Encouraging reporting of violence against women: Shifting norms, improving access
According to the World Health Organization, 1 in 3 women worldwide have experienced physical or sexual violence in their lifetimes. Low rates of reporting pose a barrier to addressing this form of violence, especially as it is commonly perpetrated by an intimate partner within the home. Recent global estimates suggest that fewer than 40 percent of women seek help after experiencing violence, and among those that do, fewer than 10 percent report such cases to the police. Individuals affected by this violence—victims and bystanders alike—may be reluctant to speak out to friends, family, or police.
In recent years, a growing number of randomized evaluations have been launched to help identify concrete measures to increase reporting and reduce the incidence of violence against women. This is just one potential response—other approaches intentionally target instead the behavior of perpetrators of this violence, while a growing strand of research is examining how cash transfers can be effective in reducing intra-household violence. We review here recent studies that have focused on two different approaches to increasing reporting, as part of a broader effort to reduce violence against women:
- Shifting social norms around gender roles and gender-based violence, and
- Opening up new channels for women to report cases to the police.
These studies are starting to build up evidence that it is indeed possible to change norms and behaviors around reporting violence against women (even when norms about the violence itself may be difficult to shift), and that increasing women’s access to the police—including through staffing stations with trained female officers—may also be an important part of the solution.
Shifting social norms
There are a number of potential reasons why women may choose not to report cases of violence. For instance, they might fear retaliation by their partner or have little faith the police will take action. Social norms may also play a role: across Africa and Asia, between 30 and 60 percent of women agree that a husband is justified in beating his wife under certain circumstances.
Trying to shift these norms could ultimately help reduce incidence in two ways: either by directly shaping the behavior of those who engage in such violence or by reducing the perceived costs of speaking out and reporting these crimes. Interventions that focus on the latter would ideally be matched by efforts to ensure that reporting leads to meaningful action by public authorities.
Increased political voice, increased police responsivenessAn earlier study from India provides promising evidence of how a shift in women’s political representation had a dramatic effect on reporting behaviors. Lakshmi Iyer (Notre Dame) and co-authors examine the impact of a 1993 constitutional amendment that called for a random third of village council positions to be reserved for women. They find that the reservation led to a 46 percent increase in the number of documented crimes against women.
They further show that this rise appears to have been primarily due to an increase in reporting of incidents to police rather than any increase in incidence—as there was no significant change in the level of crimes not targeted against women or in crimes against women such as murder that are unlikely to be affected by reporting biases. More encouragingly, the increase in women elected to these village council roles also led to heightened police responsiveness and increased arrests for crimes committed against women. (Learn more about insights from J-PAL studies on the wide-ranging effects of gender quotas in local government here.)
Changing community attitudes towards reportingWitnesses of violence against women may withhold information due to fear of community censure or being branded as a gossip. Encouraging bystanders to report violence may require shifts in norms about what it means to report it.
J-PAL affiliate Donald Green (Columbia), and co-authors Jasper Cooper (UCSD) and Anna Wilke (Columbia), evaluated the effectiveness of using a mass media intervention in rural Uganda to encourage reporting of violence against women. They held a film festival in 112 villages and, in a randomly selected subset of those villages, they screened education-entertainment videos addressing crimes against women during the festival intermissions. The short vignettes depicted violence by a husband towards his wife and appealed to viewers to speak out about these crimes, despite norms against meddling in the private affairs of others.
Following the intervention, women in villages where the videos were screened were 18 percent less likely to believe that they would face social sanctions for reporting violence against women and 22 percent more likely to say they would make a report. The researchers also found a reduction in some measures of incidence: they estimate that the likelihood of women experiencing violence in treatment communities fell by five percentage points (from a baseline of 20 percent in control communities). However, the researchers found no effects on community members’ acceptance of violence against women. Building on these results, Donald Green, Dylan Groves (Columbia), Rachel Jones (Innovations for Poverty Action), and Constantine Manda (Yale) are conducting a similar evaluation in Tanzania to measure the impact of radio programming on both attitudes and behaviors.
Enlisting village leaders as agents of changeResearchers in Peru have partnered with the government to ask whether a locally driven intervention could lead to shifts not only in reporting but also in attitudes towards violence. J-PAL affiliate Erica Field (Duke), and co-authors Ursula Aldana (Instituto de Estudios Peruanos) and Javier Romero (Duke), are partnering with the Ministry of Women and Vulnerable Populations to evaluate the effectiveness of an intensive training for rural village leaders designed to reduce acceptance of gender-based violence and provide practical methods for community members to identify and report these crimes.
