The Brazilian women’s movement and the state under the PT national governments

Certain segments of the Brazilian Women’s Movement (BWM) developed important strategic partnerships with the state under the PT federal government. This article demonstrates how these partnerships had both inclusionary and exclusionary effects on gender-centric policy outcomes. Using a mix-method approach which includes extensive in-depth interviews with BMW activists and archival data from governmental sources, the Brazilian experience illustrates that “entering the state” headed by a women-ally president facilitates funnelling some of the women’s movement’s demands into the state. However, hierarchies within the movement itself, the congressional strength of the head of executive and the activism of some of its socially-conservative legislative allies can severely temper that effect.


Introduction
Gendered public policy within present-day societies is relatively recent and has emerged to address the needs and interests of large segments, or in some countries the majority, of the population (see, for instance, Hernes, 1987;McBride 2010;Sawer, 1990). As such, "velvet triangles" (Woodward, 2004) have sprang up in several polities around the world. The latter encompass alliances among several actors, especially women's movement actors (WMAs), elected politicians, and feminist public bureaucrats, which were established with the goal of fostering gender-centric public policy. These actors engage in cooperation, coalition-building and joint activities for achieving common goals, especially policy change. Although different aspects of these "velvet triangles" or strategic partnerships have been analysed, the issue of their effect on the inclusion, prioritization, or even exclusion of the WMAs' claims that are transformed into public policies remains understudied. In other words, to what extent do these "strategic partnerships" (Halsaa, 1991(Halsaa, , 1998 open spaces for all the priorities of the women's movement to make their way into state policy? Or do these partnerships inadvertently act as a gatekeeper, and have restrictive effects on some claims as well? This article focuses on the case of Brazil, where women's movement actors developed a process of close cooperation with the central government. When the PT (Workers' Party) attained the federal executive in 2002, several WMAs severely tempered their critical stance in relation to state, and "entered" its apparatus. Although the analysis centres on the years of the two Lula presidencies (2003-2006, and 2007-2010) in particular, it evinces the role of factors that could be deployed to understand the demise of Dilma Rousseff's second presidency, and the subsequent rise to the presidency of political forces that are antisocial movements, have a misogynistic and patriarchal rhetoric and a very distinct project for Brazilian women. Using extensive qualitative and archival data, the goal is to demonstrate that feminist strategic partnerships can be a double-edged sword. In the case of Brazil, on the one hand, these "triangles of empowerment" (Vargas & Wieringa, 1998) provided the women's movement with unprecedented access to the state, particularly when it comes to resources, spaces of interaction, communication and collaboration, and to the ability to shape the formulation, approval, and implementation of gender-centric public policies. On the other hand, they acted as filter, in the sense of selecting which claims made it to the list of priorities and which did not. This filter was due to the priorities of the WMA that managed to "enter" the state, and to the preferences of other political actors who (directly or indirectly) also belonged to the strategic partnerships.
One first important lesson that arises from this case study is that in societies with high levels of (racial and/or class-based) fragmentation, the women's movement's sheer diversity may engender an internal dynamic from which gender-based demands are ranked in terms of importance and urgency. The difficulty is that the WMAs who enter the state bring with them their own worldview and associated list of gender priorities, which can ultimately result in the neglect of the demands of important contingents of women. Second, partnering with a women-friendly head of executive is clearly insufficient for advancing gender-centric public policies in some situations. In contexts of legislative minority, or when the governing coalition must rely on socially conservative party allies, some WMA's historical claims might not materialize into public policies both due to the legislative activism of these conservative allies and the temperance of social movement actors.

Velvet triangles
Social movements can benefit immensely from the arrival of their political allies to positions of political power. Mazur (2001), for instance, demonstrated that cooperation among women's movements activists, "femocrats" in women's policy agencies, and elected representatives was central for gender-based policy outcomes in several Western democracies, especially the implementation of innovative social policies and democratic practices (Giugni, 1998).
What is most interesting about these accomplishments is that they stem from what the Western-based literature refers to as "velvet triangles" (Woodward, 2004), "triangles of empowerment" (Vargas & Wieringa, 1998), "strategic partnerships" (Halsaa, 1991(Halsaa, , 1998, or "feminist strategic partnerships" (Mazur, 2001). Although these terms differ when it comes to the specific actors involved, they all refer to patterns of engagement between civil society actors and state actors aimed at the production of gender-based public policy outcomes. These concepts denote the development of alliances among differentlypositioned groups and individuals, which include elected representatives, women's policy agencies in government, women's civil society organizations and gender experts such as academics (Woodward, 2004), who find themselves engaged in complementary projects (Walby, 2011).
How do these considerations apply to the case of Brazil? This article analyses what impact the move from a stance of protest to a dynamic of systematic engagement with the state -especially the federal government -has had on state outputs, particularly as it relates to the implementation of the key claims in the Brazilian women's movement's agenda. How do the strategic partnerships help us understand which items of the women's movement agenda were transformed into public policy, and which ones did not make it? Most importantly why was the latter excluded? Finally, does this potential diversity of outcomes help illuminate some of the factors that fostered the PT removal from power, and the end of the Pink Tide in Brazil? As it will be discussed below, the elements influencing which WMA claims make it into policy, and which do not, include endogenous ones concerning the composition of the BWM and its internal dynamics as well as exogenous ones involving the composition of government and the political playing field.

