Violence against women (VAW) is an enduring obstacle for the achievement of gender equality across the world—violating women’s freedom and bodily integrity. Among the many forms of gender-based violence, intimate partner violence (IPV) is the most common, affecting at least one in three women in the world (United Nations 2015). Globally, IPV is also responsible for most female homicides (United Nations 2015). The negative consequences of VAW for victims have been extensively documented by researchers across disciplines. Scholars have demonstrated that victims of VAW suffer from a slew of physical and mental health problems (e.g., Jordan et al. Reference Jordan, Campbell and Follingstad2010). VAW also imposes costs for society as a whole—lowering income and social capital and ratcheting up medical costs (e.g., Fearon and Hoeffler Reference Fearon and Hoeffler2015).
But does intimate partner violence influence survivors’ political opinions? Despite extensive research on the effects of crime victimization on political attitudes and behaviors (e.g., Bateson Reference Bateson2012), we know surprisingly little about this relationship. Research in political science has largely focused, for example, on the determinants of policy adoption and its effects on VAW outcomes (e.g., Htun and Weldon Reference Htun and Weldon2012; Hudson et al. Reference Hudson, Bowen and Nielsen2011). This article contributes to this research by examining IPV survivors’ evaluations of government performance.
Drawing from theories of public opinion (e.g., Campbell Reference Campbell2012), I hypothesize that IPV survivors’ access to specialized public services shapes their evaluations of government performance. I argue that when states launch efforts to combat VAW, they transfer gender-based violence from the private to the public sphere, fostering a belief that government is responsible for that domain. Indeed, research has shown that the public adjusts their opinions on issues based on what they have learned from salient government initiatives on those same areas (e.g., Gusmano et al. Reference Gusmano, Schlesinger and Thomas2002). Researchers typically define salient issues as those that receive the most attention by political parties, policymakers, and news outlets (e.g., Rohrschneider and Whitefield Reference Rohrschneider and Whitefield2009). Salient legislation signals to survivors that the state should safeguard their rights as victims. Therefore, I argue that salient IPV legislation alerts survivors to the state’s role in addressing the negative consequences of victimization.
In most developing countries, however, the provision of specialized services for victims is limited, even when IPV legislation exists (e.g., Neumann Reference Neumann2017). Poor implementation of services amplifies the gap between law and practice, creating a service deficit where many survivors are unable to access adequate public services. Support services might not exist, might be located distant from one’s residence, or might deny access to survivors. As such, a shortfall in service provision might disappoint IPV survivors’ expectations of assistance heralded by the legislation. Survivors’ inability to access assistance might engender a deep sense of injustice, which activates negative emotions (e.g., Batson et al. Reference Batson, Chao and Givens2009).
Although data limitations of this study do not permit investigation of mechanisms, I suggest that negative emotional responses, and anger in particular, are potential pathways connecting the inability to use services to negative political attitudes. This is because anger is an emotion that fuels punitive attitudes (e.g., Lambert et al. Reference Lambert, Eadehb, Hanson and Olsen2019). As such, I expect IPV survivors who are unable to access services to downgrade their confidence in the state’s ability to support IPV survivors and to evaluate the central government’s performance negatively.
To test these hypotheses, this study relies on data from the 2013 Data Popular and Instituto Patricia Galvão survey conducted in Brazil, which is nationally representative of the adult population in urban areas. These data are complemented with findings from qualitative interviews conducted with survivors and service providers in Brazil. I argue that Brazil is an ideal case study, due to its very well known and comprehensive legislation on IPV (United Nations 2012) mandating the implementation of specialized services at the local level. Despite these efforts, however, Brazil still has alarming rates of VAW (Brysk Reference Brysk2018). I find that IPV survivors who had no access to specialized services are less likely to think that the state assists IPV victims and more likely to rate the performance of the central government negatively compared to nonvictims. IPV survivors with access to services, conversely, share similar views with nonvictims in both areas. As such, the empirical findings lend support to the core argument of this article. Ultimately, women’s inability to claim their rights to justice and recovery compromises the quality of democratic governance by preventing women from living lives free of fear and violence.
Given the challenges associated with observational data, which limits the ability to draw conclusions regarding directionality, the study takes several approaches to rule out reverse causality. In addition to qualitative evidence, a series of tests were run analyzing the characteristics of victims who have sought services and their municipalities, victims’ attitudes toward VAW services, and a set of placebo tests. The results lend additional support to the main findings. They also suggest that while the existence of services in a municipality increases victims’ likelihood of using them, their mere presence does not render these services accessible to all victims. Results show that Black women who have been victimized are much less likely than their white counterparts to use these institutions. Indeed, researchers have documented the stark racial inequalities permeating Brazilian society (e.g., Mitchell-Walthour Reference Mitchell-Walthour2017). Therefore, an integral strategy to combating VAW must consider the structural constraints that might undermine disadvantaged women’s access to support services. Ultimately, this study offers insights for the design and implementation of inclusive policies that advance victims’ democratic rights.
The Public and the Private: Generalized Crime and Violence Against Women
Extensive research on crime victimization has shown that being a victim of generalized crime (e.g., robbery, burglary) has important implications for people’s political attitudes. Scholars have found, for example, that crime victims evaluate regime performance negatively (e.g., Booth and Seligson Reference Booth and Seligson2007). Additionally, research suggests that crime victims participate in politics at higher rates than nonvictims (Bateson Reference Bateson2012). Yet the effect of VAW victimization on political attitudes remains understudied (Bateson Reference Bateson2012).
