Violence against Women and Stigma

Research in Latin America reveals that the vast majority of victims do not have access to or reach out for formal help from specialized public services (Kras Reference Kras2022; Htun and Jensenius Reference Htun and Jensenius2022). This is by no means a problem unique to Latin America, as many forms of VAW, including sexual violence, are vastly underreported to the authorities in the Global North as well (e.g., James and Lee Reference James and Lee2015). The absence or inaccessibility of shelters, crisis centers, specialized police stations, and other services has dire consequences for survivors and precludes efforts to effectively combat VAW. Victims are more likely to use services when they are available locally (Kras Reference Kras2022). Laws and public services for victims increase women’s sense of safety and decrease rates of violence (Córdova and Kras Reference Córdova and Kras2022; Htun and Jensenius Reference Htun and Jensenius2022). IPV survivors report positive interactions with service providers as healing and empowering, even when services are unable to provide survivors with tangible outcomes (Bell et al. Reference Bell, Perez, Goodman and Ann Dutton2011). In fact, survivors often turn to the police or other services for advice, guidance, “someone to talk to,” and protection, rather than to pursue justice against offenders (e.g., Cattaneo and Goodman Reference Cattaneo and Goodman2010; Kras Reference Kras2022). For society at large, laws and visible institutions for survivors can foster critical attitudes toward VAW (Arias Reference Arias2019; Córdova and Kras Reference Córdova and Kras2022).

What remains unclear is the effect of anti-VAW public policy on personally held stigmatizing and stereotyped opinions of VAW, as well as the extent to which one perceives that others hold these views. Stigma refers to an “attribute that is deeply discrediting” (Goffman Reference Goffman1963). While stigma is shaped by place and time, virtually all members of a society are aware of the identities that are tainted or undesirable (Major and O’Brien Reference Major and O’Brien2005). Further, the literature distinguishes between overlapping manifestations of stigma. These include public stigma, or prejudices endorsed by the public; anticipated stigma, or the fear of public stigma; and internalized stigma, wherein the stigmatized person develops a negative self-image (Gronholm et al. Reference Gronholm, Henderson, Deb and Thornicroft2017). Many identities have been subjected to stigma, including people with mental health conditions and HIV/AIDS (Clair et al. Reference Clair, Daniel and Lamont2016), and only recently has research begun to investigate VAW survivors as targets of stigma. The public stigma around VAW victims often involves judging, labeling, and pathologizing them, and it’s deeply connected to victim-blaming and beliefs that IPV is private or normal (Crowe and Murray Reference Crowe and Murray2015). The degree of stigmatization of survivors of VAW varies widely across countries. But research reveals a broader pattern of stigmatization of VAW victims that is not unique to a country or region. For example, a study with IPV victims in the United States found that 55 percent of their sample reported experiencing subtle forms of victim-blaming (Overstreet et al. Reference Overstreet, Willie and Sullivan2019).

Stigma has powerful and sweeping effects on its targets (Major and O’Brien Reference Major and O’Brien2005). For survivors of IPV, stigmatization shapes their feelings and behaviors as they recover and their help-seeking and coping processes. Internal consequences of stigma include low self-esteem, embarrassment, fear, and depression (Crowe and Murray Reference Crowe and Murray2015; Overstreet et al. Reference Overstreet, Willie and Sullivan2019). Consequently, fear of what others might think and say (anticipated stigma) is a major barrier to help-seeking for survivors of IPV (Overstreet and Quinn Reference Overstreet and Quinn2013): when people judge disclosing private information to someone as risky, they are much less likely to do so (Afifi and Steuber Reference Afifi and Steuber2009).

For example, our focus group interviews with healthcare professionals in Guatemala suggest that women engage in strategic thinking when deciding whether to disclose their experience with IPV to others. One participant said, “women don’t find support from their families … So, they just stay quiet … And even if they go to institutions, they also turn their backs on them. So, then they think, ‘What’s the point of going? It’s just shameful for me…’” [emphasis added]. When victims perceive high societal stigma toward IPV, they judge the costs of seeking help from formal and informal channels too high to bear, such as acquiring shame or experiencing gossip. Indeed, another participant in our focus group added: “Sometimes even the family blames the woman [who experienced IPV]—maybe because she is not fulfilling her duties as a wife…” Victim-blaming is, of course, a highly stigmatizing reaction to IPV. Thus, effective strategies to clamp down on VAW must inevitably confront stigma.