Methodological note
This article utilizes as one of its primary sources in-depth semi-structured interviews with thirty-nine organizations; most of which are non-governmental organizations from the BWM. Gender-centric state actors (such as representatives of the subsecretariats of public defenders, of municipal-based women's secretariats and of the national-level Secretariat of Public Policies for Women, or SPM) and Brazil's National Association of Non-Governmental Organizations (known as ABONG) were also interviewed. Those interviews took place between June 11, 2016 and October 28, 2016. They were carried out in eleven cities of the Brazilian federation and involved both urban-and rural-based women's organizations, located in small cities as well as large metropolitan areas. In order to increase the representativeness of the sample, the study included women's organizations with different missions, organizational profiles and proclivities, such as volunteer-based organizations; professionalized organizations with a cadre of paid liberal professionals and other staff; organizations primarily devoted to research and advocacy; organizations which provide direct services to women; organizations which have resorted to state funds to sustain some of their projects, and organizations with a non-state pattern of funding. When it comes to the characteristics of the individuals from the organizations selected, they were seasoned leaders in the women's movement, as well as members of the rank-and-file, middle-class and upper-middle-class women, and women from marginalized socioeconomic backgrounds. Both white and visible-minority women (especially Afro-Brazilian and Indigenous women) were among the interviewees 1 .
To create a data triangulation with the interview materials, all National Plans of Public Policies for Women, as well as the annals of the women's conferences were analysed. Those documents allow for a clearer understanding of the women's movement's objectives in terms of gender-centric public policies, as well as the degree to which the SPM incorporated those desiderata in its national plans. Finally, as the interviews reflect a subjective perception of the Lula governments' pro-women public policies, laws, bills, and presidential decrees were incorporated into the analysis. We should mention that the interviews took place a few months after the impeachment of president Rousseff and the beginning of the Temer government that abruptly ended its funding to many civil society organisations.

The women's movement, the PT and Lula
Historically, there has been a close relationship between progressive segments of the heterogeneous women's movement in Brazil (but certainly not its totality) and the Workers' Party. Several factors account for that proximity. One important element is the constitutive nature of that party itself. The PT was formally founded in 1980 amidst the final years of a military dictatorship, as a result of the amalgamation of a cluster of progressive social forces, including left-leaning intellectuals, segments of organized labour and of the Catholic Church, and a multitude of social movements (Keck, 1992;Meneguello, 1989). Among the latter one would include large sections of the women's movement (Álvarez, 1990). There is an important affinity between some WMAs and the PT and in many cases an overlapping as the result of a double social movement-party activism. Another important factor that accounts for the proximity with WMAs relates to PT's ideological profile, and its reflection on the internal functioning of the party. Its former prosocialism stance meant a staunch commitment to embracing the social struggle of the marginalized sectors, the empowerment of social movement actors and the creation of the social basis of a new type of society (Silva, 2009). The PT's functioning -based on internal democracy and the open dialogue amongst several leftist groupings -allowed for the creation of an important women cell, which increasingly made genderbased claims known and gradually accepted within the party (Godinho, 1998;Macaulay, 2006). Interestingly, in 1985, just a few years after its creation, the PT was the first party to elect a woman mayor of a state capital.
As a union leader and a charismatic figure with close ties to several social movement actors, Lula da Silva was open to listening and understanding the agenda of the women's activists within the party. The latter made him aware of the women's movement's struggles, sources of oppression, and demands. Over time, some WMAs saw in Lula a high-calibre political ally, and militated to get him elected, supporting all his five presidential candidacies. This close relationship permeated into Lula's campaign promises. In his 2002 presidential campaign, for instance, Lula emphasized the need to put in motion "a political and cultural effort to affirm in the country the principle of equality. It is insufficient to combat only the economic causes of the multiple forms of inequality. It is necessary to undertake positive actions to put to an end the different forms of discrimination against women, Afro-Brazilians, native Brazilians, disabled individuals, and sexual minorities, just to cite a few" (da Silva, 2002). In other words, Lula and segments of the women's movement gradually grew closer and closer. Even though some authors point to the fact that the PT governments enlarged Brazil's social contract and without a confrontation with landowners, bankers and industrialists , Lula did fulfil several of his campaign promises to organized women. In 2003, he created the agency that WMAs had been calling for: The Secretariat of Public Policies for Women.
Unlike its predecessors, the SPM had considerable autonomy of action, its own budget, and it was housed under the Presidency's Office. Its creation changed the relationship between the women's movement and the federal government, as it approximated both, leading to a substantial change in the pattern of engagement between these two actors (Bohn, 2010). WMAs moved away from their prior perspective of protest against the state. Interestingly, this downward moment in the BWM's cycle of contention will coincide with a period of substantial increase in the activism of socially-conservative political forces in Brazil, both inside the legislature and in the streets. WMAs formed an important "velvet triangle" with public sector actors, especially SPM's "femocrats". Included in that alliance as well were president Lula and his staff and allied elected parliamentarians. In order to create its own cadre, SPM's recruited heavily among university professors and gender experts from civil society, as well as some of the most seasoned leaders of the most established organizations within the women's movement. Thus, organizations which already had a name and a history gained access to the velvet triangle much more easily than their peers which operate at the political and socioeconomic margins.
The Secretariat sustained important lines of communication with the women's movement, and over time, became both an important space of state-civil society interaction, and one of the key actors of the PT government's progender policy. Interestingly, this alliance quickly evolved into a policy community. The women's movement actors held city-level and state-level conferences to decide on a common platform of action, which was discussed with SPM in national conferences 2 . SPM, in turn, issued national plans of public policies for women mostly based on the discussions from the conferences that took place in 2004, 2007, 2011, and 2016. Before delving into the public policy proposals that were approved, it is necessary to understand the BWM's key demands.