This study focuses on the relationship between one type of VAW and political attitudes: intimate partner violence. IPV is typically defined as VAW within intimate relationships consisting of physical or sexual aggression, or psychological or emotional abuse (United Nations 2012).Footnote 1 I argue that IPV victimization shapes survivors’ political views, but that the pathways through which IPV influences victims’ political attitudes are quite distinct from those of generalized crime. This is because until very recently, IPV was hardly recognized as a crime since states took little action to intervene in the private sphere (Htun and Jensenius Reference Htun and Jensenius2020). Generalized crime, on the other hand, has been considered a public issue since the formation of modern states (e.g., Tilly Reference Tilly, Evans, Rueschemeyer and Skocpol1985). Thus, the distinction between private and public violence refers to the physical location of victimization and the extent to which states were involved in preventing and criminalizing it.
I argue that the public versus private dichotomy that separates generalized crime from VAW is crucially important in understanding the ways through which these experiences shape attitudes. While generalized crime tends to affect people in public spaces and threaten the notion of private property that underpin free market democracies, IPV is often perpetrated by someone known to the victim in private spaces (e.g., Watts and Zimmerman Reference Watts and Zimmerman2002). Due to its public nature, protecting citizens against generalized crime has always been associated with some of the most basic functions of the state (e.g., Tilly Reference Tilly, Evans, Rueschemeyer and Skocpol1985). Conceived in this way, it is reasonable to expect public safety to be a major consideration for citizens when forming political attitudes (Carreras Reference Carreras2013). Thus, victimization on its own can engender negative political attitudes, as the state would have failed to prevent crime.
Contrary to the way they addressed generalized crime, states have historically turned a blind eye to violence taking place in the private sphere. In fact, states have often enacted legislation that institutionalized gender-based violence; for example, by protecting rapists from criminal charges if they married their victims (e.g., Htun and Jensenius Reference Htun and Jensenius2020). This notion that the private sphere is immune from state interference shielded aggressors from prosecution but also kept IPV from the public’s concern. Given IPV’s perceived private nature, the state’s responsibility to counter it is not necessarily evident to citizens in the absence of salient state action to remedy it—and even with legislation, entrenched attitudes toward IPV take time to change (e.g., Córdova and Kras Reference Córdova and Kras2022). As result, victimization on its own is unlikely to engender negative political attitudes. However, in contexts of salient laws, victims rely on the state’s response to their victimization in forming their political opinions.
There is another very important element of this distinction between VAW and generalized crime that warrants close attention. Among all types of crime victimization, VAW is the only one in which society attributes blame to victims (Watts and Zimmerman Reference Watts and Zimmerman2002). Lack of state intervention in private violence, combined with patriarchal norms, underlies the problem of VAW and the subsequent blaming of victims (e.g., Overstreet and Quinn Reference Overstreet and Quinn2013). Accordingly, scholars posit that this structural stigmatization of IPV can trigger feelings of guilt and shame among survivors (Overstreet and Quinn Reference Overstreet and Quinn2013). Because of this stigma, salient state action on VAW is of critical importance in signaling to survivors that they have victims’ rights and that the blame lies solely with the aggressor.
In recent decades, feminist movements have brought to light the public nature of gender-based violence (Htun and Weldon Reference Htun and Weldon2012). Activists have emphasized that VAW is rooted in the unequal standing of men and women in almost all societies, which requires policies that remedy it. As a result of decades of feminist mobilization at the international and domestic levels, legal responses to VAW have grown markedly. In 1993, only 8 countries criminalized IPV; today, more than 140 states have adopted some form of legislation (Brysk Reference Brysk2018, 193). The extent to which the law offers protections and resources to victims, as well as the degree of implementation, however, vary drastically across countries (e.g., Neumann Reference Neumann2017; Franceschet Reference Franceschet2010).
This discussion illustrates that VAW is quite different from other forms of generalized crime that have always been considered central political issues. Based on these considerations, I argue that IPV survivors would be more likely to link their needs as survivors to the state when comprehensive legislation has been passed and public services for victims have been established. Through these measures, the government signals to survivors that the state is responsible for safeguarding victims’ rights. This way, survivors’ inability to reach specialized assistance might foster the belief that the state has failed to fulfill its responsibility to victims.
Intimate Partner Violence Victimization and Political Attitudes
I argue that salient policies on IPV are particularly important in helping survivors form political opinions because of the private nature of IPV. Salient and comprehensive legislation that criminalizes IPV places IPV firmly in the public sphere, which prompts survivors to connect their needs to the state. This argument builds on Easton (Reference Easton1975), who has argued that specific support emerges when the fulfillment of citizens’ demands can be expected from authorities. By criminalizing IPV and establishing resources of prevention and response, states signal to women who have been victimized that they have a right to justice and recovery. Salient state action on VAW, in turn, enables survivors to define their identity as victims in terms of “rights” (Merry Reference Merry2003)—thereby raising their expectations regarding the availability of resources serving victims.
Indeed, scholars have suggested that public policies disseminate the types of information citizens need when forming opinions about issues (e.g., Campbell Reference Campbell2012). Public policies send signals to the public about the appropriateness of the state’s involvement in an issue area (Gusmano et al. Reference Gusmano, Schlesinger and Thomas2002). As intended beneficiaries of IPV policies, survivors might be more likely to rely on information derived from the law to form their political opinions, as do policy beneficiaries in general (e.g., Campbell Reference Campbell2012). Unsurprisingly, people’s evaluations of government performance depend on the government’s performance on issues they care about (Fournier et al. Reference Fournier, Blais, Nadeau, Gidengil and Nevitte2003). Supporting this thesis, Córdova and Kras (Reference Córdova and Kras2020) find that the implementation of women’s police stations designed to respond to VAW improves women’s evaluations of police performance. Thus, policies communicate to the public the issues that lie within the government’s responsibility, thereby shaping political opinions.