Extensive research examining mental health and HIV/AIDS has shed light on strategies to reduce stigma. For example, anti-stigma strategies have been categorized into advocacy, contact with the targets of stigma, and various forms of social support (Heijnders and Meij Reference Heijnders and Van Der Meij2006; Gronholm et al. Reference Gronholm, Henderson, Deb and Thornicroft2017). Governmental and structural-level interventions are also theorized to reduce public and internalized stigma; however, few studies have empirically tested this or proposed mechanisms (Murvartian et al. Reference Murvartian, Saavedra-Macías and Infanti2023). Thus, this study contributes to the literature on VAW and stigma by investigating whether state efforts to address VAW can reduce personally held stigmatizing attitudes and perceptions of the extent to which others hold stigmatizing views of VAW, even in contexts of entrenched impunity and limited resources.

Anti-VAW Legislation Policy Feedback Effects and Source Cues

This study builds on the policy feedback effect and elite cues literature (e.g., Jacobs and Mettler Reference Jacobs and Mettler2018). In essence, this approach to public opinion holds that policies, laws, and credible signals from elites can alter public attitudes and establish the parameters of behavior (e.g., Zaller Reference Zaller1992). Policies can alter attitudes around several social issues, including smoking (Pacheco Reference Pacheco2013) and same-sex marriage (Kreitzer et al. Reference Kreitzer, Hamilton and Tolbert2014). As such, legislation and messages from elites hold the potential to shape how individuals view victims of IPV as well as raise the salience of VAW as a legitimate social problem. Scholars have long argued that legislation serves purposes beyond sanctions and deterrence. Rather, laws serve as informational vehicles that reveal accepted norms and behaviors (Nadler Reference Nadler2017).

Research in the context of Latin America has shown that policies on VAW shape opinions, providing some evidence that there can be feedback effects from VAW policy. Htun and Jensenius (Reference Htun and Jensenius2022) find that women in Mexico became increasingly more intolerant of IPV and less likely to experience violence after the anti-VAW law was adopted. This finding is in line with Córdova and Kras (Reference Córdova and Kras2022), whose study finds that men in Brazil are more likely to display intolerance of IPV in municipalities with an established women’s police station. A study in Mexico by Arias (Reference Arias2019) also finds that radio messages condemning VAW led the public to reject VAW. The present study builds on these works to examine whether awareness of anti-VAW laws and other concerted efforts to prevent and remedy VAW also shape people’s views on public stigma.

While recent studies, such as those described above, have begun to investigate the policy feedback effect thesis in the Global South, most of the existing literature has focused on the United States and Europe. Similarly, the source cues literature has focused primarily on the United States, examining the impacts of the Supreme Court (e.g., Mondak Reference Mondak1994) and party members (e.g., Tesler Reference Tesler2018). The policy feedback effect literature has focused on visible and traceable policies and their impacts on public opinion, such as smoking bans in the United States (Pacheco 2012) and women’s police stations in Brazil (Córdova and Kras Reference Córdova and Kras2022). However, these institutions can leverage some level of confidence in their contexts.Footnote ¹ But in many Latin American countries, trust in political elites and the justice system is notoriously low (Lupu et al. Reference Lupu, Rodríguez, Wilson and Zechmeister2023). Distrust in government may blunt the effects of policies on attitudes and behaviors (Jacobs and Mettler Reference Jacobs and Mettler2018). Therefore, there are reasons to doubt the application of the feedback thesis in Mexico and Guatemala.