The five-session training course for village leaders covers topics such as stereotypes surrounding gender roles, norms regarding violence, and strategies to identify and prevent violence against women. These leaders–both men and women–develop and execute an action plan that includes regular awareness-raising activities within the village, house-to-house visits, and additional trainings. The idea behind the program is to engage individuals with local influence to drive changes not only in reporting of cases, but also in behaviors and attitudes that perpetuate gender-based violence more generally, leading to a reduction in incidence.
Facilitating (women's) access to police
Another way to encourage reporting of violence against women is to make it easier to approach the police. A recent evaluation by Jasper Cooper considered the impact of introducing uniformed auxiliary police officers in rural areas of Bougainville, Papua New Guinea, where previous police presence had been limited. Cooper finds that the introduction of a permanent police presence in the village leads to increased crime reporting by women, and that this is particularly true when the officer is female.
Adding female police officers without specialized training or strategic roles, however, may not be an effective way of addressing this issue. A recent lab-in-field experiment by Sabrina Karim (Cornell) and co-authors estimated the impacts of randomly varying the proportion of women officers in teams of the Liberian National Police. Teams with heavier concentrations of women exhibited greater cohesion but were no more sensitive to sexual and gender-based crimes than teams with fewer women.
In light of this, new research is examining a different question: can increasing the representation of female officers in dedicated public-facing roles, combined with special training, increase both the reporting of crimes against women and their effective prosecution?
A quasi-experimental study by Sofia Amaral (University of Munich) and co-authors presents suggestive evidence from the rapid expansion in “all-women” police stations in India from 2005-13. They find that the introduction of these units, staffed entirely by female officers trained in handling crimes committed against women, led to a 22 percent increase in reporting. There was no increase in non-gender-specific crimes, suggesting again that what changed was women’s willingness to report these crimes.
Similarly, a new study in Madhya Pradesh by J-PAL affiliate Sandip Sukhtankar (UVA), with Gabrielle Kruks-Wisner (UVA) and Akshay Mangla (Oxford), is evaluating whether opening women’s help desks—dedicated spaces in regular police stations staffed by officers trained to assist with crimes against women—can improve frontline officers’ responsiveness to women, as well as reduce crime.
Looking ahead
Taken together, the studies above provide some initial insights into promising approaches for addressing violence against women. Although influencing the core attitudes around gender-based violence may be challenging, rigorous evidence has shown that it is possible to shift both victims’ and community members’ behaviors towards reporting these crimes to police. Research currently underway, as well as new studies in this field, can provide valuable insight into the most effective ways to structure police response and on the potential approaches for changing attitudes towards violence against women, in the hopes of reducing its incidence.
Authored By
Education entertainment campaigns (“edutainment”) have the potential to spark behavioral and attitudinal change. In rural Uganda, researchers evaluated whether videos encouraging communities to speak out against violence against women (VAW) could change behavior, attitudes, and norms. Women exposed to the videos were more willing to report VAW to authorities. More broadly, women in villages where the videos were screened experienced less VAW in their household. The impact seems driven by a decline in perceived social sanctions for speaking out, despite no changes in acceptance of VAW.
Policy issue
Violence against women (VAW) is a global problem that affected at least 35 percent of women worldwide as of 2013.1 Its drivers include attitudes that make violence socially acceptable, and also low reporting rates among victims and witnesses, which prevent authorities from detecting and addressing incidents.
Mass media campaigns may help address these issues by reducing the acceptability of VAW and encouraging victims and bystanders to report—two effects that may be amplified if the campaigns’ messages are further discussed with other members of the community. However, gender attitudes tend to be deeply rooted and hard to shift by light-touch interventions like media content. This study sheds light on the effects of mass media interventions on VAW, as well as their channels of impact, by evaluating an anti-VAW educational-entertainment campaign in rural Uganda.
Context of the evaluation
Opinion polls conducted between 2001 and 2015 suggest that permissive attitudes towards intimate partner violence are widespread in Uganda.2 However, not all violence is condoned. While 31 percent of respondents in this study said a husband is justified in beating his wife when she disobeys him, only 5 percent of those who think the beating is justified would condone violence perceived as more severe than slapping. Meanwhile, 86 percent of respondents in the study’s comparison group stated that others should intervene to stop daily violence.
Nonetheless, almost a third of rural Ugandan women reported experiencing violence such as being punched, kicked, or threatened with a knife in 2011.3 Formal police and community-level actors do not seem able to prevent such violence, partly because witnesses do not speak out—only one quarter of respondents in this study said they would tell local village leaders if their cousin had been beaten by her husband, while fewer than one in five would report to police. A common justification for withholding information was fear of being branded a gossip and the perceived risk of ostracism by community members who may question a witness’s motivation for coming forward.
