“Women fear the law more than abusers”: a study of public trust in health and legal response to violence against women in Santo André, São Paulo State, Brazil

Little is known about how perceptions of government affects women’s views of violence against women. This qualitative study examines women’s perceptions of violence against women laws and trust in the health and legal sector response in São Paulo State, Brazil. Thirty interviews with female residents of Santo André, 18 years and older, were audio recorded, transcribed, coded, and analyzed for themes using MAXQDA12 software. All women were familiar with at least one federal violence against women law. They expressed widespread lack of trust in the government. They viewed the legal sector as ineffective and felt violence against women laws have little impact, or exacerbate violence. Disdain for the health sector was less prominent than for the legal sector. A generalized lack of trust in government manifests as a lack of confidence in these laws for women seeking support through legal and health sectors. Adoption of legal protections for women is a significant accomplishment. However, health and legal sector violence against women response needs to consider perceptions of government, how violence against women laws are operationalized, and the role of each sector in ensuring women’s safety. Violence against women response efforts should include standardized intersectoral procedures including referral to accessible resources and means of screening women who may be at high risk for femicide. Violence; Gender and Health; Criminal Law; Government ARTIGO ARTICLE This article is published in Open Access under the Creative Commons Attribution license, which allows use, distribution, and reproduction in any medium, without restrictions, as long as the original work is correctly cited.


Introduction
Public trust in government plays a critical role in the success or failure of policies that require the cooperation and compliance of citizens 1,2 . In Brazil, an ongoing corruption scandal is negatively impacting trust in government 3,4,5 . Citizens believe most politicians are corrupt and act in their own self-interest; trust in government institutions remains objectively low 4,6 .
In 2006, the Maria da Penha Law on Domestic and Family Violence became Brazil's first national law to define and criminalize violence against women, and establish mechanisms to punish perpetrators 7,8,9 . As a result of massive public education campaigns, over 95% of Brazilians are aware of the law's existence 10 . Several studies suggest widespread awareness of the law has resulted in greater reporting 11,12 . Yet widespread impunity for crimes against women remains a major contributing factor to high female homicide rates in Latin America, including Brazil 12,13,14 . Following the law's passage, the female homicide rate experienced a slight decrease and then rose quickly in subsequent years; a similar pattern was observed among reported rates of violence against women 12,14 .
Half (50%) of Brazilians believe the way the justice system punishes perpetrators does not reduce violence against women, and 85% agreed women who denounce an abusive partner are at greater risk of murder 10 . These attitudes and perceptions demonstrate a lack of public confidence in the legal sector response to violence against women 8 . The inadequate legal sector response is attributed, in part, to the lack of human and financial resources devoted to addressing violence against women 15,16 . Though Brazil implemented the Femicide Act in 2015 to stiffen penalties for gendermotivated killings, it continues to be recognized as a country with limited sanction for the murder of women 17,18 . Despite the existence of progressive laws, more action is needed 19 .
Lack of trust in government institutions to protect the well-being of all citizens may also be found beyond the legal sector, including the health arena. In Brazil, a universal right to health is enshrined in the 1988 Federal Constitution. The Brazilian Constitution upholds the right to health through its public health system (Brazilian Unified National Health System -SUS), which is charged with providing free and comprehensive health care through a nationwide network managed at the municipal level. Unfortunately, SUS and its beneficiaries have been impacted by negative public views that were exacerbated by the recession, during which the public health budget was subject to several rounds of budget cuts at a time when demand for services and costs of treatment increased 3 . Additionally, Brazilian Ministry of Health appointments have been used to leverage political power among warring political parties 3 . Despite the wide-ranging health consequences of violence against women, only one in five of Brazilian women who have experienced domestic or family violence seek health care 10,11,12,14,15,16,20 . Low levels of health-seeking behaviors are attributable to feelings of stigma and shame 20,21,22,23,24 .
Health professionals may reproduce stigma and shame when they hold women responsible for their victimization 25 . They may also fail to acknowledge and address violence because they view it as a private matter, further contributing to the culture of silence 21,25 . Health professionals may refer women to other services to avoid addressing the issue themselves, thereby requiring women to repeatedly recount their violent experiences risking retraumatization 26 . Such attitudes and behaviors contribute to women's low expectations for support and assistance from health professionals and low levels of abuse disclosure 25,26,27 .
Given the context of broad government corruption, pervasive violence against women, and perceived impunity despite legal protections and comprehensive health services, the aim of this study was to examine the relationship between women's perceptions of federal violence against women laws and their trust in health and legal sector response to violence against women in Santo André, São Paulo State, Brazil. We hypothesize the overall lack of trust in government institutions relates to a lack of trust in not only the legal sector response to violence, but also the ability of the health sector to provide care for women experiencing violent relationships.