Informed by these findings, I argue that subsequent to the enactment of salient IPV legislation, survivors see IPV as a policy domain in which the state must respond effectively. Comprehensive IPV laws generally criminalize IPV, necessitating reforms in the criminal justice system (United Nations 2012). Additionally, comprehensive laws require the implementation of various services for victims, including shelters, crisis centers, and medical facilities (United Nations 2012). The most thorough IPV legislation also mandates awareness-building campaigns relating to VAW (United Nations 2012). This is the case in Brazil. The information derived from salient efforts to combat VAW might promote the belief among survivors that the state must protect victims’ rights by facilitating their access to justice and recovery. However, limited implementation of services might disappoint IPV survivors who expect the state to protect their rights as victims.
Indeed, in developing countries, governments tend to lack the financial resources or political will required for widespread service provision, creating a service deficit in many regions (e.g., Neumann Reference Neumann2017). In Brazil, women’s police stations and VAW shelters have been established in about 8.3 and 2.4 percent of municipalities, respectively (Araújo and Gatto Reference Araújo and Gatto2021). The absence of safe and specialized institutions, or the difficulty in accessing them, might leave abused women with no outlets to address the emotional consequences of trauma, seek protection, or mobilize resources to leave an abusive relationship. The inability to receive specialized services guaranteed in the VAW legislation might deeply disappoint survivors, generating a profound sense of injustice.
The fieldwork for this study shows, for example, that shelters are often inaccessible to victims with children—limiting the number of children allowed to stay there at any given time. Another disturbing example is how unreliable statewide hotlines might be. Social workers at a state-funded crisis center told me the VAW hotline, housed at the crisis center, often goes without working for months at a time because the state government does not pay the phone bill. The hotline is an important way that women can access information about the places they can go for help. Specialized public services for women, even when they are available, receive precarious funding, which can render these services inaccessible in practice. Thus, the state’s failure to make services accessible to all IPV survivors and to live up to the law can produce a sense of injustice, which can fuel negative emotions like anger (Krehbiel and Cropanzano Reference Krehbiel and Cropanzano2000).
Research has consistently shown that IPV victimization is highly associated with emotional distress, post-traumatic stress disorder (PTSD), depression, and anxiety (e.g., Jordan et al. Reference Jordan, Campbell and Follingstad2010). Symptoms associated with PTSD, which can persist for years, include negative emotions (e.g., Keshet et al. Reference Keshet, Foa and Gilboa-Schechtman2019). Anger, in particular, is a common emotional response to trauma among VAW survivors, compared to other nongendered crimes (e.g., Amstadter and Vernon Reference Amstadter and Vernon2008). Anger is caused by appraisals that one has been harmed, and that another party is responsible—based on wrongdoing that someone has experienced firsthand (e.g., Batson et al. Reference Batson, Chao and Givens2009; Lambert et al. Reference Lambert, Eadehb, Hanson and Olsen2019). As intended beneficiaries of VAW policies, victims’ inability to claim rights and to access needed resources might compound negative emotions.
Extant research has shown that negative emotions shape political attitudes. For example, using survey experiments, Mullinix et al. (Reference Mullinix, Bolsen and Norris2021) find that information on police brutality prompts negative emotional responses among the public. These negative emotional responses, in turn, engender lower levels of approval of the police and greater skepticism that the police do what is right. Similarly, as a prototypical cause of anger, feelings of injustice shape people’s opinions about government, foreign policy, and political parties (e.g., Roseman Reference Roseman and Lench2018). Informed by these findings, I argue that once the state declares responsibility in providing outlets for IPV survivors to address the countless negative consequences associated with victimization, lack of access to such support cultivates a sense of injustice among survivors, who find themselves unable to claim rights. Feelings of injustice amplify negative emotions such as anger, generating negative evaluations of the government.
It is important to highlight that IPV survivors can seek help from other institutions besides the state. Existing research shows that VAW survivors might receive support from private institutions, the church, women’s organizations, or self-help groups (e.g., Kreft Reference Kreft2019). I argue, however, that even if women seek help elsewhere, the lack of access to public specialized services in light of salient IPV legislation still prompts them to develop negative views of government performance. Moreover, self-help groups and other nonprofit organizations might not be as effective as specialized public institutions in providing victims with access to justice. Needless to say, a shortfall of public services is particularly damaging for survivors who might lack the means to access private help. Thus, even when they have received support elsewhere, survivors might still feel a sense of betrayal and injustice from not having access to state resources when salient IPV legislation exists.
This theoretical framework leads to the following testable hypothesis about the likely association between lack of access to specialized services and IPV victims’ political attitudes:
Hypothesis 1. IPV survivors who have not used specialized public services will be less likely to believe that the state supports victims, compared to female nonvictims.
I further argue that perceptions that the state does not support victims would result in more negative views of the central government’s performance in particular—as comprehensive VAW legislation tends to be adopted and publicized mainly at the national level. In Brazil, for example, the executive branch promoted and enacted the domestic violence law. While the VAW law, in the context of decentralized and federal states, mandates the collaboration between the three levels of government—federal, state, and municipal—in implementing it, survivors are still more likely to perceive the central government as responsible for protecting victims’ rights (e.g., León Reference León2012). The three levels of government place emphasis on different responsibilities in implementing the law, but they are not completely independent of one another, making it difficult for citizens to correctly evaluate the levels of government according to their functions (e.g., Arceneaux Reference Arceneaux2006; León Reference León2012).Footnote 2 For example, Article 8 of the Maria da Penha Law in Brazil (Lei no. 11.340), a federal law on VAW, states that the responsibility for the implementation of the law lies within the three levels of government without clearly defining the programs and policy areas each should execute.
