview of economic, social and public safety initiatives Table 2.4 presents examples of initiatives that may have the potential to reduce women's vulnerability to violence through multi-sectoral efforts such as economic development, increased opportunities for labor force participation, education, public safety, and access to comprehensive social services once physical or sexual violence begins. Table 2.4. Examples of multi-sectoral objectives and strategies Level Objectives Specific interventions Laws and To improve social and economic laws and policies · Revisedlawsandpoliciestoimprovewomen's economic rightsto policies · Increaseinvestmentinwomen'ssocialandeconomic property, inheritance, and labor force participation development · Multi-sectoral nationalorstateplans foraddressinggender-based · StrengtheneconomicandsocialpoliciesthataddressGBV violence · Ensureacomprehensiveserviceresponsetosurvivorsof · Policies mandatingcomprehensivesocial, medical andlegalservices gender-based violence for survivors of gender-based violence Institutional To improve the response and capacity of social service · Creationorstrengtheningof governmentofficesdedicatedtothe reform and economic development programs to address GBV advancement of women · Increase attentiontoviolenceagainstwomenby · Integrationofviolencepreventionintotheprogramming ofsocialand institutions devoted to social and economic development economic development projects (e.g. urban upgrading, microcredit) · Increase attentiontoviolenceinthedesignofurban · Efforts to expand,improve and coordinateservicesforsurvivors:(e.g. planning and transport projects counseling, shelters, victim advocacy, hotlines, women's support · Improvethequalityandcomprehensivenessofthesocial groups, children's services, legal aid, batterer treatment, etc.) service response to violence 36 Community To increase community mobilization in support of · Publicsafetyprograms (e.g. camerasurveillanceinpublicspaces, Mobilization women's safety, support, access to services and economic women-only latrines or buses, neighborhood watch programs, etc.) opportunities · Strengthencollaborationbetweenpublicsector · Networksoforganizationsworkingtoimprovetherights,safetyand wellbeing of women in general and survivors in particular institutions and civil society with regard to women's rights and safety · Community-based awarenesscampaigns(e.g. mass media,workshops, · Strengthencommunity-basedwomen'sorganizationsand community theatre, protests) networks · Increase communitysupportforsurvivorsandsocial sanctions against perpetrators Individual To improve individuals' knowledge, attitudes, behaviors · Behaviorchangecommunicationstrategies(e.g. mass media behavior and access to services/economic opportunities campaigns, community based workshops, community theatre) change · Encouragefamily membersandfriendstosupportand · Economicdevelopmentprograms aimedat women (e.g. micro-credit care for victims of violence programs, small business development, housing, etc.) · Increase women's awarenessof andaccesstosocial services and economic opportunities Reforming laws and policies As noted in the Justice sector, civil codes in many countries restrict women's legal rights. An important component of multi-sectoral approaches to women's socio-economic status is the effort to strengthen women's ability to exercise their legal rights, including ownership of property, inheritance, labor force participation and access to land reform and credit. International and national advocacy efforts have produced many recent improvements in this area. For example, the Center for Reproductive Rights (2001) notes positive trends in Africa that have eliminated land reform laws favoring men (Ethiopia) and expanded inheritance rights of female children and widows (Zambia). In Latin America, Guatemala recently granted women the rights to manage marital property and decide to work outside the home without their spouses' permission. Brazil, Mexico, and Jamaica have all recently considered or enacted laws to improve women's rights to inheritance and property (Center for Reproductive Law and Policy, 2000). In many parts of the world, however, legal reform has provoked a conservative backlash and a failure to implement the law-especially when revised laws counter long standing traditions of discrimination against women. For example, when interviewed by Human Rights Watch about changes to the law of inheritance in Zambia, one respondent noted: On paper, the new law of inheritance has been a wonderful breakthrough for women's rights. In practice, though, the earlier resentment at the passing of a law which usurps customary rights has blossomed into blatant disregard of statutory law (Human Rights Watch, 2002, page 58). Similar to legal reform in other areas, this research highlights the difficulty of implementing legislative reforms and the need to complement them with other types of initiatives, especially behavior change communication strategies. Institutional reforms Women's social and economic development has been a major focus of many programs over the past few decades. However, most low-income settings still have limited social services and economic programs for women relative to the size of their populations; staff members are often poorly prepared to address violence against women; and services tend to be fragmented and difficult to access. 37 Expanding social services for women and children In many countries, public and private institutions have worked to improve social services for women and children who experience violence. In some settings, NGOs and coalitions such as the Nicaraguan Network of Women against Violence have spearheaded these initiatives (Velzeboer et al., 2003). In other settings, governments have promoted institutional reforms by establishing ministries, departments or agencies devoted to the advancement of women, including Mexico, Jamaica, Guatemala, Bolivia, Peru (Center for Reproductive Laws and Policy, 2000); these agencies often work to strengthen comprehensive services for survivors of gender-based violence. For example, in El Salvador, the Salvadoran Institute for the Development of Women is a government agency that coordinates the "Program to Strengthen the Family" (Programa de Saneamiento de la Relación Familiar), a multi-sectoral effort among public and private institutions (Valdez, 1999). In some settings, such as Nicaragua and Costa Rica, coalitions of government agencies and NGOs develop National Plans to improve the network of services for women and children affected by violence (Velzeboer et al., 2003). Typically, these efforts aim to expand, improve and integrate services such as telephone hotlines, emergency shelters, police intervention, legal assistance, counseling services, psychological care, support groups, income generation programs, programs for batterers and child welfare services. A large literature exists on caring for survivors of physical and sexual abuse within specific professional disciplines, but research on the effectiveness, quality and impact of social service programs is scarce in low and middle-income countries. Most evaluations document numbers of persons served, services provided, and types of cases reported (Inter-American Development Bank, 2002). Some conduct informal studies of quality, by gathering information on service quality, client satisfaction, or qualitative perspectives of survivors. For example, see the evaluation of health, counseling, legal aid and women's support groups by health programs in Brazil, the Dominican Republic and Venezuela (Guedes et al., 2002). Ideally, researchers would conduct randomized, longitudinal studies, such as the study by Sullivan and Bybee (1999) that demonstrated the benefits and risks of shelter-based advocacy services in the United States. Their study measured outcomes such as levels of physical violence, psychological abuse, depression, quality of life, and social support. However, almost no such research has been done in developing countries. At a community level, social service initiatives accompanied by substantial outreach efforts have sometimes increased the proportion of women who know what services exist and where, as well as the numbers of women who seek help, but little scientific research has explored the impact of expanded social services on violence prevention. Batterer programs Increasingly, NGOs and governments have attempted to reduce violence against women by organizing treatment programs for batterers, aimed at changing their attitudes and behaviors. Most are run by NGOs, but they often depend on court- mandated attendance as an alternative to criminal sanctions. Most batterer programs have been carried out in high-income countries, but increasingly they have been implemented by developing country NGOs, such as the Instituto Noos in Brazil (White, Greene, and Murphy, 2003) and CORIAC in Mexico (Morrison and Biehl, 1999). Many studies have evaluated these programs' effectiveness in high-income countries, but most evaluations have been methodologically flawed. The only randomized controlled trial to date was carried out by the United States Navy, which found no reduction in abuse compared with controls (Dunford, 2000). Battered women often identify treatment or counseling for their husbands as a high priority (e.g. Ellsberg, 2001b), but it remains to be seen whether cost-effective strategies for changing batterer behavior can be found. Shelters Many researchers and advocates have called on governments and donors to invest in shelters for women who experience gender-based violence. Typically, these facilities offer emergency refuge as well as counseling, medical and legal assistance, job training, telephone