Scholars argue that for policies and elite rhetoric to exert a persuasive function, the public must have confidence in them (Stoutenborough et al. Reference Stoutenborough, Haider-Markel and Allen2006; Mondak Reference Mondak1994). As such, we are particularly interested in investigating whether credible efforts to combat VAW can diminish perceptions of public stigma even when the public believes there is widespread impunity in their countries. Indeed, Latin Americans recognize the weakness of the law. For example, people expect more punishment for VAW in Mexico but believe that in reality, these crimes go unpunished (Barba et al. Reference Barba, Lupu and Zechmeister2023). In Guatemala, 98 percent of homicides of women remain unpunished (Wands and Mirzoev Reference Wands and Mirzoev2022). Trust in institutions is low in Guatemala, even by Latin American standards (Lupu et al. Reference Lupu, Rodríguez, Wilson and Zechmeister2023).Footnote ² This is hardly surprising given Guatemala’s long history of political violence, corruption, impunity, poverty, and neglected public services (e.g., Menjívar Reference Menjívar2011; Wands and Mirzoev Reference Wands and Mirzoev2022). Fair processes and the delivery of outcomes—both of which are lacking in Guatemala and Mexico, as indicated by the corruption index, rampant impunity, and inadequate public service provision—are associated with legitimacy (e.g., Esparza and Ugues Reference Esparza and Ugues2020). When institutions deliver, the public is much more approving of them (see Reinold Reference Reinold2022).

Thus, can political efforts to confront VAW shift perceptions of public stigma even when people have little faith in the justice and political system? We posit that awareness of anti-VAW laws and other signals of state condemnation of IPV can foster a perception that society does not stigmatize victims (public stigma) even when trust in institutions runs low. In these contexts, source credibility should be particularly important in generating shifts in perceptions of IPV survivors. When messages conveying the extent to which the state condemns VAW are transmitted to citizens through credible sources (e.g., Mondak Reference Mondak1994), citizens may perceive that society is welcoming to IPV victims and that they do not have to be ashamed (public stigma is absent). After all, credible and authoritative figures are committed to ensuring the rights and dignity of IPV victims. Credible sources of information about anti-VAW efforts are likely to be authoritative figures with whom the average person has more exposure at the local level. Such sources have some track record of delivering outcomes that citizens can observe.

Messages disseminated by police and health professionals might carry more weight and be perceived as more credible by citizens in contexts of impunity and low institutional trust. For a cue to be influential, people must believe that the cue-giver has some knowledge or authority in the subject (Gelpi Reference Gelpi2010). Both police and healthcare professionals are authorities on VAW, and in developing countries, they may be the closest people come to an authoritative figure (see Menjívar Reference Menjívar2011). In this case, messages about the state’s efforts to seriously address VAW might convey mental representations of IPV victims as deserving of rights and undeserving of ostracism when delivered by healthcare professionals or the police. Trust in doctors and nurses is higher in Latin America than in the government (e.g., Moucheraud et al. Reference Moucheraud, Guo and Macinko2021). While trust in the police in Guatemala and Mexico is very low (Lupu et al. Reference Lupu, Rodríguez, Wilson and Zechmeister2023), the police may be perceived as authorities in crime-related matters. Their “subject matter” knowledge in crime-fighting might offset the negative impact of police impunity and corruption in convincing citizens that IPV is a legitimate concern and that there is institutional will to combat this problem (reducing public stigma).

However, in highly patriarchal societies like Guatemala and Mexico (e.g., Menjívar and Walsh Reference Menjívar and Walsh2016; Menjívar Reference Menjívar2011),Footnote ³ we expect citizens to perceive male police officers and healthcare professionals as more credible in this messaging. Men have traditionally dominated the state’s security apparatus and are therefore the “default experts” in determining the seriousness of crime types. Research shows that in “male-dominated arenas,” women are seen as less competent and judged more harshly by the public (Carlin et al. Reference Carlin, Carreras and Love2020). When it comes to matters of national security and external threats, people prefer men’s leadership over women’s (Holman et al. Reference Holman, Merolla and Zechmeister2016). In terms of medicine, although the global health workforce is predominantly composed of women, men have disproportionately occupied the higher ranks of the profession (e.g., Shannon et al. Reference Shannon, Jansen, Williams, Cáceres, Motta, Odhiambo, Eleveld and Mannell2019). Thus, when the “default experts” in crime and health send credible signals of commitment to anti-VAW governmental action, citizens might be more persuaded of the seriousness of VAW and that victims are legitimate recipients of these rights.