Research shows that people tend to attribute more responsibilities to the central government than to local and state governments and can hardly differentiate federal and local competencies (León Reference León2012). This difficulty in identifying competencies generates a “national politics bias”—in which citizens attribute more responsibility to the federal government and evaluate the local government based on the performance of the central government (e.g., León Reference León2012; Gélineau and Remmer Reference Gélineau and Remmer2006). As such, even in contexts like Brazil, where state and local governments have important powers (e.g., Montero Reference Montero2001), I still expect survivors to associate satisfaction or dissatisfaction with the implementation of the VAW law, a federal law, with the central government. This discussion leads to a second hypothesis.
Hypothesis 2: IPV survivors who have not used specialized public services will be more likely to rate the performance of the central government negatively compared to female nonvictims.
Data and Methods
To test the hypotheses, this study uses data from the Data Popular and Patricia Galvão Institute survey conducted in May 2013 on the perceptions of Brazilians on violence against women. The sample consists of face-to-face interviews conducted in 100 randomly selected municipalities across the five regions of the country. The strategy for the randomization of municipalities ensured the representation of all large cities in the sample, but also the inclusion of small municipalities.Footnote 3 The analysis was restricted to female citizens only, generating a sample of 779 respondents, 136 of whom were IPV victims. It also relied on qualitative data collected from 40 in-depth interviews with survivors of IPV and VAW public service providers in Brazil.
Public Opinion Data
The two dependent variables in this analysis gauge perceptions of state support for IPV victims and more general evaluations of performance of the central government.Footnote 4 These measures correspond to the following questions available in the survey, respectively:
-
1. Where 0 means no and 1 means yes, do you think that women who have experienced domestic violence can count on the support of the state in denouncing the aggressor?
-
2. On a scale of 1, very bad, to 10, very good, how do you evaluate the government of the current president?
I argue that the first measure is adequate to test the first hypothesis because it provides important insights into whether victims have confidence in the state to do what is right in coming to victims’ aid. This is a fine-grained measure of survivors’ evaluation of government performance in responding to victims’ rights. Likewise, the second question is adequate to test the second hypothesis because it refers to the level of government that has passed and publicized the main IPV law in Brazil. Thus, this second measure gauges victims’ government performance evaluations more generally. Crucially, Dilma Rousseff was president at the time of the survey, and she publicly declared intentions to devote resources to fight VAW. These considerations might indicate that Rousseff reduced uncertainty in tracing the central government’s responsibility in responding to VAW.
On the question assessing whether victims can count on the state for support, evaluations among women were quite divided: 47 percent said no versus a slight majority, 53 percent, saying yes. On the 1–10 scale, the mean evaluation of the current president’s government was 7.7. Most female respondents expressed positive evaluations of Rousseff’s government: 28 percent rated the central government as very good, while only 3 percent rated it as very bad.
The two main independent variables for this analysis are IPV victimization and whether survivors used specialized public services. The IPV victimization question asked women whether they have been victims of IPV committed by a partner or ex-partner. For the second main independent variable, a measure was created combining the IPV victimization question and an item in the survey that asked respondents whether they used any of the 8 VAW public services.Footnote 5 This independent variable was coded as a categorical measure: 0 (nonvictims), 1 (victims and users of VAW services), and 2 (victims and nonusers of VAW services). Among the 136 victims, 51 used specialized services and 85 used none.
In addition, the analysis controlled for knowledge of the IPV law (Maria da Penha Law) for both dependent variables, as level of knowledge of the law might affect both measures of political attitudes. It included a measure gauging support for Rousseff’s party, the Workers’ Party (PT), for both dependent variables. Although the survey did not ask respondents for their party affiliations, it did include a question on evaluations of Lula da Silva government, which provides a reasonable way to assess support for the Workers’ Party. Party identification in Brazil is notoriously weak, and the PT is the only mass party (Samuels Reference Samuels2006). Lula is PT’s most emblematic politician.
The models also controlled for individuals’ sociodemographic characteristics, including employment, race, educational level, and income. Income was measured using a household wealth indicator, since the nonresponse rate of the question assessing income is high—a common problem with representative surveys (Córdova Reference Córdova2009). The household wealth indicated used here was car ownership. Also included was a question on self-reported membership in social class in all models and a measure of social class of one’s neighborhood, as complementary measures gauging socioeconomic status.Footnote 6
Perceptions of safety were also included in all models, since they influence several types of political evaluations (e.g., Booth and Seligson Reference Booth and Seligson2007). Variables asked respondents whether they thought crime and VAW were punished in Brazil. Similarly, the models testing H2 controlled for evaluations of public security. The analysis further controlled for a variable asking respondents whether they knew the location of a women’s police station in their hometown, since these are best-known VAW policy instrument in Brazil. Given that acceptability of VAW might shape attitudes toward the state’s role in protecting victims, another measure assessed citizens’ level of intolerance toward IPV. Moreover, because Rousseff’s party enacted the Maria da Penha Law, views on VAW might also shape evaluations of her government.
To test hypothesis 1, on perceptions of state support for victims, the analysis estimated logistic regression models, given that the dependent variable is a dichotomous measure. To test the second hypothesis, on evaluations of the central government, the analysis estimated ordered logit models, since the dependent variable is an ordinal measure. The models include states fixed effects to rule out any regional effects on respondents’ attitudes.Footnote 7
Qualitative Interviews
In addition to the survey, I also rely on the findings from the extensive fieldwork I carried out in the south of Brazil in the summer of 2019. During fieldwork, I conducted 5 in-depth, semistructured interviews with survivors of IPV and 35 interviews with VAW public service providers, for a total of 40 interviews. Among the service providers, I interviewed lawyers, officers of the military and civil police, sheriffs of different women’s police stations, therapists, social workers, project coordinators, and community activists.Footnote 8
The interviews were conducted across 12 municipalities in the Southern Brazilian states of São Paulo, Santa Catarina, and Rio Grande do Sul. Participants were recruited using an email and phone script. I contacted all public organizations that offer services to women in situations of violence in those municipalities and requested an interview with a service provider. Survivors were recruited through two nongovernmental organizations for victims. I visited all institutions in which interviews were conducted. But to protect the privacy and safety of survivors, I interviewed victims using video calls from the platform WhatsApp.