Details of the intervention
In rural Uganda, researchers conducted a randomized evaluation to measure the impact of a mass video campaign designed to encourage audiences to speak out against VAW. A total of 112 rural villages were randomly assigned to one of two groups:
- Anti-VAW videos group (48 villages): In these villages, three short anti-VAW videos were screened during intermission at film festivals held in video halls once per week over consecutive weekends for a little over a month. The videos were produced in collaboration with Peripheral Vision International in the local language (Luganda) and were filmed on site, enabling villagers to identify with the characters. Ranging from 4.5-8 minutes each, the videos depicted violence by a husband towards his wife and appealed to viewers to speak out about VAW in order to prevent it from escalating. The festivals showed Hollywood films, were free of charge, and were advertised using posters, fliers, and announcements through public loudspeakers when they were available. In some of these villages, the anti-VAW videos were screened in addition to another set of videos on stigma surrounding other social issues (e.g., abortion and teacher absenteeism).
- Comparison group (64 villages): In these villages, the film festival movie was screened without any anti-VAW videos during intermission.
In total, over 10,000 adults attended 670 film screenings. Two and eight months after the film screenings, the researchers surveyed 5,534 randomly sampled adults living in the catchment areas of the video halls, irrespective of their attendance of the festival. Importantly, the surveys were presented as opinion polls unrelated to the video campaign. Questions measured general attitudes about VAW, expectations around the disclosure of violence, willingness to report VAW to formal and informal authorities, and self-reported VAW incidents.
Researchers took steps to ensure the intervention and data collection protected participants’ rights and well-being. They vetted video content through focus groups and NGO consultations to reduce the risks of triggering past trauma and conducted interviews privately using general, household-level questions to safeguard privacy and prevent partner retaliation.
Results and policy lessons
The anti-VAW videos reduced the likelihood of VAW occurring in villages where the videos were shown. Results suggest that this may have occurred because the videos reduced the perception that those who speak out about VAW will face social sanctions and thereby increased individuals’ willingness to report incidents of VAW.
Willingness to report VAW: Two and eight months after the campaign, women who attended the anti-VAW screenings were 8.5 percentage points (22 percent) and then 12.6 percentage points (35 percent) more likely to say they would report violence as bystanders in hypothetical scenarios (e.g., involving the victim’s parents, a counselor, or village leader, or reporting to the police) than women in the comparison group. Men who attended the screenings were also 4 percentage points (11 percent) more likely to state they would report hypothetical incidents of VAW eight months later relative to men in the comparison group.
Incidence of violence: Eight months after the screenings, the proportion of women who reported any VAW in their household fell by 7 percentage points in campaign villages, from an average of 19 percent in the comparison villages. This effect referred to all women in the campaign villages, regardless of if they personally attended the anti-VAW screenings or not. It implied the prevention of VAW in hundreds of households in villages where the VAW campaign occurred.
Social norms, attitudes, and beliefs around VAW: Among men and women who attended screenings, there was little evidence that the anti-VAW videos affected attitudes towards VAW or norms concerning gender equality. There was also no evidence of increases in empathy for VAW victims or a change in perceptions that initial acts of domestic violence were likely to escalate to more severe forms of violence. Men and women’s perceptions of the efficacy of reporting to prevent future violence also did not change with the campaign.
On the other hand, women who watched the anti-VAW videos were 11 percentage points (18 percent) less likely to believe they would face social repercussions for intervening in a VAW incident eight months after the screening, such as scolding for gossiping. When asked about their willingness to report VAW in the presence or absence of witnesses who could support their claim, participants who attended the screening said they would report cases similarly regardless of the presence of witnesses, while those in the comparison group were more willing to report VAW if they had corroborating witnesses. This result further suggests that the campaign reduced the fear of social sanctions for reporting.
Indirect effects: While discussions between audiences and others in their social network could have amplified the effects of the anti-VAW videos by expanding exposure to its messages, there were no such indirect effects of this mass video campaign. Friends and family of audience members who did not attend the film festival were no more willing to take action to assist victims of VAW nor to report incidents to village authorities (victim’s parents, a counselor, village leader, or the police).
Taken together, these results suggest that higher willingness to report was not driven by shifts in attitudes or changes in the overall perception of the impact of reporting on VAW. Rather, it is likely that participants became more willing to report because the campaign alleviated their concerns about sanctions related to gossiping, an important barrier to reporting.
World Health Organization. 2013. "Global and Regional Estimates of Violence Against Women: Prevalence and Health Effects of Intimate Partner Violence and Non-Partner Sexual Violence." https://www.who.int/publications/i/item/9789241564625.
Demographic and Health Surveys 2001-2015.
Demographic and Health Survey 2011.