Highly credible male messengers are likely to be more influential in reducing men’s stigmatizing opinions of IPV. Research shows that institutions such as women’s police stations exert stronger effects on men’s rejection of VAW compared to women, who already have high baseline levels of rejection of VAW (Córdova and Kras Reference Córdova and Kras2022). But when it comes to perceptions of the public stigma of IPV victims, women might be more persuaded that state measures are credible—and therefore victims aren’t stigmatized— when male authorities demonstrate commitment to anti-VAW efforts. Women are more likely than men to recognize that women suffer discrimination in society (e.g., Clayton et al. Reference Clayton, O’Brien and Piscopo2019) and are more likely to personally know an IPV victim (Córdova and Kras Reference Córdova and Kras2022). As a consequence, women might find it especially hard to believe that IPV survivors are not blamed and shamed—suffer stigmatizing reactions from society, e.g., public stigma—and that their rights are protected and guaranteed.

Source cues are less influential on the opinions of those with stable attitudes (e.g., Chong and Druckman Reference Chong and Druckman2012). Women may hold crystallized beliefs regarding public stigma toward VAW and, therefore, be more likely to deliberate rather than passively rely on elite cues (e.g., Chong and Druckman Reference Chong and Druckman2012). However, even when attitudes are well-developed, cues that reveal unexpected information can powerfully persuade people to revise their judgments (Gelpi Reference Gelpi2010; Nicholson Reference Nicholson2011). Messengers are also deemed more trustworthy when they take surprising positions that are contrary to what people expect (Nicholson Reference Nicholson2011). The public expects female police officers and leaders to take gendered crimes seriously (Huber and Gunderson Reference Huber and Gunderson2023). But in societies where gender inequality is rife, anti-VAW messages from credible men might be more surprising and consequently more potent in convincing women that state action in this area is serious, reducing the public and societal stigma that victims suffer. This discussion leads to the following hypotheses:

As noted above, source cues are more likely to shape opinions in low-salience and non-polarized areas, and when people have little experience with the policy or have otherwise less crystallized attitudes (e.g., Chong and Druckman Reference Chong and Druckman2012; Nicholson Reference Nicholson2011; Tesler Reference Tesler2018). The influence of credible elite rhetoric and signaling on individuals’ opinions depends on the degree to which these attitudes are malleable. Thus, those holding some sexist attitudes should be less susceptible to anti-VAW messages, even if communicated through credible elites. Some sexist attitudes not only should produce stable opinions about women but can also be triggered through messages about policies that target women as beneficiaries (Cassese and Holman Reference Cassese and Holman2019).

Scholars have demonstrated the existence of two distinct yet interconnected forms of sexism among the public: benevolent and hostile sexism (e.g., Glick and Fiske Reference Glick and Fiske1996). While benevolent sexism endorses traditional gender stereotypes and might appear positive, hostile sexism is more broadly recognized as chauvinistic, given its derogatory beliefs toward women (e.g., Cassese and Holman Reference Cassese and Holman2019). Benevolent sexists hold paternalistic attitudes toward women rooted in the belief that women should be protected as they are best suited for conventional gender roles. Hostile sexists, meanwhile, are explicitly antagonistic toward women and perceive the relationship between men and women as a zero-sum game (Glick and Fiske Reference Glick and Fiske2001; Cassese and Holman Reference Cassese and Holman2019). Both manifestations of sexism can be endorsed by men and women (e.g., Glick and Fiske Reference Glick and Fiske2001).

Scholars have shown that both dimensions of sexism predict attitudes related to VAW. Hostile sexism is directly associated with tolerance and perpetration of domestic violence (e.g., Juarros-Basterretxea et al. Reference Juarros-Basterretxea, Overall, Herrero and Rodríguez-Díaz2019) and opposition to pro-women social spending (Beauregard et al. Reference Beauregard, Holman and Sheppard2022). While benevolent sexism enjoys greater societal acceptance, it too predicts harmful attitudes toward women (and men). Individuals scoring high on a benevolent sexism scale are more likely to blame victims of acquaintance rape (e.g., Abrams et al. Reference Abrams, Viki, Masser and Bohner2003) and to endorse the use of aggression to protect male honor (Saucier et al. Reference Saucier, Stanford, Miller, Martens, Miller, Jones, McManus and Burns2016). Because hostile sexism powerfully predicts endorsement and perpetration of VAW, we posit that hostile sexists should hold crystallized attitudes toward women, and therefore be less susceptible to cues.