The themes of all interviews were victims’ access to public services and their quality, the challenges service providers face in combating VAW, the condition of facilities, and the fluctuations in funding and personnel following changes of political parties in power in the state and local legislatures. Mental health, trauma, and emotional exhaustion were topics that emerged in every single interview, for both victims and service providers. Indeed, emotional fatigue, as well as depression was shockingly common among service providers, who often experience vicarious victimization. Last, I asked all participants about survivors’ experience with seeking help in general, their knowledge of the Maria da Penha Law, and their perceptions of state support for victims and trust in government. Interviews lasted between one and three hours.
Brazil as a Case Study
The passage of the Maria da Penha Law in 2006 marked a paradigm shift in Brazilian politics—placing VAW permanently on the political agenda. The law was passed as the result of decades of feminist mobilization that had pressured the Brazilian state to respond to VAW (Roggeband Reference Roggeband2016). The state did respond to some of these demands earlier, such as implementing women’s police stations, in 1985, aimed at working exclusively with victims of VAW. But the feminist struggle to compel the state to adopt national legislation addressing IPV was a much longer process (Roggeband Reference Roggeband2016).
The election of President Lula da Silva of the Workers’ Party in 2002 opened a window of opportunity for the feminist movement to demand federal legislation on IPV and attention to the Maria da Penha case. Maria da Penha had survived two attempted murders by her husband and was left paraplegic in 1983. The subsequent complaints she filed against her husband languished in Brazilian courts for decades (Roggeband Reference Roggeband2016). Due to this context of impunity, Maria da Penha, with the support of activists, filed a petition with the Inter-American Commission on Human Rights (IACHR) against the Federative Republic of Brazil in 1998 (Carone Reference Carone2018). In 2001, the IACHR found the Brazilian state responsible for human rights violations and urged the federal government to adopt measures to punish VAW (Carone Reference Carone2018). However, the Brazilian government complied with the IACHR’s recommendations and feminist demands only in 2004, when the executive branch under Lula proposed the Maria da Penha Law (Roggeband Reference Roggeband2016).
The Maria da Penha Law increased the costs of engaging in IPV for aggressors; created measures to empower victims, such as restraining orders; and included domestic violence in the federal penal code. The law expanded a system of integrated and specialized services designed to address the negative consequences of VAW specifically: women’s police stations, VAW crisis centers, specialized courts, special units in the prosecutors’ office, domestic violence shelters, a VAW hotline, and specialized medical units (e.g., Carone Reference Carone2018). Importantly, the law spelled out preventive measures against VAW, including the creation of awareness campaigns to diffuse the law to “all society.” Not surprisingly, the Maria da Penha Law is considered a very comprehensive piece of legislation on IPV (United Nations 2012).
The interviews with IPV survivors and service providers indicate that IPV victims tend to be highly aware of the existence of the Maria da Penha Law. Service providers overwhelmingly agreed that the law is very salient, as one coordinator put it: “everybody knows about the Maria da Penha Law.” The survey evidence also lends support to this finding. Among female respondents, 32 percent of nonvictims and 28 percent of IPV victims reported that they knew a substantial amount about the law (Data Popular et al. 2013). Almost no female respondents reported being unaware of the law.Footnote 9 Therefore it is fair to suggest that the IPV legislation in Brazil is well known among the public.Footnote 10 However, the actual implementation of specialized services promoted in the IPV law is far from ideal. I argue that while the information generated by the law raises survivors’ expectations about their access to public services, in reality, these services are insufficient, nonexistent, or overstretched.
Given Brazil’s trajectory in adopting and implementing salient legislation on IPV, I argue that it offers a particularly useful case to analyze the relationship between IPV victimization and political attitudes.
Empirical Findings
Perceptions that the State Supports Victims
Table 1 presents the findings of a set of logistic regression models testing H1 and H2. Models 1 and 2 test the correlation between IPV victimization, as well as specialized services usage, and perception of state support for victims (H1). Baseline model 1 shows that IPV victims are, on average, less likely to think that survivors can count on the support of the state to denounce aggressors. This finding is statistically significant at the p < .05 level. However, model 2 indicates that IPV survivors and nonusers of services are driving these results. Victims of IPV who have not used services are significantly more likely than nonvictims to believe that the state does not support survivors (p < .05). IPV victimization on its own does not seem to be correlated with negative perceptions of state support, except when survivors do not use services designed to respond to VAW.
The Effect of Intimate Partner Violence on Political Attitudes
***p < 0.01, **p < 0.05, *p < 0.1
Standard errors in parentheses.
Notes: These models control for states’ fixed effects. These models also control for self-reported social class, age, and whether respondents believe VAW is punished, which are all insignificant. The complete results are presented in table A4 of the online appendix.
Panel A in figure 1 illustrates the substantive findings from model 2 graphically. The predicted probability of female nonvictims of IPV believing that the state supports IPV victims in denouncing aggressors is 56 percent. This probability declines for victims with access to support services to 50 percent. The probability of believing that the state supports IPV victims diminishes substantially for victims who have not used specialized services. These survivors are 42 percent likely to express confidence in the state’s ability to come to victims’ aid. This is a statistically significant difference of 14 percentage points compared to nonvictims of IPV (p < 0.01). These results suggest that the lack of access to services in contexts of salient IPV legislation is correlated with IPV survivors’ skepticism about the state’s performance in ensuring victims’ rights.