The heterogeneous Brazilian women's movement and their main claims
In a territorially large country with a past of slavery and an experience of exploitative colonialization, with substantial contingents of Afro, indigenous and other traditional communities (Bohn & Grossi, 2018), as well as of descendants of white European immigrants, the women's movement could not help but reflect the constitutive diversity of the Brazilian nation. The WMAs are diverse when it comes to their composition, activities, ideological proclivities, repertoires of protest, specific demands and partisan affiliations (e.g. as not every group has had a close relationship with the PT). The women's movement in Brazil is generally conceived as a plural entity made up of various groups and political subjects (Barbosa-Araujo & Adrião, 2015). There are middle-class and highly educated (preponderantly white) feminists working in academia and in highly-institutionalized NGOs; women in political parties, unions and federations; and a plethora of (including racialized) women's movement actors spread throughout the country (Vargas, 1992). In this sense, the women's movement reflects Brazilian society's hierarchical structures as shaped by race, class, sexuality and geographical regions. Despite this inherent diversity, there are points of convergence in the movement's agenda. On a macro-level, its goals are the democratic participation of citizens in civic life and public debates; promotion and defence of human rights and improving the quality of life of marginalized populations through social organization and development at a local level. In terms of a more gender-based agenda, nearly all the WMAs in-terviewed describe that the movement struggles to create a united front to challenge patriarchal attitudes, institutions and practices while representing a wide variety of interests of women with diverse identities and backgrounds.
Since the 1970s, there has been an effort to increase the network of resources that enable women's integration into the workforce, such as the expansion of the number of day cares; to push for the development of a policy to address the issue of violence against women; to improve healthcare, particularly for children and women in all phases of their reproductive life (Blay, 2017;Pinto, 2003). The issue of the decriminalization of abortion has always occupied centre-place for most women's movement actors. Abortion in Brazil is a crime in all circumstances, subject to a jail sentence. In cases of rape or incest, if the growing foetus lacks a brain, or if the life of the mother is in danger due to the pregnancy, the practice of abortion -still a crime -is devoid of penalties (mothers and doctors do not go to prison in those cases). The feminist segment of the Brazilian women's movement -which is its largest contingent but not its totality -has always insisted that it is past time to change the existing legislation, especially given the harsh reality of maternal death and a whole array of health-related ailments stemming from the use of unregulated, clandestine abortion clinics, which disproportionately affect impoverished women. In fact, although the official statistics estimate that approximately 5 percent of maternal deaths derive from abortive practices (Ministry of Health, 2013, p. 47), it is widely known that, since it is an underground practice, the numbers are much bigger, as the phenomenon is under-reported (Martins et al., 2017).
In addition to these movement-level demands, some WMAs have groupspecific demands. Below are some illustrations of those claims. Afro-Brazilian women, for instance, have, for a very long time, pressed for the expansion of services to deal with sickle cell anaemia, especially the retraining of healthcare providers (Criola, 2007;Lopes & Werneck, 2010). This genetic condition disproportionately affects Afro-descendent individuals, and its effects are particularly deleterious for women in their child-bearing years. Lamentably, the Ministry of Health in Brazil, in its official reports, recognizes that women with such condition are often not given access to specialized services "due to lack of information among healthcare professionals about the disease" (Ministry of Health, 2015, p. 10).
Indigenous women in Brazil have a severe problem with accessing services, particularly but not exclusively, healthcare, both for basic needs (such as prenatal and postnatal care) and preventative measures, such as services to diagnose and treat uterine cancer and HIV (Pereira et al., 2015). When healthcare does arrive in their communities (some of which are very secluded), the medical teams often lack female healthcare providers, which deters indigenous women from utilizing the service. Ultimately, those situations exemplify the inadequacy of the existing healthcare policy for this segment of Brazilian women, due to the precarious offering of services, and the lack of culturesensitive training of medical personnel (Ferreira, 2013).
Rural-based Brazil women, which encompass several traditional communities and individuals from different social classes, also have problems with deficient access to public services. Brazil's innovative policy to deal more effectively with the problem of violence against women, the so-called police stations for women -where female officers with gender-sensitive training handle cases -concentrate in highly-urbanized areas. The distances in this almost continental country mean that rural women who are victims, or in danger, of gender-based violence are left to their own device. In addition, a substantial number of rural women lack basic citizenship documents (such as birth certificates, social security cards, etc.). Finally, as a reflection of essentially patriarchal practices, most rural lands are registered under the name of husbands only, which in practice means that those women are unable to access credit to move their agricultural business forward when their husbands are absent, such as when they temporarily move to urban centres in search of job opportunities (Deere & Leon, 2003).
Ethnicity and place of residence are not the only factors that create obstacles for contingents of women to access public services. Sexuality is also a major impediment. Sexual-minority women in Brazil face severe difficulties to navigate through the network of social services in the areas of healthcare, education and public security, especially due to the discrimination that they face (Toledo, 2008). Irigaray and Freitas (2011) chronicle different forms of discrimination at the workplace, which are accentuated by other elements (e.g. race, ethnicity, and class belonging). LGBTQ organizations call for the need to work towards a change of mentality and values, and a culture of acceptance of diversity. Which of those movement-level claims and group-specific demands did the strategic partnerships -led by the PT under Lula -transform into public policies? Prior to delving into the analysis of the actual gender-based policies of Lula I and Lula II, a few words about his support coalition and Brazil's rather unique party and electoral systems are needed.