Methods
This analysis was part of a larger study examining individual experiences and community perceptions of violence against women. Detailed information on the methods of the parent study are available elsewhere 28 .

Study site
Santo André, a municipality located in metropolitan São Paulo, has approximately 710,000 inhabitants (Instituto Brasileiro de Geografia e Estatística. http://cod.ibge.gov.br/4BW, accessed on 23/ Jan/2019). Santo André has one specialized police station for women, three service locations for psychological and social support for survivors of violence against women, one women's hospital, one health unit focused on women's health (also one of our study sites) and four legal service locations whose services include legal counseling and services 29 .
It boasted the only Municipal Secretariat for Women's Policies (SPM) in the region from 2014 to 2016. The SPM was a government body charged with articulating, formulating, and implementing policies and programs that address gender inequality and promote gender equality 30 . It fostered intersectoral coordination and collaboration in addressing violence against women before it was absorbed into the Department of Social Policies in 2017 following the election of the Brazilian Social Democracy Party -part of a broad shift away from the progressive left-wing Worker's Party which previously dominated 30 . SPM was a key partner for this project.

Data collection
A native Portuguese speaker conducted 30 in-depth interviews with participants who self-identified as women, were aged 18 and older, and resided in Santo André. Participants were recruited using convenience and respondent driven sampling from public health clinics in three neighborhoods. Additional local partners included the Municipal Secretariat of Health, and the ABC School of Medicine (FMABC), who facilitated access to data collection sites.
Using an in-depth interview guide described elsewhere, participants were asked open-ended questions about individual experiences and community perceptions of violence against women, health care access and quality, and federal legislation addressing violence against women 28 . Followup questions, probing, and prompts were used to explore topics such as interactions with clinic staff and police officials. The World Health Organization (WHO) protocol for conducting research on violence against women was applied throughout recruitment and data collection processes to ensure the safety of everyone involved 31 .

Data management and analysis
In-depth interviews were audio recorded, transcribed verbatim in Portuguese, and de-identified. Data were coded and analyzed in Portuguese for deductive and inductive themes using MAXQDA12 software (https://www.maxqda.com/). The codebook was refined over multiple rounds of coding, with two Portuguese-speaking coders discussing discrepancies between each round; a subset of transcripts was coded by a second coder in the final iteration to ensure consistency. For the purposes of this analysis, we examined the relationship between the violence against women, law, and healthcare codes using means of modified grounded theory 32 . We performed multiple close reads of the coded segments for the three codes through which inductive themes were identified. We developed thick descriptions for each theme for developing our resulting theory.

Study ethics
The study protocol was reviewed and approved by Emory University's Institutional Review Board (Atlanta, United States), the Santo André municipal government, and Plataforma Brasil, the Brazilian national institutional review board (CAAE 57344616.0.0000.5484). All participants provided written informed consent before each interview and were informed they could withdraw from the study at any time. There was minimal risk to participants given that participation and data were kept confidential. All participants were given violence against women-related educational materials and contact information for the clinic-based support person responsible for violence against women case management.

Results
Thirty in-depth interviews were conducted with women at three public health posts in Santo André; all women were familiar with at least one federal violence against women law. Among the participants, 30% (n = 13) reported personal violence against women experiences, consistent with previous studies 14,22 . As depicted in Figure 1, participants described an overall distrust in the Federal Government characterized by a perception of corruption and inefficacy. This distrust seemed to relate to negative views of violence against women laws -including legal sector inefficacy -and finally mixed satisfaction with the health sector and its utility in responding to violence against women cases.