The Effect of IPV Victimization and Access to Specialized Services on Attitudes
The in-depth interviews with IPV survivors and specialized service providers in Brazil corroborate these findings. A common theme throughout the interviews was precisely the lack of implementation or the inaccessibility of services. For example, when asked what she thought about the Maria da Penha Law, an IPV survivor stated that she knew about the law, but added “for me this law is talk, I could not find help.” Another IPV survivor shared a similar experience: “I knew about the law, but the state did not protect me, the shelter ran by the girls [activists] did.” These two survivors also told me that they felt a deep sense of injustice when they were going through the process of help seeking.
Relatedly, victims told me that they felt incredibly isolated in the process of help seeking, often having to find help outside the state. Indeed, one victim, who is Indigenous, eventually left Brazil to escape her extremely violent ex-husband after trying to seek help from public services countless times. She could not find a safe place to stay in her town, which lacked an IPV shelter, and did not find support at the local Women’s Police Station, which denied her help for “lack of evidence” against her ex-husband, who was physically abusing her, stalking her, and threatening her. Her case illustrates that VAW services not only are inaccessible for many women but can also revictimize them—especially women who have faced historical discrimination. When I asked if she used public services, she said no, since they refused to help her; she added, “the Maria da Penha Law does not exist in my state.”
I also spoke to two military police officers who work exclusively with the implementation of the Maria da Penha Law. They told me that the lack of services for victims in their region forces them also to serve as “therapists.” When asked about their role in combating VAW, they responded that they help to restore victims’ trust in the system, especially the police. They, too, expressed skepticism about the state’s willingness to support survivors. Likewise, an NGO coordinator emphasized that by the time survivors arrive at the NGO facility, they no longer think they can rely on the state for help.
Thus, in general, the interviews did suggest that survivors encounter many obstacles in accessing their rights and that many of them end up with no specialized assistance. Others look for help elsewhere. But their disappointing trajectory in trying to access their rights damages their political evaluations. The interviews show that even service providers overwhelmingly expressed low confidence in the states’ willingness to come to victims’ aid.
Performance Evaluations of the Central Government
Models 3 and 4 in table 1 present the results of the ordered logistic regression models testing H2. Baseline model 3 shows that on average, IPV victims have more negative evaluations of the central government compared to nonvictims. This result is statistically significant at the p < .05 level. However, this result masks the relationship between access to public services for victims and evaluations of the government. As we can see in model 4, survivors who report not having used specialized services are significantly more likely than nonvictims to evaluate the performance of the central government negatively (p < .01). Conversely, IPV victims who have had access to services are not more likely than nonvictims to evaluate the central government negatively. IPV victims appear to downgrade their evaluations of the central government when they do not have access to services designed to safeguard their rights as victims.
Predicted probabilities are further estimated and the results presented graphically in panel B in figure 1. The predicted probability of female nonvictims’ having high evaluations of the central government is 29 percent, while for IPV victims who have used VAW public services the probability is 27 percent. However, evaluations of central government performance diminish even further for IPV victims who have not used specialized public services. Their predicted probability of having high evaluations of the central government is 21 percent (p < 0.01). This is a statistically significant difference of 8 percentage points compared to nonvictims. In sum, IPV victimization, coupled with the lack of access to support services for survivors, appears to be consequential for women’s evaluations of performance of the central government.
Other notable findings from the analysis in table 1 include the relationship between the awareness of a women’s police station (WPS) location and perceptions that the state supports IPV victims. Women who know where a WPS is located in their town are more likely to think the state aids victims than women who are not aware of the location of a WPS (p < 0.05). These findings might suggest that credible political action aimed at promoting women’s safety contributes to women’s positive evaluations of the state’s performance in combating VAW.
Additional Analyses
A core assumption of my theory is that IPV survivors who reported not using any of the public services for VAW had no access to them. Data limitations, however, do not permit further investigation into victims’ reasons for refraining from using services. Therefore it is worth examining the validity of the core assumption of this study. The following analysis first explores the characteristics of survivors who have and have not sought help from specialized services, as well as the characteristics of their municipalities. Then it examines whether survivors with and without access to services share similar views on women’s police stations—the most publicized institution for VAW survivors in Brazil. It also conducts a set of placebo tests to rule out the possibility that IPV survivors who have not used specialized services are simply less trusting of public services in general.Footnote 11
Individual and Municipal Characteristics
This survey does not allow for testing why some survivors have not made use of specialized services, but I can explore possible explanations by looking at whether certain groups of women are more likely to use them. Likewise, I can analyze empirically whether victims in municipalities with established services are more likely to use them. Accordingly, I estimate a multilevel model predicting access to specialized services that controls for municipal- and individual-level characteristics.Footnote 12 The results indicate that Black women are, on average, less likely to receive specialized support than white women. Black women who have been victims of IPV are only 8 percent likely to have received assistance from specialized public facilities (see figure 2, panel A). By contrast, white IPV victims are 44 percent likely to have received specialized assistance. Strikingly, this is a statistically significant difference of 36 percentage points (p < 0.01). Survivors who identify themselves as multiracial are 29 percent likely to have accessed services for victims; however, this result is not statistically significant.
Individual and Municipal Characteristics Predicting Victims’ Access to Services
These results demonstrate that Black women who have experienced IPV are systematically excluded from having access to specialized services for victims. Afro-Brazilians have historically faced obstacles in attaining housing, health care, education, and economic opportunities (e.g., Mitchell-Walthour Reference Mitchell-Walthour2017). Against a backdrop of persistent racial discrimination and economic inequality, Black women who have been victimized might face additional barriers that undermine their access to services, compromising their health and safety. Indeed, research has demonstrated that femicide rates in Brazil are much higher among Black women than white women (e.g., Oliveira et al. Reference Oliveira, Rodrigues and Aguiar2016). As femicide is often the last stage of a cycle of violence, preceded by other forms of VAW, the high rates of femicide among Black women illustrate the importance of a robust, accessible, and inclusive public service network for women in situations of violence. VAW services, including the police, play a crucial role in preventing domestic violence from escalating (e.g., Saxton et al. Reference Saxton, Jaffe and Olszowy2020). The lack of access to services has real consequences for the lives of Black women in Brazil.