hotlines, and other services. Most rigorous evaluation studies on the effectiveness and quality of shelters come from settings such as the 38 United States, where over 1200 were in operation as of 1998 (Chalk and King, 1998). In high-income settings, survivors consistently rated shelter services highly, relative to mixed reviews of other institutions such as health services, police, courts, etc. Many, however, question the feasibility and cost-effectiveness of shelters in middle and low-income settings, where they remain rare (Larrain, 1999). Alternatively, women's groups have organized informal strategies that help women and children find refuge among friends, family or volunteers. On the other hand, activists often question why victims should be forced to leave their homes and neighborhoods, rather than perpetrators; and, shelters may not be an option for women who have no hope of financial survival apart from their families (International Center for Research on Women, 2002). Middle-income countries often have more shelters than low-income countries; even so, funding constraints generally do not allow wide coverage. For example, Turkey has a population of over seventy million, but has only 14 shelters for women experiencing family violence (Amnesty International, 2004). Shelters may be particularly important in settings where "honor" killings are common, because informal refuge systems may not provide adequate protection for women at risk of being killed by their families. Unfortunately, in Jordan and elsewhere, officials have sometimes "protected" women in danger of family violence by putting them in prison (Human Rights Watch, 2000). Increasing attention to gender-based violence within micro-credit institutions Many governments and NGOs have turned to microcredit programs as a way to increase economic opportunities for women in low and middle-income countries. For many years, microcredit programs did not consider violence against women relevant to their work. This began to change after ethnographic research in the early 1990s reported that violence against women was pervasive in the Bangladesh communities where the Grameen Bank worked, and women who received loans were frequently left (temporarily) incapacitated by beatings (Schuler, Hashemi and Badal, 1998). At that time, researchers note, Bank staff virtually ignored the problem for fear of provoking conflict in the community. More recently, a growing number of micro-credit organizations have implemented institutional reforms by educating their staff about gender issues and including the issue of violence against women in their programming. For example, BRAC (an NGO in Bangladesh with a large micro-credit program) developed in-service training for staff, set up legal aid services, created libraries for information on gender and violence, and began to address gender throughout its work (Murshed, 1998). (More information about the impact of microcredit on individual women and men is presented in the Individual Behavior Change Strategies section below. Multi-sectoral initiatives in refugee settings Increasing evidence of sexual violence in refugee settings (see review by Dugan, Fowler and Bolton, 2000) prompted international agencies to review the way they designed refugee camps and distributed goods. The U.N Special Rapportuer on Violence Against Women, Radhika Coomaraswarmy has noted that "poorly lit camps, latrines at unsafe distances and lack of privacy" contribute to "hostile living conditions for women" (Marshall, 1995). UNHCR guidelines (1999) recommend the following: locate latrines, water points and fuel collection areas in accessible places; make special arrangements for housing unaccompanied women and girls; lock washing facilities; provide adequate lighting on paths used at night; provide security patrols; and avoid shared communal living space with unrelated families. UNHRCR guidelines also recognize the need to place women in charge of distributing goods so that men cannot pressure women to exchange sexual favors in return for vital necessities, such as food (UNHCR, 1991). These guidelines have been widely field-tested, albeit in informal ways. Since 1996, the International Rescue Committee has launched multi-sectoral gender-based violence programs in Tanzania, Republic of Congo, Guinea, Kosovo, East Timor, and Liberia. The Sierra Leone program exemplifies their approach; it includes: a) counseling, medical and legal services for survivors; b) teacher training in primary and secondary schools; c) support groups for women; d) awareness-raising 39 among law enforcement and health providers; e) mass media campaigns; f) lobbying for legal reform; and g) skill-building for women. Although they have not systematically evaluated the project, a consultant recently conducted an informal assessment by reviewing program documents and interviewing key informants. That assessment concluded that the program had increased access of survivors to free emergency medical care and forensic exams; increased awareness and sensitivity of staff; convinced local NGOs to address gender-based violence in their programming; allowed some women to start their own businesses; contributed to the successful prosecution of 10 cases of post-conflict sexual assault cases; and improved the response of local courts to survivors who bring cases (Guedes, 2004). Community mobilization Many initiatives have aimed to improve the response to violence against women by mobilizing the community through mass media campaigns, community outreach, community-based awareness raising campaigns, and service provision. Multi-sectoral initiatives to mobilize a better community response to gender-based violence One example of a multi-sectoral community-level prevention effort was developed by an NGO called CEDOVIP (Michau and Naker, 2003; Raising Voices, 2003). First tested in Kampala, Uganda, "Raising Voices" aims to promote change at the community level, through five phases, namely: 1) Conducting a Community Assessment to gather information on attitudes and beliefs about domestic violence, build relationships in the community and prepare staff for the project; 2) Raising Awareness within the community and professional sectors (i.e. social and health services, law enforcement, teachers, religious communities, etc.) of domestic violence and its negative consequences; 3) Building Networks of support within the community and professional sectors; 4) Integrating Action against domestic violence into everyday life and within institutions; and 5) Consolidating Programs and activities to ensure their sustainability, continued growth and progress. A preliminary qualitative evaluation suggested that this approach succeeded in changing male behavior because of reduced tolerance to violence by local councils, police and the community at large. For example, men reported fear of being publicly shamed by having their personal issues exposed. These findings are preliminary, however, and more extensive external evaluation is planned for 2004. Urban upgrading to improve the safety of public spaces In some settings, public and private institutions have tried to increase the safety of the physical environment. For example, in the City of Johannesburg (South Africa) the Metropolitan Police and private groups collaborated to install camera surveillance of high crime sites in the inner city (United Nations Habitat, n.d.). By 2000, this system included 360 cameras in key spots around the city. Camera surveillance in South Africa appears to have reduced crime levels, but may have simply displaced it elsewhere. Little information is available on specific outcomes related to women's safety, however. In general, few evaluations of public safety efforts exist, even in high-income countries. For example, a systematic literature review by the National Research Council (2004) found no published evaluations of the impact of urban upgrading on prevention of violence against women, even though the 1994 Violence Against Women Act allocated $25 million dollars for better lighting, emergency phones, and increased police in public areas in the U.S. Improving roads and transportation Researchers have only recently begun to evaluate the impact of transportation projects on women from a gender perspective. For example, case studies sponsored by the International Forum for Rural Transport and Development explored efforts to increase women's access to bicycles in India; the construction of roads in rural Nepal; and the transportation options of women in Bangladesh, where harassment on public buses is common (see review by Asian Development Bank, 2003). Case studies suggest that better transportation options for women can increase their mobility, safety, confidence, income-generating 40 opportunities, and, in some cases, access to health services. A more direct link between transport and violence would involve efforts to reduce harassment on public buses and trains. For example, all-women buses have been introduced in some settings, such as the "Lady Bus initiative" in Thailand (United Nations Center for Human Settlements, 2001). Similarly, women may benefit from projects that reduce the distance that girls and women travel along insecure routes to reach distant sources of water and firewood. Little information exists, however, about the effectiveness of these initiatives on improving women's safety. Water and Sanitation projects In many low-income cities, large proportions of the urban population lack access to adequate (or any) sanitary facilities, including, for example, nearly one-quarter to one-half of the population in Indian cities (United Nations Habitat, 2003). Burra, Patel, and Kerr (2003) profiled a relatively successful ten-year