President Lula's support coalition
Since redemocratization, Brazil's party system has been extremely fragmented, with many parties holding seats in the Chamber of Deputies (the lower house, with 513 representatives) and the Senate (81 representatives). In the Congress formed in the aftermath of Lula's first presidential victory in 2002, nineteen parties had seats in the Chamber (Câmara dos Deputados, 2017) and ten in the Senate (Senado Federal, 2018). Lula's PT controlled 18 percent of seats in the Chamber and 16 percent in the Senate. Needless to say, Lula -as did all presidents before him since 1986 -had to rely on a coalition of parties (seven, including the PT) to govern. This institutional configuration has been aptly labelled as "coalition-based presidential system" (Abranches, 1988). The latter is challenging for presidents, as, on the one hand, they need to engage in a lot of give-and-take to maintain their ideologically-and programmatically-heterogeneous base of support. They might have to cater, and eventually make concessions, to legislative allies who are important to strengthen the base but hold divergent public policy priorities in some areas. On the other hand, Brazil's directly-elected presidents also need to deal with an illusion of power; i.e. the popular perception of a powerful president. Lula I, for instance, was elected in the second round of presidential election with 61.3 percent of the popular vote (TSE, 2018a). From the perspective of some segments of the Brazilian population (including some social movement actors), governing is essentially a matter of political will from the part of the president; and there is not much awareness of the impact of the party system fragmentation on the head of government's ability to pass bills in a bicameral Congress. This rather distorted perspective was magnified over the course of the 2006 presidential election, which Lula won with 60.8 percent (TSE, 2018b), also in the second round. Lula II remained entangled in a highly-heterogeneous coalition, with distinct ideological proclivities and public policy proposals. 3 In addition to his traditional social movements and labour ally base, who or which political parties were part of Lula I's and Lula II's coalition? First, a group of businessmen, congregated in the Liberal Party and other parties. The group was supportive of Lula I's vicepresident (José Alencar, himself a business leader) and welcomed Lula's proposal of improving Brazil's social contract, decreasing extreme poverty and eliminating hunger ). Another important, more sizeable group, which has experienced substantial growth in terms of congressional seats over the decades, was the caucus of Brazilian evangelicals, who are spread over distinct political parties, especially PSC (Social Christian Party), PR (Republican Party), and PRB (Brazilian Republican Party). That group also supported strongly Lula's social policies. However, it has been notorious for its opposition to the decriminalization of abortion and illicit drugs, the use of stem cells in research, and the approval of homosexual civil unions. How did these coalition partners work against the strategic partnership between the Lula presidencies and the women's movement actors? As we will see below, conservative actors within the PT-led coalition were opposed to many of the BWM's principal claims, and actively opposed laws in congress, used their influence on the executive to bargain for their support on other issues, actively mobilized their followers and wielded influence in the national media on public opinion. What claims from the movement's agenda were transformed into public policies?