The network of public resources for IPV survivors is limited even in the municipalities that provide them. The ones that do exist are typically situated in more centralized locations, which are generally distant from marginalized communities on the outskirts of town. Indeed, during my fieldwork, service providers often voiced concern about marginalized women’s access to services. These practitioners expressed frustration with the fluctuation and instability of funds allocated to combating VAW, depending on the political parties in municipal, state, and federal legislatures. Budget cuts undermine their outreach efforts. A crisis center in a major city with millions of inhabitants that I visited lost its state-provided cars for outreach campaigns with the election of a more conservative governor—which is consistent with findings from existing research (Araújo and Gatto Reference Araújo and Gatto2021). Without funds to reach poorer or remote communities, the accessibility of VAW public services is seriously diminished, especially for Black women. In short, because results from table 1 indicate that Black women are not more predisposed to express negative attitudes toward government performance, I argue that Black women are less likely to use support services because such services are unavailable to them.
The existence of services at the local level is also important in predicting victims’ access to specialized services. Panel B of figure 2 shows the results of the municipal predictors from the multilevel model. The figure shows that IPV survivors in cities with shelters, WPS, and crisis centers are 41 percent likely to use services, while only 19 percent of victims in places with no services are (p < 0.01). Survivors in places with both a WPS and a crisis center or only a crisis center are 39 percent and 45 percent likely to have used specialized services, respectively (p < 0.01). Furthermore, survivors in towns with only a WPS are 29 percent likely to have used services, although this is not statistically significant.
These findings suggest that crisis centers might be more effective at reaching out to the population that needs them. While WPS are the most widely known public service for victims, many municipalities with WPS are not able to increase IPV victims’ access to them. WPS seem to be much more effective when other support services for victims are also present—highlighting the importance of a network of support services at the local level. With these findings, I am more confident in the argument that survivors who did not use services do not have access to them.Footnote 13
Views on the Effectiveness and Utility of WPS
I also ran a set of logistic regression models analyzing IPV victims’ attitudes toward WPS. If IPV victims with and without services share similar views on the potential benefits of services, we can rule out the possibility that some victims avoid services due to their negative views on them. Results show that while both IPV victim groups are significantly less optimistic than nonvictims about the effectiveness of WPS (p < 0.01), they still agree that WPS are the most effective specialized service. Remarkably, the predicted probability of nonvictims’ reporting that WPS are the most effective support service is 89 percent. This probability decreases to 70 percent for victims who have used services and to 78 percent for victims who have not used them. In general, female citizens consider WPS the most effective public institution for victims in Brazil.Footnote 14 This suggests that survivors who have not used services are still aware of the potential benefits of public support services addressing VAW.Footnote 15
The analysis further examined IPV victims’ views on what kinds of assistance other IPV victims should seek first when deciding how to cope with the aggression. The results of a logistic regression model show that nonvictims and both IPV victim groups are equally as likely to select WPS, rather than seeking help from friends and family or regular police. These results indicate that IPV victims, regardless of their access to services, are equally as likely to see support services as effective in combating VAW.
Placebo Tests
To rule out the possibility that IPV survivors and nonusers of VAW public services are less trusting of public institutions in general, leading them to avoid public services for victims, a series of placebo tests examined IPV survivors’ evaluations of public health care, security, and transportation. The results show that IPV victims who have not used VAW services are not more likely than nonvictims to evaluate these three public services negatively. These results are presented in table A11 in the online appendix.
Discussion
These results lend strong support to my theoretical expectations. When states pass and publicize comprehensive federal legislation on VAW, survivors expect implementation of necessary resources at the local level. Efforts to combat VAW are essentially meaningless if women do not have safe spaces to seek shelter, legal assistance, mental health professionals, and justice. These resources must be accessible and thoroughly implemented for the political elite’s promises of combating VAW to be perceived as credible among survivors. A shortfall of public services is an indication that VAW legislation is merely cheap talk. Consistent with previous literature (e.g., Campbell Reference Campbell2012), the findings of this study show that in contexts of salient VAW legislation, survivors’ access to services has implications for their political attitudes. Further, as survivors with no services still indicate that they are important, and as their existence in a municipality in fact predicts access to them, the systematic lack of access to VAW services among Black women points to a structural story.
This study raises questions as to whose rights VAW legislation protects. If services are concentrated in certain areas within cities and political volatility leads to the unpredictability of budgets and resources, limiting services’ reach to communities far from the center, then we must ask how racism is built into the design and implementation of these policies. This compromises the democratic rights of Black women in Brazil by endangering their lives. Indeed, one of the most unsettling implications of the lack of access to services among Black women is their high rates of femicide compared to white women (Oliveira et al. Reference Oliveira, Rodrigues and Aguiar2016). In general, research suggests that the existence of services like WPS dampens rates of IPV and increases women’s sense of safety (Córdova and Kras Reference Córdova and Kras2020, Reference Córdova and Kras2022). The existence of services, then, is consequential for women’s lives, the prevalence of IPV, women’s perceptions of safety, and survivors’ evaluations of government performance.
I believe that these results speak to the quality of democratic governance, as the implementation of VAW laws effectively safeguards women’s right to live lives free of violence and fear. The state’s inability to make VAW laws effective can foster a democracy of “low-intensity citizenship” (O’Donnell Reference O’Donnell1993), where women, and Black women in particular, face constraints in accessing their democratic rights. When victimized women cannot claim the legal protections they are entitled to, essential rights and guarantees that constitute full democratic citizenship are jeopardized (O’Donnell Reference O’Donnell1993).