program to build community designed, built and managed toilet blocks in urban slum areas of various Indian cities (funded in part by the World Bank). Women's cooperatives played a key role, both in advocating for the toilets and by ensuring that they met women's needs. No information directly linked this project with specific outcomes related to gender-based violence. However, before the toilets were built, many women and men in urban slums were forced to defecate in outdoor public spaces. To obtain some privacy, many women waited until nightfall. Thus, the public toilets may have improved women's safety by eliminating the need to go out at night for sanitary reasons. Individual behavior change strategies Evaluations targeted at individual behavior change highlight the complexity of violence prevention. Changing and measuring reported attitudes and awareness appears to be much easier than changing or measuring violence-related behaviors. Mass media campaigns to increase awareness of women's rights Organizations (mostly NGOs) around the world have used mass-media campaigns to raise awareness of gender-based violence, promote nonviolent behavior and encourage women and men to be more supportive of their friends and family members who experience violence. These campaigns have included international campaigns (such as the 16 Days of Activism Against Gender Violence Campaign), national campaigns (such as the annual campaigns conducted by the Nicaraguan Network of Women Against Violence), and local campaigns conducted within individual communities. These campaigns often appear to successfully raise awareness and increase knowledge. For example, national surveys suggest that-at least partly due to the annual campaigns--women in Nicaragua are increasingly aware of their rights. A Demographic and Health Survey (DHS) found that more than two thirds of women believed that violence against women was never justified and that they should be able to refuse sex with their partner for any reason. While it is difficult to sort out the contribution of multiple initiatives on changes in awareness and attitudes, the DHS found that one quarter of the women surveyed were able to repeat one or more messages from the annual campaigns (Ellsberg, Liljestrand, and Winkvist, 1997; Rosales, 1999; Herrera, 2001). Once again, however, researchers have not yet been able to measure the impact of these initiatives on the levels of violence against women. Moreover, experiences from other fields suggest that while these types of mass media campaigns play an important role in awareness- raising, strategies such as `edutainment' are more likely to change behaviors over time (CDC, 2004). Multi-sectoral initiatives to reduce violence against women throughout the community Association Najdeh, an NGO in Lebanon, operates 26 service centers for Palestinian refugees that offer vocational training, literacy, preschools, and income generation programs. In 1999, they launched a domestic violence initiative using action research, staff training, community awareness-raising, and four counseling centers. Community-based baseline and follow-up surveys found some attitudes and beliefs improved over the course of the initiative, including attitudes about whether physical violence against 41 women is ever justified, whether women have a right to decide whom to marry, whether women have the right to accept work outside the home, and whether women could turn to Najdeh for help with domestic violence. Within three months of the surveys, the reported levels of violence increased slightly between baseline and follow-up (from 16% to 18%), as did the percent of respondents who said that they would opt for keeping domestic violence situations "at home" (Guedes, 2004). This evaluation highlighted the challenge of changing levels of violence in the short run as well as the difficulty of evaluating prevention initiatives and interpreting indicators of success or failure. Empowering women through microcredit Earlier, this paper discussed the need to increase attention to gender-based violence within micro-credit institutions. In addition, a substantial body of research has investigated the impact of micro-credit on violence against women at the individual level. Initial studies suggested that micro-credit could reduce women's vulnerability to violence by increasing their access to economic resources, by improving their social status in the household, and by making their lives more visible within the community (Schuler et al., 1996). However, subsequent multivariate analysis and qualitative research found that the link between micro-credit and violence was more complex (Hashemi, Schuler and Riley, 1996; Schuler, Hashemi and Badal, 1998). While participation in microcredit programs appeared to increase women's empowerment over time, levels of violence did not decline, and in some cases even rose. Ultimately, researchers concluded that--similar to other types of empowerment initiatives--micro-credit programs appear to work in two directions at once. On the one hand, they reduce women's vulnerability to violence by strengthening their access to resources and making women's lives more public; on the other hand, they may increase the risk of violence by challenging patriarchal norms and escalating conflict in the household. Some micro-credit programs are trying to reduce the potential risks of exacerbating violence associated with micro-credit. For example, RADAR (South Africa) has integrated HIV/AIDS and gender-based violence prevention into an existing microcredit program for women in poor rural communities. Since RADAR is designed as a prospective, randomized community intervention trial, it may--in the future--contribute to a richer understanding of how to provide the benefits of micro-credit while mitigating the risks (RADAR, n.d.). 42 CONCLUSIONS AND RECOMMENDATIONS The review highlights both the limits and progress made in building a knowledge base about effective-or at least promising-ways to reduce levels of intimate partner and sexual violence. In general, gender-based violence prevention has received far less attention than treatment for survivors. Empirical evidence about effective interventions is scarce, though numerous research projects are underway that may contribute to the scientific knowledge base in the near future. While recognizing the limited number of high quality studies on program effectiveness, this review has attempted to highlight emerging good practices. Table 2.5 summarizes some of the most promising approaches profiled, as well as typical challenges and pitfalls by sector. 43 Table 2.5. Potentially promising approaches to GBV prevention and typical pitfalls / problematic approaches by sector Potentially promising approaches Typical pitfalls and problematic approaches JUSTICE Pressure on governments to comply with international human rights agreements Failure to allocate resources for implementing new laws and policies Educating law enforcement and the public about new laws Failure to monitor and evaluate implementation or impact of laws/policies Broad investment in strengthening law enforcement response to GBV (protocols, Underinvestment in police and courts training, etc.) Long-term efforts to educate police/ judiciary on the implications of GBV Family courts that require survivors to attempt reconciliation with abusers Specific GBV related legal reforms (special stations, cells or courts, etc.) Separate women's police stations without broader law enforcement reforms Comprehensive medico-legal system reform (e.g. forensic nursing) Policies restricting collection of legal evidence to forensic physicians Networks and alliances between legal, social, and health organizations Failure of law enforcement to coordinate with social and medical services Efforts to publicize enforcement of laws that protect women's rights Lack of legal aid services for divorce and child support Reform of informal justice systems (e.g. traditional courts and councils) Practices that allow traditional authorities to block women's access to courts HEALTH Policies facilitating access to emergency contraception, STI/HIV prophylaxis and Government mandates without allocated funding safe abortion Policies clarifying providers' roles and responsibilities in cases of GBV Mandatory reporting requirements for adult survivors of GBV Broad institutional reforms to improve the health care response to GBV Routine screening for GBV without broader institutional reforms Networks and coalitions for research, referrals, advocacy and education Failure of the health sector to coordinate with other community services Community education that aims to improve awareness, knowledge, attitudes Health education programs that lack a human rights framework behaviors, and access to services related to GBV as a public health problem Reproductive / HIV education for youth that addresses gender and GBV Reproductive health programs that fail to address gender or sexual coercion Mass media 'entertainment-education programs' Campaigns that use "macho" imagery (e.g. for condom promotion) EDUCATION Implementation and enforcement of sexual harassment laws and policies Vague, un-enforced or non-existent national sexual harassment policies Improved school infrastructure (more rural schools; more female teachers; safe, Schools that ignore parents' concerns about girls' safety working latrines for girls, etc.) The "whole school" approach to educator training about GBV Educator training limited to a single session or of poor quality School-based counseling and referrals Schools with no links to external GBV services School-based programs focused on changing male gender norms Abuse prevention programs focused