Poverty eradication measures
The Bolsa Família initiative is a federal program which aims at transferring revenue directly to poor families to reduce hunger and the intergenerational transmission of poverty. Ninety-three percent of BF recipients are women who are also responsible for submitting to the conditions of the program. It is important to remember that the number of women head of households nearly doubled since 1995 from 23 percent to 42 percent (IPEA, 2018). These women in the lowest socio-economic category live in a context of extreme social marginalization (Pinto et al., 2011). A secondary goal of the program is to contribute to the self-esteem and empowerment of women in terms of allowing them access to more material resources and thus increasing their capacity to make choices for themselves and their families (Moreira et al., 2012). Feminist critics consider that although it produces some positive results such as access to a better diet, goods and credit, it uses women to administrate poverty rather than empowering them to overcome it (Cisne, 2013). Nevertheless, as the interviewee below remarks, this financial support is an effort to redistribute social value (Htun & Weldon, 2010) from men to women, and contributes to the autonomy of millions of women: …it was remarkable that the PT governments gave so much priority to gender policies, especially transversal policies such as the Bolsa Família and Minha Casa, Minha Vida… which essentially transformed women into protagonists […] There may be women who do not see these programs as gender policies, but without a doubt they create a lot of power for women inside the family, the power to negotiate with their partners, the power to impose some decisions (Women's movement actor; interview in São Paulo, on June 22, 2016).

Violence against women
There has been substantial progress in addressing this issue since the approval of the Maria da Penha Law in 2006, which the representatives of the women's organizations that are part of this study, such as the one below, consider a watershed moment: …historically [the law] represents a tremendous advancement […] If you looked at the years before [the approval of the law], violence against women was not even considered violence (Women's movement actor; interview in Cabo de Santo Agostinho, on October 24, 2016). This piece of legislation has fostered the expansion of a comprehensive array of social services dedicated to helping women (and their children) in situation of violence, such as police stations, shelters, and other specialized services (Bohn, forthcoming). In addition, the Maria da Penha Law made it easier for battered women to legally detach themselves from their aggressor, as they can obtain a divorce in the same court that deals with child custody (issues that were previously handled separately).

Sexual tourism and sex trafficking
For many years, successive Brazilian governments advertised Brazilian tourism to foreigners with sexist images of women, which had deleterious effects on sex trafficking and tourism. One of the actions of the Lula's (and Rousseff's) administrations was to create state-sponsored groups around the prevention of, and struggle against, human trafficking. Special posts at airports offer support to victims. In addition, media campaigns and promotional material were elaborated to call attention to the problem and involve the different forces of order at different levels of governments across the country.

Labour Rights for Domestic Workers
The most common occupation for low-income (often visible minority) women is that of domestic worker. The relationship between employers and domestic workers is heavily conditioned by Brazil's institutionalized slavery past and goes beyond economics to embrace social relations and Brazilian culture. The Lula (and Rousseff) administrations introduced new measures to provide basic protection for some seven million domestic workers previously excluded from Brazil's labour laws. A constitutional amendment limits domestic workers to a forty-four-hour week. It defined other rights as well -basic entitlements such as an eight-hour working day, the right to the minimum wage, a lunch break, social security, severance pay, and protections to pregnant workers.