Hostile sexists should view VAW as an exaggerated problem and hold stigmatizing attitudes toward victims, while at the same time, they perceive that society does not discriminate against IPV victims. Recognizing that a group is stigmatized implies that they experience discrimination and that they are legitimate recipients of special protections (e.g., Mangum and Block Reference Mangum and Block2021), which is at odds with the hostile sexists’ worldview. Hostile sexists view women as exaggerating their problems and might perceive anti-VAW laws as a sign that women demand “special favors.” Therefore, we expect hostile sexists to resist persuasion and display highly stigmatizing opinions of IPV, while downplaying public stigma. Those without such underlying attitudes, meanwhile, are much more susceptible to the messenger’s credibility when forming opinions about IPV.

Case Studies: Mexico and Guatemala

In Latin America, violence against women is a persistent concern. The WomenStats database indicates that women’s physical security across the region is low. However, VAW is especially acute in Central America and Mexico, for example, and less so in some countries of the Southern Cone, highlighting the country-level variation in women’s physical security in the region. Central America carries the dubious distinction of having some of the highest femicide rates globally. In Guatemala, there were 624 femicides recorded for 2022 alone, while in Mexico, an estimated total of 3,754 women and girls were murdered in the same year (Kloppe-Santamaría and Zulver Reference Kloppe-Santamaría and Zulver2023). Indeed, one of the driving forces behind women’s decisions to migrate to the United States is the staggering scale of gender-based violence in the region (Menjívar and Walsh Reference Menjívar and Walsh2019).

In Guatemala, IPV is also disturbingly common and widely accepted (Menjívar Reference Menjívar2011). Indeed, in our focus group interviews with healthcare providers, participants highlighted that women in small towns accept IPV as part of marriage— “if their mom endured [IPV], they think so should they,” one participant said. Further, Guatemalan women have the lowest labor market participation in the region (Beck Reference Beck2021; Menjívar Reference Menjívar2011). Economic vulnerability severely hinders IPV victims’ ability to start a new life. In Mexico, a representative survey found that 27.4 percent of female respondents had experienced domestic abuse in the previous year (Htun and Jensenius Reference Htun and Jensenius2022). Low service delivery, weak rule of law, and limited political commitment to implementing VAW policies, combined with criminal violence, result in very low levels of reporting and help-seeking among survivors (Frías Reference Frías2022; Morse et al. Reference Morse, Colombini, Lentz, Soriano, Morales, Olavarrieta and Rodríguez2025; Htun and Jensenius Reference Htun and Jensenius2022; Kloppe-Santamaría and Zulver Reference Kloppe-Santamaría and Zulver2023). Indeed, generalized insecurity in Mexico, rooted in criminal violence and the heavy-handed state response it has elicited, is found across several indicators, such as levels of lethal violence against local-level public officials (Ley Reference Ley2018) and crime victimization (Lupu et al. Reference Lupu, Rodríguez, Wilson and Zechmeister2023). Despite the link between criminal violence and VAW, criminal violence has dominated public and scholarly debates about Mexico (Kloppe-Santamaría and Zulver Reference Kloppe-Santamaría and Zulver2023).

As elsewhere in Latin America, Mexico and Guatemala have adopted anti-VAW laws that set forth clear state obligations. In Guatemala, the 2008 law guarantees resources for comprehensive support centers for women and legal assistance for victims. Similarly, a strong anti-VAW law was adopted in 2007 in Mexico. The law defined the crimes of femicide and domestic violence and established a national system to prevent, punish, and eradicate VAW. Further, Mexico has implemented notable measures to curb VAW. For example, following similar approaches adopted by other countries in the region, some states in Mexico have established women’s police stations designed to serve victims of VAW. However, despite these efforts, impunity for VAW and crimes in general remains an intractable problem. For example, only 3 percent of murders of women have resulted in sentences in Mexico, versus 11 percent for all homicides (Barba et al. Reference Barba, Lupu and Zechmeister2023). The WomenStats database highlights the justice system’s failure in reporting, investigating, and prosecuting crimes against women. The database also observes that women in situations of violence in Mexico came to expect little to no response from the authorities.