solely on girls School-based sexual and reproductive health programs that encourage a critical School-based health education focused primarily on anatomy consciousness about gender and violence MULTI-SECTORAL Laws/policies to enhance women's ability to exercise economic rights Policies that restrict women's economic rights or privilege men's access to economic opportunities National Plans for a comprehensive approach to gender-based violence Inadequately funded, implemented or evaluated national GBV plans Networks and coalitions for expanding social services Lack of investment in social services; lack of coordination Micro-credit and income-generating programs that integrate attention to GBV Micro-credit institutions that ignore the implications of GBV Multi-sectoral approaches to GBV services for refugee women and girls Poorly designed refugee camps (e.g. poorly planned lighting, security, male- controlled distribution systems and location of water and firewood) Attention to women's needs and priorities in transport/infrastructure projects Transport and infrastructure projects aimed solely at men's priorities Community mobilization and mass media campaigns to change attitudes and increase access to social services 44 Cross-cutting conclusions The sectoral structure of this review was designed to identify good practice approaches for policy makers, since they generally work within a specific sector. The specificity of this sectoral approach, however, should not obscure some important overarching conclusions. Of first and foremost importance are what Guedes (2004) identifies as "guiding principles" for work on gender-based violence: · Ensure that all programs and projects prioritize survivors' safety and autonomy · Employ a human rights perspective in order to explicitly challenge prevailing norms that make violence acceptable within a society · Ensure that interventions are culturally appropriate before transferring interventions from one cultural milieu to another Other important lessons learned include the importance of: · Employing a multi-sectoral approach. This paper reviewed initiatives from each individual sector separately; however, one consistent finding from all sectors is the need for collaboration between law enforcement, legal aid services, health care organizations, public health programs, educational institutions and agencies devoted to social services and economic development. This collaboration appears to be important not only for providing an integrated service response to women who experience violence, but also for developing effective strategies for reducing levels of violence against women in society. · Working at different levels. Similar to the point above, effective approaches to prevention generally require working at different levels: individual, community, institutional, and laws and policies. For example, simply changing the penal code may not be effective if law enforcement institutions remain weak, if communities resist changes in women's legal rights, and if individual women remain unaware of the laws or unable to access services. This review highlights similar examples of the need to work at each of these levels in every sector. · Creating partnerships between government and nongovernmental agencies. This review highlights many examples of the benefits derived from collaboration between government and civil society. Both of these groups have an essential role to play and are unlikely to change levels of violence working in isolation from one another. · Addressing norms, attitudes and beliefs at all levels of society. Attitudes that condone or tolerate violence against women and blame the victim are deeply entrenched throughout society in nearly all parts of the world-albeit to varying degrees. Changing these attitudes and beliefs is a challenging, long-term process that requires a sustained commitment by institutions providing services, as well as organizations with the capacity to harness mass media strategies. · Targeting young people. Evidence suggests that young people are more open to changing their views about the acceptability of violence than older adults. Youth-oriented education programs may represent one of the most important strategies for reducing violence against women in the long run. Similarly, schools and universities have an opportunity to improve the response of the next generation of professionals by integrating attention to violence into the training of lawyers, judges, physicians, nurses, psychologists, teachers, etc. This approach may ultimately be more effective than trying to change the attitudes and practices of experienced professionals (although both are needed). · Demonstrating the developmental impact of GBV. Gender-based violence is clearly a women's issue and an issue of fundamental human rights. But it is also a serious public health issue and an 45 important barrier to the socio-economic development of many countries. Rigorous research that documents the health and developmental impacts will attract new actors to the fight against gender- based violence, with a concomitant increase in the visibility of the issue and in the resources devoted to it. · Building the knowledge base through rigorous evaluation. The dearth of evidence about effective programs to address gender-based violence leaves policy makers and program managers without the ability to make informed decisions. Lack of data not only impedes evidence-based decision-making, but makes it more difficult to argue for allocating increased resources for preventing and responding to gender-based violence. Researchers and programmers need to collaborate on more rigorous evaluations-particularly in the area of prevention. Too often, gender-based violence initiatives have not been based on a clear hypothesis or theory about how their strategies may produce results; even fewer have tested their theories through baseline and follow-up data collection (much less control groups and longitudinal designs). Comparative advantage of multilateral institutions and bilateral donors Bilateral donors and multilateral institutions can play an important role in addressing gender-based violence in developing countries. First, international actors are in a unique position to encourage science- based evaluations of gender-based violence programs, to share the results of these evaluations across countries, and to use those findings to promote investment in effective prevention and treatment initiatives. In addition, research on the health and socioeconomic costs of gender-based violence may encourage governments to invest in prevention. Second, international actors can promote effective public-private partnerships and, in particular, coalitions between governments and NGOs. As this review notes, these types of coalitions appear to be essential for developing an effective community or nationwide effort to reduce violence and assist survivors. Third, this review highlights the value of integrating attention to violence into existing multi-sectoral programs and institutions devoted to justice, health, education, social services, and economic development. Within the individual sectors in which multilateral institutions and bilateral donors work, specific recommendations from this review (see Table 2.5) represent possible avenues of technical and financial support. Especially promising avenues include: a) Integrating actions to prevent gender-based violence and ensure survivors' access to justice in reforms of the judicial sector; b) Encouraging governments to revise their legislation in accordance with CEDAW and other international human rights agreements; c) Encouraging national and local government agencies to develop and enforce stronger sexual harassment policies for schools and universities; d) Including attention to sexual violence and harassment in initiatives to boost girls' school enrolment; e) Ensuring that attention to gender-based violence (particularly sexual coercion) is fully integrated into HIV/AIDS prevention programming and services; f) Promoting rigorous impact evaluations of gender-based violence initiatives where possible. The best hope for reducing worldwide levels of violence against women lies in mobilizing all levels of society-from international donors and national governments, to grassroots women's organizations, private firms and local governments. The challenge is not only to raise awareness of violence against women, but to maintain a long-run commitment by all these actors to address gender-based violence as an impediment to economic development, a public health problem and a violation of fundamental human rights. 46 REFERENCES Note: all websites were accessed in September 2004. Abrahams, N, Rachel Jewkes, R Laubsher. 1999. "I Do Not Believe in Democracy in the Home: Men's Relationships with Abuse of Women." Centre for Epidemiological Research in South Africa. Medical Research Council: Tyberberg. Abaunza, H. 2002. "Puntos de Encuentro: Communication for Development in Nicaragua." Sexual Health Exchange 2002-1. KIT Information Services, Royal Tropical Institute: Amsterdam American Medical Association. 1992. Diagnostic and Treatment Guidelines on Domestic Violence. Chicago. Amnesty International. 2004. Turkey: Women Confronting Family Violence. (EUR 44/013/2004) London. Available at: http://web.amnesty.org/library/index/engeur440132004 Asian Development Bank. 2003. "Assessing the Impact of Transport and Energy Infrastructure on Poverty Reduction." Available at: www.adb.org/Documents/Events/2001/RETA5947/draft_final_report.pdf Åsling-Monemi, Kajsa et al. 2003. "Violence Against Women Increases the Risk of Infant and Child Mortality: A Case-Referent Study in Nicaragua." The Bulletin of the World Health Organization 8:10-18. Barker, Gary. 2003. "How Do We Know if Men have Changed? Promoting and Measuring Attitude Change with Young Men. Lessons From Program H in Latin America." Paper Presented at the Expert Group Meeting on 'the Role of Men and Boys in Achieving Gender Equality'. United Nations: Brasilia, Brazil. Bennell Paul, Karin Hyde, and Nicola Swainson. 