Property titles and housing
Since the massive migrations from countryside to cities, which intensified in the 1960s, Brazil has had an enormous housing deficit. Traditionally lowincome state-subsidized housing has been allotted to male heads of households contributing to grave wealth inequalities within Brazilian homes. In 2006, under Lula I, the Ministry of Cities adopted the Resolution 004/2006, which gives priority to low income families with a women head of household within their public housing programs (SPM, 2007). In 2007, housing movement activists proposed that joint property titles be given when distributing public housing to low-income families so that women are protected against losing their home in the case of separation and divorce (Verschuur, 2011); which it is of great importance to rural women. In 2008, this measure (of joint titling) was integrated into the ambitious federal housing program Minha casa, Minha vida, whose goal is to build housing units for low-income families.

Gender-sensitization of the media
The representation of women in the media reflect sexist attitudes and the objectification of women found in social values and beliefs. One of the longstanding goals of the women's movement is to change misogynistic Brazilian culture. The SPM supported civil society organizations in their research and knowledge mobilization around issues such as the invisibility of women in media reports on the world cup, and the representation of women in TV commercials. The issue was taken up at the national conferences on communications and women's public policies (e.g. it is priority 8.1 in the second national plan of public policies for women); which led to seminars aimed at the social control of the media, the promotion of non-discriminatory practices in the media industry, and the democratization of the means of communications.

Women in politics
The issue of the inclusion of women in politics was taken up in the second national plan with the goal of promoting and strengthening egalitarian, plural and multi-racial participation in spaces of power and decision through a law of electoral quotas. In 2009, a mini-electoral reform was passed (Law 12.034/2009), stipulating that all parties must allocate 5 percent of party funding to the political training of women, that 10 percent of the party program should involve women, and that all parties should have a minimum of thirty percent and maximum of 70 percent of candidates of each sex (Rocha, 2012).

Lessons
What lessons can we draw from the approval of these measures? The strategic partnership did bear fruit. First, the institutional strength of the presidency and the relative independence of the women's policy machinery (the SPM) allowed for the rather rapid approval of certain gender-centric public policies, which has been perceived by women's movement actors -as noted -as a substantial accomplishment. Second, women's movements' claims that are central to the movement but not contentious to other strategically-positioned actors -such as the issue of violence against women, for instance -can more easily be transformed into a public policy proposal, approved and implemented. The same applies to non-contentious claims that are supported by other social movement actors. Examples of those in the case just analysed were changes in the titling of properties (which were supported by unions and the MST, the Movement of Landless Workers); the expansion of domestic workers' labour rights (which organized labour backed); and the placing of disadvantaged women (who are heads of household) in the priority pool for public housing projects (the measure had the backing of the housing movement). Interestingly, many of those claims are material or economic in nature, and thus aligned with the classbased, larger goals of the PT governments.

Reproductive rights
Numerous religious-based politicians and parties -some of which, as mentioned before, were part of president Lula's (and Rousseff's) coalition -have held staunch pro-life positions. As a result, access to safe, though illegal, services is limited to middle and high-income women. In addition, many women who are eligible for induced abortion under the law cannot obtain services. Approximately one million women each year seek abortions to end unwanted pregnancies (Ministry of Health, 2011, p. 8). Although the issue of abortion has been addressed in the National Conferences on Women's Policy 4 and president Lula demonstrated openness to it, it was not implemented by the PT while in the federal government. Interestingly, a legislative proposal (PL 1135/1991) aimed at decriminalizing the practice of abortion (by removing it from the Penal Code) stalled in Congress for over a decade since its introduction in 1991. With the Workers' Party's arrival in power, important segments of the BWM pressed for the decriminalization and the congressional approval of a new regulatory framework. However, other actors, such as the powerful CNBB (National Conference of Roman Catholic Bishops) and the growing caucus of Christian-evangelical legislators -some of whom (it bears repeating) were part of president Lula's governing coalition -showed their strength, particularly when the proposed bill reached the Chamber of Deputies' Committee on Social Security and the Family (CSSF) -and later the Justice, Constitution and Citizenship Committee (CCJC). The arrival of the bill at the CSSF (in October 2005) coincided with a moment of great debility for president Lula I, as since June of that year, his administration had been embroiled in a large-scaled corruption scandal, known as "monthly stipend" (the government was being accused of exchanging resources siphoned off from state-owned enterprises for votes in Congress, as way to eschew the inherent debilities of the "coalitionbased presidential system"). Pro-life allies of president Lula stood by his side in that moment of grave political instability. Their intent to stop the bill aimed, in their words, at "killing innocent children from conception up until birth" (Portal da Família, 2005) was not strongly questioned by president Lula, who badly needed their support at that point in time (Machado, 2016). BWM actors did take note of that: We [from the women's movement] have some platforms of action, some needs that were not fulfilled by the [PT] government, because the latter lacked strength […] A key example is the abortion issue (Women's movement actor; interview in São Paulo, on June 20, 2016).
There were demands that did not make it, such as the abortion issue for instance, because it is a hot topic; and because it was a coalition government. Therefore, within the limits of the coalition under which they [SPM people] were operating, they did the best that they could […] I think (Women's movement actor; interview in Porto Alegre, on June 16, 2016).
Not only an enlargement of the set of reproductive rights did not materialize in Brazil under the PT, but also the risk of retrocessions increased in the course of the Pink Tide in Brazil. In 2015, during president Rousseff's second administration, some socially-conservative political leaders in Congress proposed to expedite the passage of a bill proposal (PL 5069/2013), which, in addition to criminalizing the prescription of abortive methods (such as the so-called nextday pill), would impose several additional hurdles on women seeking an abortion in cases where the pregnancy was the result of rape. The proposal, in essence, amounted to the hardening of the already very restrictive legal framework. In response to this legislative activism from the part of these conservative groups, organized women -particularly, the young feminists -initiated a cycle of massive protests in several cities. This "feminist spring" evinced the profound limitations of the PT-led velvet triangle in Brazil, and signalled the beginning of a moment of contention, after a period of substantive collaboration between the BWM and state actors.
Although no further legislative action has been taken concerning bill proposal 5069 since October 2015, it is ready to be voted on by lower house, as its constitutionality has been attested by the CCJC (which in 2015 was controlled by socially conservative forces). Furthermore, it is important to mention that some of Lula's pro-life allies slowly shifted their allegiance away from the Workers' Party in the course of president Rousseff's administration, and clustered around right-wing political leaders thereafter. These socially-conservative allies also contributed decisively to the defeat of other important demands put forth by the BWM.