The government is not trusted and the legal system is ineffectual
In addressing violence against women, participants described an overarching distrust in the broader government often referred to as "the law" or "politics" without clear distinctions between strata of government. This was linked to views of inefficacy, corruption and scandals, including as high as the office of the president. This distrust was sometimes described as a hopelessness, including futility in basic political processes like voting, which is mandatory in Brazil. One participant described how this feeling influences her civic engagement:

Figure 1
Proposed model for transfer of distrust across government, legal and health sectors in Brazil.
Cad. Saúde Pública 2020; 36(10):e00114019 "Also, the government -so far, I always voted -I'm never going to vote for anyone ever again. I don't believe in politics, it gets worse and worse. When I vote now, I'm going to void my vote".
Overall distrust in the Federal Government was linked to the perceived ineffectiveness of the Brazilian legal sector. Though participants recognized there is a legal code in Brazil, participants questioned the general application of the law as illustrated by one participant's comments: "Well, there is law on paper, right? But, the law in Brazil is problematic". Participants linked the general lack of government efficacy to the feasibility of legal protections for violence against women survivors questioning how women could expect the government to act in their interests when the same government demonstrates inability to ensure protections for everyday citizens across diverse issues: "It's like people say: 'In Brazil, what law?' There is no law for more serious things, but you are going have one to protect women?".
This included the perception that domestic violence charges made according to existing laws are viewed as trivial and without force. One participant described blatant disregard for enforcement of laws designed to protect women: "Today it has turned into, 'Oh, you got a Maria da Penha charge, so what?' You know?".
Participants perceived the legal sector as failing to uphold fundamental laws, even in the presence of violence against women legislation. One participant pointed out possible benefits and protections that survivors derive from existing violence against women laws can be completely undermined, and even reversed by the legal sector's inability to ensure full compliance with these laws, including the consequences for perpetrators: "The Maria da Penha law captures them [perpetrators], the justice... legal system frees them". This distrust was described by participants as extending into the justice system's ability to protect women who experience extreme forms violence such as femicide. In the absence of trust in the rule of law, desperation may lead to extralegal adaptive behaviors such as self-defense and in extreme cases retaliatory violence may emerge, as described by one participant: "There are many women who nowadays no longer trust in the justice system, and there are women who end up killing their husband in order not to die".
While retaliatory violence may be uncommon, this quote highlights the desperation women may feel. Lacking trust in the efficacy of the broader government and legal protections leaves women in violent relationships with the untenable choice between self-protection and the realistic possibility of their own injury or death at the hands of a violent partner.

Violence against women laws do not protect women and may exacerbate violence
Beyond distrust of the efficacy of the broader government and legal sector, participants also questioned the validity of violence against women laws themselves. Violence against women laws were described as having a limited positive impact on women's lives. This included legal provisions such as the type of support provided; women indicated the support provided by police may not fully address women's needs. Additionally, the law's influence was not viewed as fully permeating the private lives of survivors. An excerpt from one participant illustrated the issue of scope: "It ends up that women are intimidated by the law. Because she knows that a lot of times... the man that is inside the house is extremely aggressive, vengeful, and that if she goes to make a report then she has to come back home. And then what? Who will be there to support her? The law doesn't do it, the law is there in the police station, and when she comes home?".
Within the purview of these laws, possible interventions such as restraining and removal orders were seen as useless. Other participants contrasted the perceived trivial impact of violence against women laws with the grave realities of violent relationships, further illustrating the perceived ineffectiveness of legal provisions on immediate physical danger as described by one participant: "Sometimes the man beats up the woman and they give a paper [restraining order], and a little while later he is beating her again. What good does it do?".
Some participants suggested women often fear the consequences of attempting to make use of laws due to widespread impunity for perpetrators of violence. Numerous participants recounted stories of how women who sought to seek justice through violence against women laws were further Cad. Saúde Pública 2020; 36(10):e00114019 traumatized by ineffective interactions with health and legal sector actors -or worse -abused or murdered by their partners in retaliation. One participant went so far as to say the strong possibility of violent retaliation leads women to "fear the law more than the abusers".
Notably, participants perceived the enactment of violence against women laws as increasing the risk for women seeking legal support. Participants linked women's reporting and ineffectual laws with an increased likelihood of femicide. One participant described this phenomenon: "Oh, I have my doubts because, like that, we see a lot of cases where the woman goes, reports, and the police does nothing, and two, three months later you receive the news that the person was murdered by the husband, you know? But she went back, she tried, she reported it, but no...".
Participants pointed out punitive actions such as jail time may anger a perpetrator without addressing the source of his violent behavior, resulting in worse outcomes for effected women. One participant made this point: "It doesn't work because... the guy gets locked up, then when he gets out he goes there and kills the woman". Participants emphasized rather than reforming a perpetrator, or addressing the source of the aggression, temporary punitive solutions like incarceration could actually worsen a perpetrator's behavior and put a woman at greater femicide risk.
"And why do they kill? Because after he gets locked up he gets more aggressive. So much so that he goes and kills the woman, right? That's the truth".