2002. The Impact of the HIV/AIDS Epidemic on the Education Sector in Sub-Saharan Africa: A synthesis of the findings and recommendations of three country studies. Centre for International Education, University of Sussex: Brighton. Berliner, D. 2002. "Nicaraguan Youth Empowerment Through Mass Media." Unpublished Summer 2002 Research Report for the Center for Latin American Studies, University of California: Berkeley. Billings, Debra et al. 2002. "Constructing Access to Legal Abortion Services in Mexico City." Reproductive Health Matters 10(19):86-94. Birdthistle, Isolde, and Cheryl Vince-Whitman. 1998. Reproductive Health Programs for Young Adults: School-Based Programs. Focus on Young Adults Research Series. Focus on Young Adults: Washington, D.C. Available at: www.pathfind.org/site/PageServer?pagename=Publications_FOCUS_Publications Blumel, D.K. et al., 1993. "Who Pays? The Economic Costs of Violence against Women." Women's Policy Unit, Office of the Cabinet: Queensland. Boender, Carol et al. 2004. The So What Report: A Look at Whether Integrating a Gender Focus into Programs Makes a Difference to Outcomes. Interagency Gender Working Group Task Force Report: Washington, D.C. Bolen, Rebecca. 2003. "Child Sexual Abuse: Prevention or Promotion?" Social Work 48(12):174-185. Bott, Sarah, Alessandra Guedes, and Ana Guezmes. (Forthcoming). "The Health Service Response to 47 Sexual Coercion/Violence: Lessons from IPPF/WHR Member Associations in Latin America." In: Jejeebhoy, Shireen, Iqbal Shah, and Shyam Thapa, Eds. Sexual Violence and Young People: Perspectives from the Developing World. Zed Publications: London. Bott, Sarah et al. (2004). Improving the Health Sector Response to Gender-Based Violence: A Resource Manual for Health Care Professionals in Developing Countries. International Planned Parenthood Federation, Western Hemisphere Region: New York. BRAC. n.d. Unpublished information available on BRAC's website. Available at: http://www.brac.net/ Brown, A. Widney. 2001. "Obstacles to Women Accessing Forensic Medical Exams in Cases of Sexual Violence." Unpublished background paper to the Consultation on the Health Sector Response to Sexual Violence, WHO Headquarters, Departments of Injuries and Violence Prevention and Gender and Women's Health, Geneva, June 2001. Available at: www.hrw.org/backgrounder/wrd/who-bck.pdf Buel, Sara. 2001. "Treatment Guidelines for Healthcare Providers' Interventions with Domestic Violence Victims: Experience From the USA." International Journal of Gynecology & Obstetrics 78(Supplement No. 1): S39-S44. Burra, Sundar, Sheela Patel, and Thomas Kerr. 2003. "Community-Designed, Built and Managed Toilet Blocks in Indian Cities." Environment & Urbanization 15(2) October: 11-32. Buvinic, Mayra, and Andrew Morrison. 1999. Technical Notes on Violence Prevention. Inter-American Development Bank, Sustainable Development Department: Washington, D.C. Buvinic, Mayra, Andrew Morrison, and Michael Shifter. 1999. "Violence in the Americas: A Framework for Action." In: Morrison, Andrew, and María Loreto Biehl, Eds. Too Close to Home: Domestic Violence in the Americas. Inter-American Development Bank: Washington, D.C. Campbell, Jacqueline et al. 2002. "Health Consequences of Intimate Partner Violence." The Lancet 359(9314):1331-1336. Campbell, Jacqueline. 2000. "Promise and Perils of Surveillance in Addressing Violence Against Women." Violence Against Women 6(7):705-727. Campbell, C, and C MacPhail. 2002. "Peer Education, Gender and the Development of Critical Consciousness: Participatory HIV Prevention by South African Youth." Social Science and Medicine 55:331-345. CDC. 2003. "Costs of Intimate Partner Violence against Women in the United States." Centers for Disease Control and Prevention, National Center for Injury Prevention and Control: Atlanta, Georgia. CDC. 2004. "Behavior Change Communications" In: CDC. Global AIDS Program Strategies. Centers for Disease Control and Prevention National Center for HIV, STD and TB Prevention Global AIDS Program: Atlanta. Available at: www.cdc.gov/nchstp/od/gap/strategies/2_7_bcc.htm Center for Reproductive Law and Policy. 2000. Women of the World: Laws and Policies Affecting Their Reproductive Lives. Latin America and the Caribbean, Progress Report. New York. Available at: http://bookstore.reproductiverights.org/womofworlawa2.html Center for Reproductive Law and Policy. 2001. Women of the World: Laws and Policies Affecting their Reproductive Lives. Anglophone Africa. Progress Report. New York. Available at: 48 http://bookstore.reproductiverights.org/womofworlawa.html Center for Reproductive Law and Policy. 2002. Bringing Rights to Bear: An Analysis of the Work of the UN Treaty Monitoring Bodies on Reproductive and Sexual Rights. New York. Available at http://bookstore.reproductiverights.org/brinrigtobea.html Chalk, Rosemary, and Patricia King, Eds. 1998. Violence in Families: Assessing Prevention and Treatment Programs. National Research Council and the Institution of Medicine. National Academy Press: Washington, D.C. Counts, Dorothy Ayers, Judith Brown, and Jacqueline Campbell. 1992. Sanctions and Sanctuary: Cultural Perspectives on the Beating of Wives. Westview Press: Boulder, Colorado. Dahlberg, L. 1998. "Youth Violence in the United States: Major Trends, Risk Factors and Prevention Approaches." American Journal of Preventive Medicine 14(4): 259-72. d'Oliveira, A., and L. B. Schraiber. 2001. "Violence Against Women and Brazilian Health Care Policies: A Proposal for Integrated Care in Primary Care Services." International Journal of Gynecology & Obstetrics 78(Supplement Number 1):S21-S26. Dreyer, A et al. 2001. "What Do We Want to Tell Our Children About Violence Against Women? Evaluation Report for the Project Developing a Model `Gender and Conflict' Component of the Primary School Curriculum." School of Public Health, University of the Western Cape: South Africa. Cited in Mirsky, Judith. 2003. Beyond Victims and Villains: Addressing Sexual Violence in the Education Sector. The Panos Institute: London. Dube, S. et al. 2001. "Childhood Abuse, Household Dysfunction, and the Risk of Attempted Suicide throughout the Life Span: Findings from the Adverse Childhood Experiences Study." Journal of the American Medical Association 286(24): 3089-96. Dugan, Julie, Carolyn Fowler, and Paul Bolton. 2000. "Assessing the Opportunity for Sexual Violence Against Women and Children in Refugee Camps." The Journal of Humanitarian Assistance. Available at: www.jha.ac/articles/a060.htm Dunne, Michael et al. 2003. "Is Child Abuse Declining? Evidence from a Population Based Survey of Men and Women in Australia." Child Abuse and Neglect 27:141-152. Ellsberg, Mary et al. 1999. "Domestic Violence and Emotional Distress Among Nicaraguan Women: Results from a Population-Based Study." American Psychologist, 54: 30-36. Ellsberg, Mary et al. 2000. "Candies in Hell: Women's Experiences of Violence in Nicaragua." Social Science and Medicine 51(11):1,595-1610. Ellsberg, Mary et al. 2001a. "Researching Violence Against Women: Methodological and Ethical Considerations." Studies in Family Planning 32(1):1-16. Ellsberg, Mary et al. 2001b. "Women's Strategic Responses to Violence in Nicaragua." Journal of Epidemiologiy and Community Health 55:547-555. Ellsberg, Mary et al. (Forthcoming). Researching Violence Against Women: A Practical Guide for Researchers and Advocates. World Health Organization, Center for Health and Gender Equity: Washington, D.C. 49 Ellsberg, Mary, Jerke Liljestrand, and Anna Winkvist. 1997. "The Nicaraguan Network of Women Against Violence: Using Research and Action for Change." Reproductive Health Matters 10:82-92. EngenderHealth. 2003 (unpublished). "MAP Evaluation Report." New York. Felitti, VJ et al. 1998. "Relationship of Childhood Abuse and Household Dysfunction to Many of the Leading Causes of Death in Adults. The Adverse Childhood Experiences (ACE) Study." American Journal of Preventive Medicine 14(4): 245-258. Ferrando, Delicia, Nery Serrano, and Carlos Pure. 2002. "Perú: Salud Reproductiva en Comunidades. Educando y Empoderando a Mujeres de Escasos Recursos: Evaluación de Impacto de Medio Término del Proyecto ReproSalud." Unpublished report, cited in: Boender, Carol et al. 2004. The So What Report: A Look at Whether Integrating a Gender Focus into Programs Makes a Difference to Outcomes. Interagency Gender Working Group Task Force Report: Washington, D.C. Finkelhor, David, and Nancy Strapko. 1992. "Sexual Abuse Prevention Education: A Review of Evaluation Studies." In Willis, Diane J., E.Wayne Holden, and Mindy Rosenberg, Eds. Prevention of Child Maltreatment: Developmental and Ecological Perspectives, pages 150-167. John Wiley and Sons: New York. Finkelhor, David, and Jennifer Dziuba-Leatherman. 1995a. "Victimization Prevention Programs: A National Survey of Children's Exposure and Reactions." Child Abuse & Neglect 19:129-139. Finkelhor, David, and Jennifer Dziuba-Leatherman. 1995b. "The Effectiveness of Victimization Prevention Instruction: an Evaluation of Children's Responses to Actual Threats and Assaults." Child Abuse and Neglect 10:141-153. Garcia-Moreno, Claudia. 