Meanwhile, in Guatemala, scholars observe a “patchwork” of implementation of services for victims, with some implementation in urban areas and complete neglect in rural, heavily Indigenous regions (Beck Reference Beck2021). Our fieldwork corroborates Beck’s findings. Participants in our focus group highlighted that there are many government “initiatives” but only on paper. Further, anti-VAW legislation in Guatemala is in direct tension with persistent discriminatory legal practices and laws around property and divorce, creating barriers to justice for women victims of violence (Menjívar and Walsh Reference Menjívar and Walsh2016). In both Mexico and Guatemala, trust in the police is notoriously low (Lupu et al. Reference Lupu, Rodríguez, Wilson and Zechmeister2023). Even still, research in Guatemala suggests that 77 percent of survey respondents expressed a preference for increased police presence (Denny et al. Reference Denny, Dow, Pitts and Wibbels2023). In our focus group in Guatemala, participants lamented the absence of policing in their municipality when we asked about reporting VAW to the police, noting that they don’t even go when it is a domestic violence call. The WomenStats database notes that law enforcement’s insufficient capacity leads to high levels of impunity in Guatemala. Adding to these serious structural problems, Guatemala ranks the highest in the region in levels of tolerance for domestic violence, with 58 percent of Guatemalans either approving or “understanding” a husband hitting an unfaithful wife (Azpuru Reference Azpuru2015). It is not surprising that participants in a study with service providers in Guatemala identified VAW as being deeply rooted in entrenched conservative and patriarchal values (Wands and Mirzoev Reference Wands and Mirzoev2022).

In terms of help-seeking behavior, survey evidence from Mexico suggests that only 15 percent of women who experienced violence reported the crime to the authorities (Htun and Jensenius Reference Htun, Jensenius, Brinks, Levitsky and Victoria Murillo2020). In this study, women cited “shame” and “keeping it quiet” as reasons for not reporting, which fit in Overstreet and Quinn’s (Reference Overstreet and Quinn2013) anticipated stigma model. Other reasons provided for not reporting IPV include low trust in authorities and insufficient information about available options for victims (Htun and Jensenius, Reference Htun, Jensenius, Brinks, Levitsky and Victoria Murillo2020). Our interviews with healthcare professionals in Guatemala largely confirmed these findings from Mexico. Focus group participants cited shame, fear, and scant resources as reasons for women to keep the violence “quiet.” Our participants also added that victims do not want to be blamed for the violence: the community always blames her, they said in unison. Thus, a powerful mix of poverty, inaccessibility of services, insufficient implementation of the law, impunity and corruption, sexism, and stigmatization prevents women experiencing IPV from seeking help.

Notwithstanding the codification of legislation aimed at eradicating violence against women in Guatemala and Mexico, both countries are unable or unwilling to make good on these promised rights to various degrees. In all, the absence of a functioning criminal justice system in both countries has contributed profoundly to the persistence of generalized insecurity, high rates of VAW, low institutional legitimacy, discretionary sentencing, and low help-seeking among victims (e.g., Frías Reference Frías2022; Wands and Mirzoev Reference Wands and Mirzoev2022). For these reasons, Mexico and Guatemala are appropriate cases to test our theoretical framework. That is, information about state efforts to combat VAW diffused through credible individuals should still shape stigmatizing opinions about victims, even in contexts of rampant impunity and widespread distrust in institutions and authorities.

Empirical Strategy

To test these hypotheses, we rely on an original experiment embedded in a public opinion survey conducted in Guatemala and Mexico in October 2024. In all, 1,204 individuals were surveyed. The online survey was administered through NetQuest, which maintains a large respondent pool by offering tokens that can be traded for goods in exchange for occasional participation in surveys. Subjects from all 31 states and the federal district took part in the survey in Mexico. However, due to NetQuest’s limited reach in Guatemala, only subjects from cities were sampled. Furthermore, due to the small panel, only 401 Guatemalans participated in the survey.Footnote ⁴ As a result, the Mexican sample is balanced across age, gender, socioeconomic status, and region of residence, mirroring the demographics of the general population. In contrast, the Guatemalan sample is wealthier and more urban compared to the general population. We recognize that this is a limitation of our Guatemalan sample, as research notes that women in Guatemala face more entrenched forms of social stigma when experiencing and reporting IPV in rural areas (Beck Reference Beck2021). We also rely on focus group interviews conducted in Guatemala in 2024 with 10 women from a local organization working with women’s health. We draw on illustrative insights from these interviews to substantiate our theoretical claims and contextualize the survey findings. We provide a detailed explanation of our fieldwork methodology and focus group procedure in the Supplementary Material (Box A1).