Satisfaction with health sector services is mixed
In addition to a description of distrust in the legal system, participants' discontent bled into descriptions of interactions with the health sector, where women sometimes seek services following violent experiences. Women described dissatisfaction with the health system, sometimes linking it directly to the broader government health system itself. One participant described her feeling: "It's really bad, the service sucks... I've seen cases in my family and known people that die and do not get to do a blood test... The public service, the [SUS], it sucks, it gets worse and worse".
Related to these concerns were complaints about the quality of service rendered by health professionals as well as their bedside manner. Although some participants were pleased with personnel, others had deeply dissatisfying experiences. Participants described incompetence in some cases as illustrated by one participant's description of doctor visits: "They do not ask anything. You go to the doctor, you're dying of pain, you're running a fever... they look at you, but don't really look".
There is a double entendre in this participant's description of her experience. She maintains that health professionals are not performing their core functions while at the same time implies health professionals treat patients in an inhumane manner. Similarly, some participants described stigmatizing and shaming attitudes among health personnel. One participant even described being demeaned by a health professional who criticized her clothing, the number of children she had, and her use of a public assistance program: " Others had relatively positive experiences with personnel although they complained about the inefficiency of the health system, which was differentiated from the quality of care itself. This was described by one participant: "The doctors here, the nurses, are very nice, understanding, and very attentive, too; the problem itself is the delay of the service. If you are making an appointment it takes two or three months. That is the difficulty".
Some women described how the administration of the health system or the government itself was at fault rather than those health personnel who are assigned to a specific post: " Participants shared mixed views around violence disclosure to health professionals. Those who said they would not disclose violent experiences to a health provider gave a few reasons for their discomfort. One participant described the importance of rapport.
"Let's suppose, you do not know the doctor, it's the first time you go, the doctor asks you. You will not want to respond... You may be in the worst situation of your marriage; you will not want to vent with a doctor".
Another participant described having a deep sense of shame around her violent experiences, and that any health provider that she disclosed to would need to make her feel comfortable and demonstrate empathy and understanding for her experience.
" Some participants didn't expect health providers to be able to link them to appropriate resources or that there could be any positive results from disclosing to a health provider. One participant plainly stated she did no't think any good would come of disclosing, saying, "No, it was not going to help anything".