2002. "Dilemmas and Opportunities for an Appropriate Health-Service Response to Violence Against Women." The Lancet 359:1509-14. Garcia Moreno, Claudia et al. 2003. "Responding to Violence Against Women: WHO's Multi-Country Study on Women's Health and Domestic Violence." Health and Human Rights 6(2): 112-127. Gazmararian, JA et al. 1995. "The Relationship Between Intendedness and Physical Violence in Mothers of Newborns." Obstetrics and Gynecology, 85:131-138. Gillegan, James. 2000. "Violence in Public Health and Preventive Medicine." The Lancet 355:1802-1804. Godenzi, A. and C. Yodanis. 1998. "Report on the Economic Costs of Violence against Women." University of Fribourg: Fribourg. Greaves, L. et al. 1995. "Selected Estimates of the Costs of Violence against Women." Centre for Research on Violence against Women and Children: London: Ontario. Grisourapong, S. 2002. "Establishing a One-stop Crisis Center for Women Suffering Violence in Khonkaen Hospital, Thailand." International Journal of Gynecology and Obstetrics 78(Supplement 1):S27-S38. Guedes, Alessandra. 2004. "Addressing Gender-based Violence from the Reproductive Health/HIV Sector: A Literature Review and Analysis." USAID, Bureau for Global Health: Washington, D.C. Available at: www.prb.org/pdf04/AddressGendrBasedViolence.pdf 50 Guedes, Alessandra et al. 2002. "Gender-based violence, Human Rights, and the Health Sector: Lessons from Latin America." Health and Human Rights 6(1):177-194. Guedes, Alessandra, Sarah Bott, and Yvette Cuca. 2002. "Integrating Systematic Screening for Gender- based Violence into Sexual and Reproductive Health Services: Results of a Baseline Study by the International Planned Parenthood Federation, Western Hemisphere Region." International Journal of Gynecology and Obstetrics 78: 557-563. Hakimi, Mohammad et al. 2001. Silence for the Sake of Harmony. Domestic Violence and Women's Health in Central Java, Indonesia. Yogyakarta, Indonesia. CHN-RL GMU. xvi. Haque, Y, and J. Clarke. 2001. "The Woman Friendly Hospital Initiative in a Bangladesh Setting: Standards for the Care of Women Subject to Violence." International Journal of Gynecology & Obstetrics 78(Supplement No 1): S45-S50. Hashemi, Syed, Sidney Ruth Schuler, and Ann Riley. 1996. "Rural Credit Programmes and Women's Empowerment in Bangladesh." World Development 24(4):635-653. Heise, Lori, and Claudia Garcia-Moreno. 2002. "Violence by Intimate Partners." In: Krug, Etienne et al., Eds. World Report on Violence and Health. World Health Organization: Geneva. Pages 89-121. Heise, Lori, Jacqueline Pitanguy, and Adrienne Germain. 1994. "Violence against Women: The Hidden Health Burden." World Bank Discussion Paper #255. World Bank: Washington, D.C. Heise, Lori, Kirsten Moore, and Nahid Toubia. 1995. Sexual Coercion and Reproductive Health: A Focus on Research. The Population Council: New York. Heise, Lori, Mary Ellsberg, and Megan Gottemoeller. 1999. "Ending Violence Against Women," Population Reports, Volume XXVII, Number 4, Series L, Number 11. Available at: http://www.infoforhealth.org/pr/l11/violence.pdf Heise, Lori. 1998. "Violence Against Women: an Integrated Ecological Framework." Violence Against Women 4(3):262-290. Hernández Rosete, D. 1998. "Pobreza Urbana y Violencia Doméstica en Hogares de la Ciudad de México." Acta Sociológica (enero-abril): 25-43. Herrera, M. 2001. "Women's Network Against Violence, Nicaragua." Paper presented at the Gender Violence, Health and Rights in the Americas Symposium in Cancun, Q.R., Mexico, June 4-7, 2001. Human Rights Watch. 1997. South Africa: Violence Against Women and the Medico-Legal System. New York. ----. 1999. Crime or Custom? Violence Against Women in Pakistan. New York. ----. 2000. What will it take? Stopping Violence Against Women: a Challenge to Governments. New York. ----. 2001. Sacrificing Women to Save the Family? Domestic Violence in Uzbekistan. New York. ----. 2001. Scared at School: Sexual Violence Against Girls in South African Schools. New York. ----. 2002. Suffering in Silence: the Links Between Human Rights Abuses and HIV Transmission to Girls 51 in Zambia. New York. ----. 2003a. Uzbekistan: From House to House. Abuses by Mahalla Committees. New York. ----. 2003b. Just Die Quietly. Domestic Violence and Women's Vulnerability to HIV in Uganda. New York. ILANUD. n.d. The United Nations Latin American Institute for the Prevention of Crime and the Treatment of Offenders. ILANUD Activities in 2002 and 2003 Work Programme. San José, Costa Rica. Inter-American Development Bank. 2002. Inventario de Programas de Atención y Prevención de la Violencia a Nivel Municipal. [Violence Prevention Programs Database]. Electronic document. Washington, D.C. Available in Spanish only at: www.iadb.org/sds/publication/publication_1535_s.htm International Center for Research on Women. 2002. Domestic Violence in India, Number 5: Women- Initiated Community Level Responses to Domestic Violence, Summary Report of Three Studies. Washington, D.C. Available at: www.icrw.org/docs/DVIndia_Report5_702.pdf Jacobs, Tanya, and Rachel Jewkes. 2001. "Vezimfilho: A Model for Health Sector Response to Gender Violence in South Africa." International Journal of Gynecology & Obstetrics 78(Supplement No 1): S51- S56. Jaffe, P, and M Sudermann. 1995. "Child Witness of Women Abuse: Research and Community Responses." In: Stith S, and M Straus, Eds. Understanding Partner Violence: Prevalence, Causes, Consequences, and Solutions. Families in Focus Services, Vol. II. National Council on Family Relations: Minneapolis. Jejeebhoy, Shireen, and Sarah Bott. 2003. "Non-consensual Sexual Experiences of Young People: A Review of the Evidence from Developing Countries." Regional Working Paper #16. Population Council, Regional Office for South and East Asia: New Delhi. Available at: www.popcouncil.org/pdfs/wp/seasia/seawp16.pdf Jewkes, Rachel. 2000. The HIV/AIDS Emergency: Department of Education Guidelines for Educators. Department of Education: Pretoria, South Africa. Jewkes, Rachel, Purna Sen, and Claudia Garcia-Moreno. 2002. "Sexual Violence." In: Krug, Etienne, et al, Eds. World Report on Violence and Health. World Health Organization: Geneva, pages 149-181. Jewkes, Rachel. 2002. "The Epidemiology of Rape and Sexual Coercion in South Africa: an Overview." Social Science and Medicine 55(7):1231-44. Jewkes, Rachel. 2002a. "Preventing Domestic Violence." Editorial. British Medical Journal 324:253-254. Jewkes, Rachel. 2002b. "Intimate Partner Violence: Causes and Prevention." The Lancet 359:1423-1429. Joint United Nations Programme on HIV/AIDS. 2000. National AIDS Programmes: A Guide to Monitoring and Evaluation. UNAIDS: Geneva. Jubb, N, and WP Izumino. 2002. "Women and Policing in Latin America." A Revised Background Paper Prepared for the Women and Policing in Latin America Project. Kerr, R, and J. McLean. 1996. "Paying for Violence: Some of the Costs of Violence Against Women in 52 B.C." Ministry of Women's Equality: British Colombia. Kim, Julia. 2000. "Rape and HIV Post Exposure Prophylaxis: The Relevance and the Reality in South Africa." Discussion Paper presented at the World Health Organization Meeting on Violence Against Women and HIV/AIDS: Setting the Research Agenda. Geneva, 23-25, October 2000. Kim, Julia and Mmatshilo Motsei. 2002. " `Women Enjoy Punishment:' Attitudes and Experiences of Gender-Based Violence among Primary Health Care Nurses in Rural South Africa." Social Science and Medicine 54(8): 1243-54. Kishor, Sunita, and Kiersten Johnson. 2004. Profiling Domestic Violence - A Multi-Country Study. ORC Macro: Calverton, Maryland. Koenig, Michael et al. 2003. "Women's Status and Domestic Violence in Rural Bangladesh: Individual- and Community-Level Effects." Demography 40(2): 269-88. Korf, D.J. et al. 1997. "Economic Costs of Domestic Violence against Women." Dutch Foundation of Women's Shelters: Utrecht, Netherlands. Koss, M.P. 1993."Detecting the Scope of Rape: A Review of Prevalence Research Methods." Journal of Interpersonal Violence 8(2):198-222. Krug, Etienne et al., Eds. 2002. World Report on Violence and Health. World Health Organization: Geneva. Larraín, Soledad. 1999. "Curbing Domestic Violence: Two Decades of Action." In: Morrison, Andrew, and María Loreto Biehl, Eds. Too Close to Home: Domestic Violence in the Americas. Inter-American Bank: Washington, D.C., pages 105-130. Larraín, Soledad et al. 1997. Relaciones Familiares y Maltrato Infantil. UNICEF: Santiago, Chile. Laurence, L., and R. Spalter-Roth. 1995. "Measuring the Costs of Domestic Violence Against Women and the Cost-Effectiveness of Interventions: An Initial Assessment and Proposals for Further Research." Institute for Women's Policy Research: Washington, D.C. Leach, Fiona et al. 2003. An Investigative Study of the Abuse of Girls in African Schools. Policy Division of the Department for International Development: London. Levack, A. 2001. "Educating Men in South Africa on Gender Issues." Unpublished report. SIECUS: New York. Levinson, David. 1989. Family Violence in Cross-Cultural Perspective. Sage Publications: Thousand Oaks, California. Levinson, David. 1989. "Family Violence in Cross-Cultural Perspective." In: Bernard, H. Russell, Ed. Frontiers of Anthropology. Vol 1. Pages 435-455. Sage Publications: Newbury Park, California. Lozano Ascencio, Rafael. 1999. "The Impacts of Domestic Violence on Health: Mexico City." In: Morrison, Andrew, and María Loreto Biehl, Eds. Too Close to Home: Domestic Violence in the Americas. Inter-American Development Bank: Washington, D.C. Magnani, Bob et al. 2003. "The Impact of Life Skills Education on Adolescent Sexual Risk Behavior." 53 Horizons Research Summary. Population Council: Washington, D.C. Marshall, Ruth. 1995. "Refugees, Feminine Plural." Refugees Magazine 100. UNHCR: Geneva. McLeer, S.V. et al. 1989. "Education is Not Enough: A Systems Failure in Protecting Battered Women." Annals of Emergency Medicine 18(6): 651-653. McPhedran, Marilou et al. 2000. The First CEDAW Impact Study: Convention on the Elimination of All Forms of Discrimination Against Women. Final Report. Center for Feminist Research and the International Women's Rights Project, York University: Toronto, Canada. Medical Research Council. 