Rights of sexual minorities
The bills to legalize homosexual civil unions in Brazil either stalled or were defeated in Congress. After many years being examined in the Chamber of Deputies since it was introduced in 1995 (by a PT congresswoman), bill proposal PL 1151/1995, which aimed at giving legal recognition to homosexual civil unions, was removed from the lower house's agenda on May 31, 2001 as a result of an agreement among party leaders. With the intent of permanently defeating it, in 2007, a Christian evangelical politician (congressman Russomano) requested that it was placed back on the agenda for voting. Similarly, PL 6297/2005 presented on November 30, 2005 (also by a PT congressman)whose goal was to extend pension rights to partners in homosexual relationships -was defeated on November 20, 2013 in the Committee for Human Rights and Minorities. Interestingly, that defeat was primarily articulated by another Christian evangelical congressman, Pastor Eurico. It was in the Judiciary where some inroads were made regarding the issue.
Regarding the question of civil unions, it is important to mention that the Brazilian laws allow (heterosexual) couples in the so-called stable unions to have similar rights as (heterosexual) couples who are legally married. In 2011, the Federal Supreme Tribunal (STF) recognized that homosexual couples that can prove that they are part of stable unions have the same rights as those enjoyed by heterosexual couples in stable unions. In 2013, another branch of the Judiciary, the CNJ (National Justice Council), determined that homosexual couples in stable unions can convert those relationships into a civil union, and, therefore, access rights that are identical to those that heterosexual couples who are legally married have. Although these two measures represented important gains for sexual minorities in Brazil, in practice, homosexual civil unions still do not legally exist in the country. Moreover, it is essential to emphasize that it has been the judicial power in Brazil which, in response to the pressure from LGBTQ activists, is increasing the body of rights for this segment of Brazilian population.
Another important theme for sexual minorities in Brazil is the struggle against discrimination. In 2001, a PT congresswoman introduced a bill proposal to criminalize any discriminatory actions against homosexuals, including in the workplace (PL 5003/2001) -which, as noted before, has traditionally been one of the key demands of sexual minority women. LGBTQ advocates hailed when it was approved (albeit with changes) as PL5003-B/2001 in the lower house in 2006 and sent to the Senate for approval -where it received the nickname of "Law against homophobia". Several Christian Evangelical pastors, Roman Catholic leaders, some Brazilian army officials (the military do not recruit sexual minorities to its ranks in Brazil), and others manifested publicly against the bill. Evangelical leaders were especially active. They organized public rallies against its approval; called upon evangelical voters to punish in the ballot box any senator who would vote in favour of the bill; and threatened a judicial battle in case of its approval, as in their view, the proposed piece of legislation would jeopardize their freedom of expression and of belief. The bill was ultimately archived in the Senate on December 23, 2014. 5 Finally, still related to the objective of decreasing discrimination against sexual minorities, the women's movement had expressed the need to foster a non-sexist education, especially through curricular changes, and gendersensitive training of teaching staff. In response, the SPM supported the governmental program "Gender education and sexual diversity." However, congress took this initiative off the agenda concerning goals around gender inequalities, in its discussions of the National Education Plan. This happened following the objection of Christian Evangelical and Roman Catholic congressmen to educational videos (commissioned by the Ministry of Education) to inform students about transsexuality and bisexuality. In 2011, the sitting presi-dent, PT's Rousseff, decided that schools would not receive those videos. Regarding this topic, several groups echoed the sentiment of this leader of a women's organization, who had the post-Dilma government as a frame of reference: …our organization works to create a non-sexist education […] This platform is a big challenge nowadays […] Even under the Dilma government, in order to approve the National Education Plan, the word "gender" had to be removed due to the strength of fundamentalist religious groups […] Thus, in the Plan we had to go back using the terminology of masculine and feminine, there is no such a thing as gender... (Women's movement actor; interview in Rio de Janeiro, on June 14, 2016).