Discussion
While Brazil has made significant strides in regard to addressing violence against women through federal legal provisions, these strides may be comprised in part by recent recession and a political conflict including a current administration that has been accused of divesting from health and other programs addressing violence against women. Our data suggest within the context of Santo Andréin a country that has received international attention both for prevalence of violence and government scandals -widespread lack of trust in the government may be connected to women's perceived efficacy for violence against women response in the health and legal sectors. The historical and current political corruption, which continues to dominate the consciousness of Brazilians may contribute to a broader sense of distrust in the government. Given the perspectives of the participants in this study, it appears this broader distrust in the government interplays with specific distrust in the legal system, views of federal violence against women laws, and the health system suggesting violence against women interventions need to be intersectoral and multifaceted in nature.
Participants tended to characterize women in Brazil as having a broad distrust in the government which has implications for major pathways to reporting violence against women, namely the legal and health sectors. Women do not trust the legal sector or find its responses effective despite existing provisions; concerns were focused on the implementation and operationalization of existing laws. Survivors of violence have many reasons not to disclose experiences of violence such as stigma, shame, privacy concerns, economic dependency, fear of retaliation, and sometimes to protect the violent partner 27,33,34 . When weighing the perceived benefits of reporting against these barriers, women in violent relationships question the likelihood of a positive outcome given the demonstrated inability of health and legal sectors to effectively comply with the mechanisms enshrined in violence against women laws. A lack of gender sensitization regarding violence against women for responders such as police has resulted in women being blamed for their experience, or turned away by those who do not view it as a real crime 15,16 . The lack of public confidence in legal sector response to violence against women seems justified when one considers women disproportionately experience extreme violence in the home environment; the legal sector response is known to be inadequate regarding such crimes 12 .
The adoption of federal violence against women laws in Brazil are assumed to have an impact on curtailing violence against women 7,14,17,27 . Indeed, violence against women laws are necessary but not sufficient to address the problem. In our data, these effects were seen as relating to deterrence, potential protections from violence, and the disturbing unintended consequences of exacerbated violence. All participants were familiar with the Maria da Penha Law; however, none recalled the more recent Femicide Law. While women did recall some protective measures provided by the law, they viewed the punitive mechanisms as having a minimal impact on the protection of women who Cad. Saúde Pública 2020; 36(10):e00114019 experience violence. While the Brazilian public may positively view the principles behind violence against women laws, the laws themselves cannot have a meaningful impact on targeted beneficiaries while impunity reigns. For now, violence against women laws exist on paper but their protections are limited in practice. If these views of the government systems and actors are widespread they could result in deeper distrust, violence permissiveness, and a decrease in disclosure and reporting, rather than deterrence. Ineffectual and inconsistent punitive measures for perpetrators may lead to severe consequences such as retaliation. In fact, in many cases violence against women laws were perceived as exacerbating violence-specifically via retaliatory action. Participants linked reporting, ineffective laws and an increased likelihood of femicide. While this is counterintuitive, similar perceptions have been observed in other Latin American countries and studies in Brazil have corroborated this perception 10,14,35 .
Participants' descriptions of their discontent with the health sector were less prominent than the legal sector. Yet the lack of enthusiasm for the health sector response was clear among some participants. This pattern suggests the distrust in government not only permeates multiple levels of the government, but also transfers across sectors of government (both legal and health) when it comes to violence against women. This discontent may be exacerbated by stigma and a lack of trained medical professionals 25,27 . As seen in our data, when health professionals disrespect or mistreat women, a strategic window of opportunity for violence disclosure and subsequent violence prevention and response closes. Even with the best intentions, transference of distrust to the health system may negatively impact perceptions of health sector efficacy for violence prevention and response 36 . Distrust in the system may be compounded by negative experiences with health professionals or vice versa.
The transference of distrust undermines the ability to protect women by delegitimizing two major pathways for survivors of violence to disclose and seek support; these results also point toward intersectoral and multifaceted approaches to address violence against women in Brazil. Though interventions for violence against women are often conceived within separate disciplines, negative perceptions of a single sector (e.g. legal) may cause real consequences that reverberate through the complementary responses of other sectors (e.g. health). This speaks to the necessity of interventions that cross sectors. There has been growing attention to the development of more streamlined, integrated approaches to multi-sector violence against women response 37,38 . The reinstitution of the SPM at the local level would be a welcome and feasible first step in this regard. In addition to these integrative approaches, there should also be attention to the transference of negative perceptions across sectors, particularly in contexts such as Santo André where health and legal sectors are principally public services under the purview of the State. Regardless of how integrated these services truly are in practice, they are perceived as affiliated with a government that women generally do not trust. If left unaddressed this transference may have profound effects -beyond the current negative repercussions -on the effectiveness of violence against women prevention and response by both the health and legal sectors.
Many factors on multiple levels of the social ecology are theorized to affect women's decisionmaking regarding disclosure of violence 38,39 . In the absence of viable governmental support women experiencing violence may even resort to extralegal adaptive behaviors in the pursuit of self-protection despite the known legal consequences; alternatively they may suffer violent abuse and even death at the hands of intimate partners. Women are more likely to report violence when it is witnessed by children or when it results in bodily harm 25,27,34 . These patterns taken together may make women more likely to delay reporting violence until they are at higher risk of more severe consequences of violence against women including femicide. Response to violence against women needs to focus on reducing perceived barriers to reporting and maximizing the benefits of reporting to reduce the burden on women. The effective implementation and enforcement of existing legal protections is the starting point. In addition, service providers need to have a system in place to identify cases that are at high risk for immediate physical harm or femicide to facilitate disclosure and allocation of services to women who most need them. Additionally, experiences of violence survivors who have utilized specialized services should be examined in order to understand the degree to which distrust in the health and legal system is true based on the survivors' own lived experiences.