Findings from my fieldwork, which are consistent with previous literature (e.g., Kreft Reference Kreft2019), suggest that often survivors themselves, along with civil society, fill the void left by the state. In Brazil, the Olga Benário Movement, comprising primarily of feminist activists, lawyers, and survivors, has established “unofficial” crisis centers for victims of VAW across several cities in Brazil—like for example, the crisis center and shelter Mulheres Mirabal in Porto Alegre (Fernández Reference Fernández2021). They offer shelter, legal assistance, and economic empowerment classes to women fleeing violence. Some of these shelters receive assistance from the local or state government; for example, with electricity bills. Relatedly, many survivors and activists become Promotoras Legais Populares—or community victim advocates—aiding survivors in navigating the criminal justice system or finding resources to escape an abusive relationship. These examples illustrate the role of social movements as agents of change—pressuring, complementing, reinforcing, and even replacing the state.
In Brazil, the persistent advocacy of women’s movements eventually led to the creation and adoption of the Maria da Penha Law. Their efforts at the local level continue to pressure governments to create services for women, or they establish the services themselves. Anecdotal evidence from some of my interviews suggests that survivors whose frustration with trying to access services compels them to find support from women’s self-help groups later become involved with these causes. An extremely fruitful path forward would be to investigate how the negative evaluations of government performance heralded by the lack of access to services, and the subsequent sense of betrayal, mobilize women to join social movements and advocacy groups.
Conclusions
The empirical results of this study lend strong support to my hypotheses, along with clear policy implications. To effectively combat VAW, credible legislation must go hand in hand with budget allocations to guarantee an even implementation of support services. Not only must services be established to ensure victims’ rights, but they need to have the resources necessary to provide quality and effective support for victims. Moreover, a shortfall in the provision of a system of support services undermines the effectiveness of VAW legislation, which compromises victims’ health and safety. With the fifth-highest rate of femicide in the world (Brysk Reference Brysk2018), Brazil is a textbook example of how the gap between VAW law and practice has tangible consequences for women’s lives. As femicide generally materializes in a context of recurring violence, public support services can interfere with this cycle (e.g., Saxton et al. Reference Saxton, Jaffe and Olszowy2020). The accessibility of public services reveals the degree to which victims’ rights are enforced in actuality, thereby influencing survivors’ political evaluations.
The mere presence of limited resources in centralized areas within cities is not sufficient in enhancing all victims’ access to justice and recovery, as Black women are much less likely to access services than white women. A multitude of obstacles, including structural inequality, racial discrimination, and poverty, might undermine Black women’s ability to claim their rights (e.g., Mitchell-Walthour Reference Mitchell-Walthour2017). State action on VAW can increase marginalized women’s access to support services by devising exhaustive outreach campaigns that work closely with marginalized communities (e.g., Kasturirangan et al. Reference Kasturirangan, Krishnan and Riger2004). As such, an integral and inclusive strategy in combating VAW must also acknowledge and account for the intersecting obstacles Black women face that limit their access to specialized services.
The rise of far right populist Jair Bolsonaro to the presidency in 2019 constitutes a major threat to women’s democratic rights in Brazil. The dismantling of social policies for women and minorities that has been taking place since he assumed the presidency is gradually scaling back gains made in the expansion of democratic rights to historically marginalized groups. As this article has argued, Brazil’s legal guarantees for women in situations of violence are comprehensive and multifaceted. The problem is the lack of proper implementation of inclusive social services accessible to all victims. Bolsonaro’s government is intensifying the precariousness of these services and expanding the gap between VAW law and practice. His government cut millions from the budget of the Ministry of Women, Family, and Human Rights, which is responsible for implementing policies on VAW. In short, Bolsonaro’s opposition to social spending threatens even further women’s access to resources to escape violence (e.g., Hofmeister and Fleck Reference Hofmeister and Fleck2021).
While these challenges and setbacks are alarming, there are reasons to be hopeful. First, important VAW bills have been recently introduced and passed in the National Congress that would offer further guarantees to victims. An example is a bill that passed in the Brazilian lower house that offers legal guarantees against institutional revictimization to sexual violence survivors (Clavery et al. Reference Clavery, Calgaro, Garcia and Felipe Barbiéri2021). Regular introduction of VAW bills demonstrates that the idea that VAW is political has been institutionalized. In 2019 alone, 134 bills were introduced in the congress dealing with VAW—including bills seeking to expand WPS and other services (Kras Reference Kras2021).
Second, there has been organized resistance to Bolsonaro’s attacks to social programs for women among policymakers—from both left and right parties (e.g., Alegretti Reference Alegretti2021). And third, feminist organizations and civil society have been pressuring state governments to improve and expand social services for women in situations of violence for decades and continue to do so. For example, women activists in the city of São Leopoldo in southern Brazil have been demanding a WPS in their city since the mid-1990s. Their relentless efforts paid off in 2019 when the first WPS finally opened in the city. These examples suggest that the belief that VAW is political and warrants state intervention has developed deep roots in Brazilian society, which might protect social services for victims in the long run.
It is important to stress, however, the limitations of this study. To properly test some of the links and causal mechanisms proposed in this research, future scholarship using methodologies such as experiments embedded in public opinion surveys is needed. Additionally, this study has focused exclusively on Brazil, which is a unique case, as VAW is arguably the most salient women’s rights issue and services have been established for decades. Future research should include country cases that have less well known IPV laws and services. Mexico and El Salvador, for example, also have alarming rates of VAW (e.g., Brysk Reference Brysk2018). However, they have adopted services such as WPS much more recently (e.g., Ríos Reference Ríos2020). Variation in the extent to which state action on IPV is salient to the public would enable us to better understand the connection between IPV victimization, access to services, and political attitudes.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/lap.2022.26
Conflicts of interest
Helen Kras declares none.