2003. "South African Gender-Based Violence and Health Initiative: Final Report to the Rockefeller Foundation." Medical Research Council: Pretoria, South Africa. Mehrotra, Aparna. 1998. Gender and Legislation in Latin America and the Caribbean. United Nations Development Programme, Regional Bureau for Latin America and the Caribbean: New York. Mensch, Barbara, and Cynthia Lloyd. 1998. "Gender Differences in the Schooling Experiences of Adolescents in Low-Income Countries: The Case of Kenya." Studies in Family Planning 20(2):167-184. Mensch, Barbara, Judith Bruce, and Margaret E. Greene. 1998. The Uncharted Passage: Girls' Adolescence in the Developing World. Population Council: New York. Available at: www.popcouncil.org/pdfs/passage/passage.pdf Mesquita da Rocha, Martha. 1999. "Dealing with Crimes against Women: Brazil." In: Morrison, Andrew, and María Loreto Biehl, Eds. Too Close to Home: Domestic Violence in the Americas. Inter- American Bank: Washington, D.C. Meyer, Heather, and Nan Stein. 2000. "Review of Teen Dating violence Prevention." [Electronic publication.] National Violence Against Women Prevention Research Center, Wellesley Centers for Women, Wellesley College. Available at: www.musc.edu/vawprevention/research/teendating.shtml Mgalla, Z et al. 1998. "Protecting School Girls Against Sexual Exploitation: A Guardian Programme in Mwanza Tanzania." Reproductive Health Matters 7(12):19-30. Michau, Lori, and Dipak Naker. 2003. Mobilization Communities to Prevent Domestic Violence: A Resource Guide for Organizations in East and Southern Africa. Raising Voices: Nairobi, Kenya. Mirsky, Judith. 2003. Beyond Victims and Villains: Addressing Sexual Violence in the Education Sector. The Panos Institute: London. Mitra, Nishi. 1998. Best Practices Among Responses to Domestic Violence: A Study of Government and Nongovernment Responses in Madhya Pradesh and Maharashtra. International Center for Research on Women: Washington, D.C. Molnar, B. et al. 2001. "Child Sexual Abuse and Subsequent Pathology: Results from the National Comorbidity Survey." American Journal of Public Health 91(5): 753-60. Morley, Rebecca, and Audrey Mullender. 1994. Preventing Domestic Violence to Women. Police Research Group Crime Prevention Unit Series: Paper Number 48. Home Office Police Department: London. Available at: www.homeoffice.gov.uk/rds/prgpdfs/fcpu48.pdf 54 Morrison, Andrew, and María Beatriz Orlando. 1999. "Socioeconomic Costs of Domestic Violence: Chile and Nicaragua. " In: Morrison, Andrew, and María Loreto Biehl, Eds. Too Close to Home: Domestic Violence in the Americas. Inter-American Development Bank: Washington, D.C. Muirhead, Debbie, Lilani Kumaranayake, and Charlotte Watts. 2001. "Economically Evaluating the 4th Soul City Series: Costs and Impact on HIV/AIDS and Violence Against Women." Institute for Health and Development Communication and the London School of Hygiene and Tropical Medicine: South Africa and London. Murshed, Rehana. 1998. "Gender Intervention as Applied in BRAC's Organisation and Programmes" Women in Agriculture and Modern Communication Technology - Proceedings of a Workshop. March 30 - April 3, 1998, Tune Landboskole, Denmark. Available at: www.husdyr.kvl.dk/htm/php/tune98/1- Rehana.htm National Research Council. 2004. Advancing the Federal Research Agenda on Violence Against women. National Academy Press: Washington, D.C. Nayak, Madhabika B., et al. 2003. "Attitudes Toward Violence Against Women: a Cross-Nation Study." Sex Roles 49: 333-342. Office of Women's Policy, Northern Territory Government. 1996. "The Financial and Economic Costs of Domestic Violence in the Northern Territory." KPMG Management Consulting: Northern Territory, Australia.. Owen, Bruce, and Jorge Portillo. 2003. "Legal Reform, Externalities and Economic Development: Measuring the Impact of Legal Aid on Poor Women in Ecuador." Discussion Paper Number 02-32, Stanford Institute for Economic Policy Research: California. Parenzee, Penny. 2001. "While Women Wait . . . Monitoring the Domestic Violence Act." Nedbank ISS Crime Index 5(3) May-June. Pence, Ellen. 1995. "Criminal Justice Response to Domestic Assault Cases: A Guide For Policy Development." Domestic Abuse Intervention Project: Duluth, Minnesota. Raising Voices. 2003. "Impact Assessment. Mobilising Communities to prevent domestic violence." Kampala, Uganda. Ramsay, Jean et al. 2002. "Should Health Professionals Screen Women for Domestic Violence? Systematic Review." British Medical Journal 325:314-326. Rashid, Maria. 2001. Giving Men Choices: A Rozan Project with the Police Force in Pakistan. Working Paper Series on Men's Roles and Responsibilities in Ending Gender Based Violence. Working Paper Number 2. Rozan: Islamabad. Rennison, Callie Marie. 2003. "Intimate Partner Violence. 1993-2001." Bureau of Justice Statistics, Crime Data Brief. United States Department of Justice: Washington DC. Renton, Linnea et al. 2000. Safe Crossing: The Stepping Stones Approach to Involving Men in the Prevention of Violence and HIV Transmission. Action Aid: London. Rogow, Debra, and Judith Bruce. 2000. "Alone You Are nobody, Together We Float: the Manuela Ramos Movement." Calité/Calidad/Quality, Number 10. Population Council: New York. Available at: 55 www.popcouncil.org/publications/qcq/qcq10.pdf Rosales, J et al. 1999. Encuesta Nicaragüense de Demografía y Salud, 1998. Instituto Nacional de Estadísticas y Censos: Managua, Nicaragua. Sadasivam, Bharati. 2000. "Community Justice: West Bengal's Women Draw on Village Tradition to Stop Domestic Violence." Ford Foundation Report, Special Issue on Women Now Its a Global Movement. Winter Issue 31(1). Sánchez, Fabio et al. 2004. "Los Costos de la Violencia Intrafamiliar en Colombia." Universidad de los Andes, Centro de Estudios sobre Desarrollo Económico: Bogotá, Colombia. Sathar, Zeba, and Cynthia Lloyd. 1993. "Who Gets Primary Schooling in Pakistan: Inequalities Among and Within Families." Research Division Working Paper Number 52. Population Council: New York. Scheepers, Esca. 2001. Soul City 4 Impact Evaluation - Violence Against Women Volume I. Houghton, South Africa, Institute for Health and Development Communication. Soul City: South Africa. Available at: www.soulcity.org.za/downloads/SC4%20VAW%20Volume%201.pdf Scheepers, Esca, and Nicola Cristophides. 2001. Soul City 4 Impact Evaluation - Violence Against Women Volume II. Houghton, South Africa, Institute for Health and Development Communication. Soul City: South Africa. www.soulcity.org.za/downloads/SC4%20VAW%20Volume%202.pdf Schuler, Sidney Ruth, Syed Hashemi, and Shamsul Huda Badal. 1998. "Men's Violence Against Women in Rural Bangladesh: Undermined or Exacerbated by Microcredit Programmes?" Development in Practice 8(2):148-157. Schuler, Sidney Ruth et al. 1996. "Credit Programs, Patriarchy and Men's Violence Against Women in Rural Bangladesh." Social Science and Medicine 43(12):1729-1742. Shane, Barbara, and Mary Ellsberg. 2002. "Violence Against Women: Effects on Reproductive Health." Outlook 20(1). Program for Appropriate Technology in Health: Washington, D.C. Available at: www.path.org/files/EOL20_1.pdf Shaw, Matthew, and Michelle Jawo. 2000. "Gambian Experiences with Stepping Stones: 1996-1999." Participatory Learning and Action Notes 37. Pages 73-78. Available at: www.iied.org/sarl/pla_notes/pla_backissues/documents/plan_03714.pdf Shaw, Matthew. 2001. Stepping Stones: The Gambia Adaptation. United Kingdom Medical Research Council: The Gambia. Shaw, Mathew. 2002a. "A Qualitative Evaluation of the Impact of the Stepping Stones Sexual Health Programme on Domestic Violence and Relationship Power in Rural Gambia. Unpublished paper presented at the 6th Global Forum for Health Research: Arusha, Tanzania. Shaw, Mathew. 2002b. " 'Before We Were Sleeping But Now We Are Awake': The Stepping Stones Workshop Programme in the Gambia." In: Cornwall, Andrea and Alice Welbourn, Eds. Realising Rights: Transforming Approaches to Sexual and Reproductive Well-being. Zed Books: London. Pages: 128-140. Shepard, Melanie, and Ellen Pence, Eds. 2000. Coordinating Community Responses to Domestic Violence: Lessons from Duluth and Beyond. Sage Publications: Thousand Oaks, California. 56 Shramajibee Mahila Samity. 2003. "Shalishi in West Bengal: a Community-Based Response to Domestic Violence." Economic and Political Weekly. Review of Women Studies, April 26:1665-1673. Singhal, Arvind et al. 2004. "Harnessing the Entertainment-Education Strategy in Africa: The Soul City Intervention in South Africa." In Okigbo, Charles, Ed. Development and Communication in Africa." Rowman and Littlefield: Boston, Massachusetts. Snively, Suzanne. 1994. "The New Zealand Economic Cost of Family Violence." Coopers and Lybrand: Wellington, New Zealand. Sorenson, Susan. 2003. "Funding Public Health: The Public's Willingness to Pay for Domestic Violence Prevention Programming." American Journal of Public Health 93: 1934-38. South African National Department of Education. 2001. "Opening Our Eyes: Addressing Gender-Based Violence in South African Schools, a Module for Educators." Gender Directorate, Pretoria: National Department of Education: Pretoria, South Africa. Stanko, E.A. et al., 1998. "Counting the Costs: Estimating the Impact of Domestic Violence in the London Borough of Hackney." Brunel University: Middlesex, United Kingdom. Strauss, Murray et al. 1980. Behind Closed Doors. Doubleday: New York Strauss, Murray, Richard Gelles, and Christine Smith. 