Limitations
Our participants were not known survivors of violence therefore they may not have had insights into the specialized services available to survivors. However, one third of our participants did disclose their own personal experiences of violence. As a result, our data reflect a mixture of insights grounded in personal experiences and the perceptions of others experiences. The sample was sufficient to see indications of saturation on the themes included in the study; however, due to the qualitative and locally specific nature of the sample the findings are not generalizable. Regardless of this limitation, the findings presented around trust and violence against women services were salient in this sample and may be transferable to other populations in Brazil, which should be explored in future empirical studies.

Conclusion
While the Brazilian State has made laudable efforts to protect women through laws such as the Maria da Penha Law (2006) and Femicide Law (2015), the perceptions of the efficacy of these laws, as well as legal and health services appears to be impacted by a general distrust in the Brazilian government 7,11 . This may be compounded by a current administration that has been accused of divesting from programs to address violence against women. Having laws on the books was the first step in improving the systems that address violence against women; however, these laws need to be enforced. Particular attention should be paid to the perception of violence against women laws as they are currently implemented and the palpable concerns of women who seek support. The results of this study suggest violence against women response in contexts like Santo André may be improved through greater enforcement of existing legal protections, increased quality of care within the health system, greater availability of trusted non-governmental services, and addressing the generalized mistrust in the government. Trust may be improved by addressing these concerns and ensuring that the laws and legal and health systems are adequately meeting the needs of women. Through improved practices such as better tools to detect violence and women at high risk of femicide, police, judges, and health professionals may be able to better identify women who need services and provide them assistance. Such services may include legal, economic, and material support (such as housing and food) as well as referrals to health services including psychosocial support. Improvement in overall trust in the system may facilitate the ability of professionals across sectors to truly address the needs of women who face threats of violence.
Violence against women interventions should consider the integrated nature of governmental trust and its effects on the perception of the legal and health sectors. The legal sector should address the perceived illegitimacy of violence against women laws, perception of inadequate implementation of violence against women laws, and specific concerns among survivors including the fear of retaliation. The health sector should consider the transference of distrust from the legal sector and personnel concerns regarding disclosure readiness and violence against women sensitization. Both legal and health professionals operating within public systems of the Brazilian government are State actors. Their individual actions therefore are reflections of the State and vice versa. The potential to improve public trust lies largely in the hands of these well-intentioned public service officials as well as a transparent government free from corruption 40 .
The adoption of violence against women laws in Brazil have sought to better protect Brazilian women who continue to be disproportionately affected by violence against women and femicide. This analysis serves as further evidence of the need for additional policies and programming, and a robust multisectoral and multifaceted response in the prevention and response to violence against women.
Cad. Saúde Pública 2020; 36(10):e00114019 Contributors D. P. Evans conceptualized the study and its design, participated in data analysis, and drafting and critical revisions to the manuscript. C. D. X. Hall and N. DeSousa participated in data analysis, and drafting and critical revisions to the manuscript. J. D. Wilkins, participated in data collection and critical revisions to the manuscript. E. D. O. Chiang participated in data analysis and critical revisions to the manuscript. M. A. F. Vertamatti conceptualized the study and its design, participated in data collection and critical revisions to the manuscript.