The theoretical argument proposes that in contexts of chronic impunity and weak institutions, credible anti-VAW messages can shape attitudes toward IPV victims, generating perceptions of lower societal stigma toward VAW. We propose that male authority figures, especially those with expertise in fighting crime and public health, might tap into a reservoir of subject matter credibility to transmit these messages to the public. We test this argument with a 2 × 2 experimental design with two additional control groups (Table 1). The vignettes manipulate the messenger’s gender and occupation (police officer or medical doctor).Footnote ⁵ Importantly, the aspects of the anti-VAW law in their messages are slightly manipulated to either prime respondents to the criminal aspects of the anti-VAW legislation or its public health components. The control groups either answer questions about VAW stigma without reading a vignette or after reading about a non-governmental organization (NGO) (no legislation primed). These messengers were selected based on the traditional authorities tasked with responding to VAW (the justice system and healthcare professionals), but also as a result of the findings of our fieldwork in Guatemala. In the towns we visited, the police are largely absent. However, NGOs and medical facilities are often present. Nevertheless, as articulated above, we expect police and healthcare messengers to exert stronger influence on opinions given their public authority, subject matter expertise, and power to change outcomes at the local level. Vignettes and questions are available in the online appendix. Additionally, Table A1 in the supplementary material presents balance tests that verify randomization.

Table 1. Vignettes

After reading a randomly assigned vignette, respondents were presented with several statements gauging attitudes toward VAW and views of the public’s stigma toward IPV victims. On a scale from 1 (strongly disagree) to 5 (strongly agree), stigmatizing attitudes held by the participants were assessed with the following statements: 1) Intimate partner violence is a private matter and society should not interfere (mean: 1.8; SD: 1.12); 2) Men do not hit women for no reason (mean: 2.42; SD: 1.33); and 3) it is more serious when a stranger kills a person during a robbery than when a man kills his wife because of jealousy (mean: 2.3; SD: 1.4). These three items were combined into an index assessing personally held beliefs about IPV. For views on the extent of public stigma toward IPV, respondents were asked their level of agreement or disagreement with the following statements: 1) Society does not take victims of intimate partner violence seriously (mean: 3.6; SD: 1.2); and 2) Society thinks victims of intimate partner violence should keep the experience secret (mean: 2; SD: 1.3). These statements were also combined into an index assessing perceptions of public stigma. These items were inspired by previous research on the components of the stigma (e.g., Overstreet et al. Reference Overstreet, Willie and Sullivan2019; Day et al. Reference Day, Edgren and Eshleman2007).

The models control for several relevant individual-level characteristics and attitudes, asked before the experimental vignettes. Importantly, models account for trust in political institutions and views of impunity. Indeed, as expected, the sample displayed very high levels of distrust in institutions. For example, 76.25 percent of the combined sample selected no or very little trust when asked about the police. When asked how likely it would be for someone to be punished by the authorities if they committed a crime, a plurality of the sample (43 percent) selected “not very likely.” To test hypotheses 2 and 4, we include gender as a moderator in the models, along with its interaction with the experimental treatments.

Similarly, to test the effect of sexism on views of IPV, we use the benevolent and hostile sexism scale developed by Glick and Fiske (Reference Glick and Fiske1996). The hostile sexism index on a 5-point scale included: 1) women are too easily offended, 2) most women fail to appreciate all that men do for them, and 3) women exaggerate their problems (index mean = 3.07 SD=1.04). Benevolent sexism index included: 1) many women have a quality of purity that few men possess, 2) men should be willing to sacrifice their own well-being in order to provide financially for the women in their lives, and 3) in a disaster, women should be rescued first (index mean= 3.02; SD= 0.9). Benevolent sexism is used as a control in the models, while hostile sexism is included as a moderator to test H5.