Lessons
Although the strategic partnerships can be credited for spearheading important policy responses, they can thwart other policy changes. Given the role played by the institutional political framework, the head of government's allies are also (albeit sometimes unwanted) partners in the velvet triangles. This, in practical terms, entails that -as the issues of abortion or the expansion of the rights of sexual minorities illustrated -a president's conservative allies can effectively block bill proposals whose content ranks very high on the women's movement's list of priorities. Furthermore, the blockage of those claims illustrates the influence of religion into state policy making and the need to include this factor when studying strategic partnerships.

Conclusions
After 2003, the women's movement in Brazil sustained a strategic partnership with state actors, particularly its women-friendly and politically powerful ally, president Lula. This empirical case demonstrates that the concepts of "velvet triangle", "strategic partnership" and "triangle of empowerment", conceived from the analysis of experiences from advanced capitalist democracies, are useful to countries from the Global South. In fact, the strategic partnership proved essential to the approval of several policies that matched key claims historically put forward by the BWM. However, it is also clear that claims that were not on the top of the list for women's movement actors who actually "entered the state" -such as some of the demands of the afro-Brazilian and indigenous women -were not funnelled in by the strategic partnership with state actors, which exemplifies the fact that velvet triangles cut both ways. On the one hand, they make it easier for certain demands to be translated into public policies, i.e. those which are not contentious and are a top priority for the more influential WMAs. On the other, they also syphon out issues to which political allies are not sympathetic, or to which politically-connected WMAs are inat-tentive. In addition, this case illustrates that the nature of the political system, especially the party system, matters a great deal for the analysis of any strategic partnership. Women's movement actors allied with governments which have minorities in Congress, may face considerable difficulty to approve and/or implement certain demands, even if they have agenda-setting capacity. Thus, having a women-friendly, head-of-government ally is clearly insufficient in a fractured party system, particularly "coalition-based presidential systems", as is the case of Brazil.
Under the Workers' Party's presidencies, social movement actors -especially WMAs -entered the state and initiated a process of strong collaboration with state actors, which paralleled a downward moment in their cycle of contention. Paradoxically, some of the PT's powerful socially-conservative allies reacted to some of the BWM's historical claims and, in fact, utilized these very demands to flex their legislative muscles and intensify the mobilization of their supporters. This is another double-edged sword. As supporters of the PT at the federal level, these allies helped the Lula's and Rousseff's governments pass legislation in congress. On the other hand, they directly impacted the concrete functioning of the triangle of empowerment in Brazil and ended up acting as a powerful veto coalition when it comes to some of BWM's policy goals.
Due to the preferences of these socially-conservative allies and particularly their extraordinary mobilizational capacity, at moments, there were clashes around public policy preferences, which president Lula (and later Rousseff) had to promptly solve. Sadly for the women's movement, as the interviewee below eloquently articulates, both did so in ways that prioritized the maintenance of their mosaic-like network of congressional support -and not their gender-centric commitments: …Independently of the fact that the Dilma government committed substantial errors in relation to organized women, the PT alliance with evangelicals and Catholics is horrible, is terrible. It [the alliance] started with the Lula's administration, not with Dilma's, this cannot be denied… (Women's movement actor; interview in São Paulo, on June 13, 2016).
These fractures and contradictions within the coalition-based presidentialism, which president Lula was able to skilfully mend or solve, proved insurmountable in the course of president Rousseff's second administration, contributing to the tumultuous and controversial process of impeachment, and the abrupt end not only of PT's reign in the federal government, but also of public policies aimed at creating a more socially just and gender-equitable Brazil. * * * Simone Bohn is associate professor of Political Science at York University. She co-edited Mothers in Public and Political Life and is currently working on a SSHRC-funded project on Brazil's women's policy agency.