1990. Physical Violence in American Families; Risk Factors and Adaptations to Violence in 8,145 Families. Transaction Publishers: New Brunswick. Sullivan, CM, and D Bybee. 1999. "Reducing Violence Using Community-Based Advocacy for Women with Abusive Partners." Journal of Consulting and Clinical Psychology 67(1):43-53. Thomas, DQ. 1994. "In Search of Solutions: Women's Police Stations in Brazil." In: Davies, M, Ed. Women and Violence: Realities and Responses Worldwide. Zed Books: London, pages 32-42. Thornberry, TP et al. 2001. "The Importance of Timing: The Varying Impact of Childhood and Adolescent Maltreatment on Multiple Problem Outcomes." Development and Psychopathology 13(4): 957-79. UNFPA. 2002. "Communication/Behaviour Change Tools: Entertainment-Education." Programme Briefs No. 1, January 2002. New York. Available at: www.unfpa.org/upload/lib_pub_file/160_filename_bccprogbrief1.pdf UNHCR. 1999. Reproductive Health in Refugee Situations: An Inter-Agency Manual. United Nations High Commission for Refugees: Geneva. Available at: www.who.int/reproductive- health/publications/interagency_manual_on_RH_in_refugee_situations/index.en.htm UNHCR. 1991. Guidelines on the Protection of Refugee Women. United Nations High Commissioner for Refugees: Geneva. UNICEF. 2003a. Partnerships in Practice: From AGEI to UNGEI. Summary Report of a Workshop on Girls' Education. Bamako, Mali, 17-20 November 2003. New York. Available at: www.unicef.org/girlseducation/UNICEF_Mali_Meeting_Report_Final.pdf UNICEF. 2003b. Gender and Education for All-the Leap to Equality. EFA Global Monitoring Report 2003/4. UNESCO: Paris. 57 UNICEF. 2004. Strategies for Girls' Education. New York. Available at: www.unicef.org/publications/English_Version_A.pdf United Nations General Assembly. 1993. Declaration on the Elimination of Violence Against Women. Proceedings of the 85th Plenary Meeting, Geneva, Dec. 20, 1993. United Nations: Geneva. United Nations Habitat. n.d. "Best Practices Database in Improving the Living Environment." (Electronic database). Available at: www.bestpractices.org United Nations Center for Human Settlements. 2001. The State of the World's Cities Report. UNESCAP: Geneva. (HS/619/01E). United Nations. n.d. Web page of the Division of the Advancement of Women, United Nations Department of Economic and Social Affairs. Available at: www.un.org/womenwatch/daw/cedaw/index.html Usdin, Shereen et al. 2000. "The Value of Advocacy in Promoting Social Change: Implementing the New Domestic Violence Act in South Africa." Reproductive Health Matters 8(16):55-65. Available at: www.soulcity.org.za/downloads/RHM%20Article.pdf Valdez, Enrique. 1999. "Using Hotlines to Deal with Domestic Violence in El Salvador." In: Morrison, Andrew, and María Loreto Biehl, Eds. Too Close to Home: Domestic Violence in the Americas. Inter- American Bank: Washington, D.C. Velzeboer, Marijke et al. 2003. Violence Against Women: The Health Sector Responds. Pan American Health Organization: Washington, D.C. Villanueva, Zarela. 1999. "Legislative and Judicial Reforms Regarding Domestic Violence: Costa Rica." In: Morrison, Andrew, and María Loreto Biehl, Eds. Too Close to Home: Domestic Violence in the Americas. Inter-American Development Bank: Washington, D.C. Walker, E et al. 1999. "Adult Health Status of Women HMO Members with Histories of Childhood Abuse and Neglect." American Journal of Medicine 107(4): 332-339. Ward, Jeanne. 2002. If Not Now, When? Addressing Gender-base Violence in Refugee, Internally Displaced and Post-conflict Settings. A Global Overview. The Reproductive Health for Refugees Consortium: New York. Warshaw, Carole, and Anne Ganley. 1998. Improving the Health Care Response to Domestic Violence: A Resource Manual for Health Care Providers. The Family Violence Prevention Fund: San Francisco. Watts, Charlotte et al. 1997. Strengthening the Health Sector Response to Women Experiencing Violence in Zimbabwe. Musasa Project: Harare. Wellesley Centers for Research on Women and DTS. 2003. "Unsafe Schools: A literature Review of School-Related Gender-Based Violence in Developing Countries." United States Agency for International Development: Washington, D.C. Available at: www.usaid.gov/our_work/cross- cutting_programs/wid/pubs/unsafe_schools_literature_review.pdf White, Victoria, Margaret Greene, and Elaine Murphy. 2003. "Men and Reproductive Health Programs: Influencing Gender Norms." U.S. Agency for International Development: Office of HIV/AIDS: Washington, D.C. Available at: http://www.synergyaids.com/SynergyPublications/Gender_Norms.pdf 58 Windom, Cathy. 1989. "The Cycle of Violence." Science 244: 160-166. World Bank. 2002. "Infrastructure & Gender Equality: Gender and Development Briefing Notes." The World Bank, Gender and Development Group: Washington, D.C. World Health Organization. 1999. "Summary of International and Regional Human Rights Texts Relevant to the Prevention and Redress of Violence Against Women." Commissioned by the Global Commission on Women's Health. Geneva. Yodanis, C, A Godenzi, and EA Stanko. 2000. "The Benefits of Studying Costs: A Review and Agenda for Studies on the Economic Costs of Violence against Women." Policy Studies 21(3): 263-76. 59 ANNEX A Physical assault on women by an intimate male partner, selected population-based studies, 1982-2003 Country Year of study Coverage Sample Study Age Proportion of women Size population* (years) physically assaulted by a partner (%) Previous Ever 12 months Africa Ethiopia 2002 Meskanena Woreda 2261 III 15-49 29 49 Kenya 1984-87 Kisii District 612 V >15 42 Namibia 2002 Winhoek 1367 III 15-49 16 31 Nigeria 1993 Not stated 1000 I 31a South Africa 1998 Eastern Cape 396 III 18-49 11 27 Mpumalanga 419 III 18-49 12 28 Northern Province 464 III 18-49 5 19 1998 National 10190 II 15-49 6 13 Tanzania 2002 Dar es Salaam 1442 III 15-49 15 33 Mbeya 1256 III 15-49 19 47 Uganda 1995-96 Lira & Masaka 1660 II 20-44 41d Zambia 2001-02 National 3792 III 15-49 27 49 Zimbabwe 1996 Midlands Province 966 I >18 17d Latin America and the Caribbean Barbados 1990 National 264 I 20-45 30a,c Brazil 2001 Sao Paulo 940 III 15-49 8 27 2001 Pernambuco 1188 III 15-49 13 34 Chile 1993 Santiago province 1000 II 22-55 26d 1997 Santiago 310 II 15-49 23 Colombia 1995 National 6097 II 15-49 19d 2000 National 7602 III 15-49 3 44 Dominican Republic 2002 National 6807 III 15-49 11 22 El Salvador 2002 National 10689 III 15-49 7d Guatemala 2002 National 6595f VI 15-49 8 Honduras 2001 National 6827 VI 15-49 6 10 Haiti 2000 National 2347 III 15-49 21 29 Mexico 1996 Guadalajara 650 III >15 27 Monterrey 1064 III >15 17 Nicaragua 1995 Leon 360 III 15-49 27 52 1997 Managua 378 III 15-49 33 69 1998 National 8507 III 15-49 13 30 Paraguay 1995-96 National 5940 III 15-49 10 Peru 2000 National 17369 III 15-49 2 42 2001 Lima 1019 III 15-49 17 50 2001 Cusco 1497 III 15-49 25 62 Puerto Rico 1995-96 National 4755 III 15-49 13e Uruguay 1997 National II 22-55 10c North America Canada 1993 National 12300 I >18 3d,e 29d,e United States 1995-96 National 8000 I >18 1.3c 22c Asia and Western Pacific Australia 1996 National 6300 I 3d 8dk Bangladesh 1992 National (villages) 1225 II <50 19 47 1993 Two rural regions 10368 II 15-49 42+I46 2003 Dhaka 1373 III 15-49 19 40 2003 Matlab 1329 III 15-49 16 42 Cambodia 1996 Six regions 1374 III 15-49 16 60 2000 National 2403 III 15-49 15 18 Country Year of study Coverage Sample Study Age Proportion of women Size population* (years) physically assaulted by a partner (%) Previous 12 Ever months India 1998-99 National 90303 III 15-49 10 19 1999 Six states 9938 III 15-49 14 40 Indonesia 2000 Central Java 765 IV 15-49 2 11 Japan 2001 Yokohama 1276 III 18-49 3 13 Papua New Guinea 1982 National, rural villages 628 IIIf 67 Philippines 1993 National 8481 IV 15-49 10 1998 Cagayan de Oro City & 1660 II 15-49 26 Bukidnon Republic of Korea 1989 National 707 II >20 38 Samoa 2000 National 1204 III 15-49 18 41 Thailand 2002 Bangkok 1048 III 15-49 8 23 2002 Nakonsawan 1024 III 15-49 13 34 Vietnam Ha Tay province 1090 III 15-60 14 25 Europe Azerbaijan 2001 National 5533 III 15-44 8 20 Georgia 1999 National 5694 III 15-44 2 5 Finland 1997 National 4955 I 18-74 9bc 22bc Netherlands 1986 National 989 I 20-60 21a Norway 1989 Trondheim 111 III 20-49 18 Republic of Moldova 1997 National 4790 III 15-44 8 15 Romania 1999 National 5322 III 15-44 10 29 Russia 2000 Three provinces 5482 III 15-44 7 22 Serbia & Montenegro 2003 Capital 1189 III 15-49 3 23 Sweden 2000 National 5868 III 18-64 4 18 Switzerland 1994-96 National 1500 II 20-60 6c 21c Turkey 1998 E & SE Anatolia 599 I 14-75 58a Ukraine 1999 National 5596 III 15-44 7 19 United Kingdom 1993 North London 430 I >16 12a 30a Eastern Mediterranean Egypt 1995-96 National 7123 III 15-49 13 34 Israel 1997 Arab population 1826 II 19-67 32 West Bank/Gaza 1994 Palestinian population 2410 II 17-65 52 Key Study population: I = all women; II = currently married/partnered women; III = ever-married/partnered women; IV = women with a pregnancy outcome; V = married women - half with pregnancy outcome, half without; VI women who had a partner within the last 12 months aSample group included women who had never been in a relationship and therefore were not in exposed group bAlthough sample includes all women, rate of abuse is shown for ever-married/partnered women (number not given) cPhysical or sexual assault dDuring current relationship eRate of partner abuse among ever-married/partnered women recalculated from author's data fWeighted for national representativity gWithin the last five years Source: Ellsberg, Mary, and Lori Heise, et al. Forthcoming. Researching Violence Against Women: A Practical Guide for Researchers and Advocates. World Health Organization, Center for Health and Gender Equity: Washington, D.C. 61