Public Policy Frames and Gender mainstreaming: an analysis of childcare policy in Brazil during the left turn (2003-2016)

Brazil was one of the countries that integrated Latin America’s left turn, a period in which social policies have become central. During the Labor Party’s Government (Partido dos Trabalhadores) (2013-2016), were developed institutional conditions to mainstream gender in public policies, which embraced the issue of the sexual division of labor. However, did it mean an effective reorientation of the childcare policy towards gender equality perspectives? This article aims to reflect upon this question, drawing on the gender mainstreaming concept. It is understood as a process of incorporation of feminist perspectives into the public policy framing, regarding the (re)definition of both the public problem and the course of state action. To do so, we carried out a qualitative study of gender mainstreaming on childcare policy (daycare centers and leaves), focusing on official discourses, mainly through documentary analysis. Based on the results analyzed, we identified the coexistence of two frames: “education and childcare rights” and “promotion of women's economic autonomy”. Since the first one has prevailed, we conclude that gender mainstreaming was marginal in the childcare policy, during the analyzed period.


Introduction
At the turn of the 21st century about two-thirds of the Latin American population were governed by left-wing coalitions (Filgueira & Martínez Franzoni, 2017;Weyland, Madrid, & Hunter, 2010).This phenomenon, which has cooled off in recent years, emerged as a counterpoint to the neoliberal hegemony that had been imposed on the region in the 1990s.Despite differences between the left-wing governments in each country, they shared their defense of state action for putting into effect political projects for income redistribution, the expansion of rights and social justice.
The expansion and innovation of social policies were central to these political projects (Huber & Stephens, 2012).This did not mean, however, that all social issues were incorporated into left-wing projects, and that when they were incorporated the result was homogeneous.Regarding gender inequalities, even though the left turn in Latin America provided a favorable environment for policies aimed at tackling the problem, the results have been mixed.Prioritizing the political agenda for gender equality was not a common denominator of these governments in the region (Blofield, Ewig, & Episcopo, 2017).Specifically in relation to policies aimed at linking work and family responsibilities with women's economic empowerment, there were notable variations between countries, especially with regard to the reorganization of the provision of care for facing up to the sexual division of labor (Blofield et al., 2017;Filgueira & Martínez Franzoni, 2017).
Brazil was a part of this left turn.The Labor Party (PT) in the country led the Federal Government from 2003 to 2016 1 .Some of the Brazilian feminist movements took part in the coalition of political subjects that led to the PT being elected, and the path of many feminist activists in the country has even been marked since the re-democratization by their "dual militancy"; they were active in the various movements, and in parties and trade unions (Souza-Lobo, 1991;Alvarez et al., 2003;Silva, 2016).Policies for women were instituted during the PT administrations by the creation of the Department of Policies for Women (SPM), which reported directly to the President of the Republic's Office, and enjoyed the status of a ministry (Bandeira, 2005).The SPM was responsible for coordinating government actions for women, such as those provided for in the three editions of the National Plan of Policies for Women (PNPM), in which gender mainstreaming was established as a management strategy (Brazil, SPM, 2004;2008;2013).The PNPM took as its basis the guidelines issued by the National Policy Conferences for Women (CNPM), which involved nearly 200,000 women in the process in just one edition, from the local to the national stages (Brazil, SPM, 2008).
Construction of the sexual division of labor as a public problem that demands that the State become co-responsible for resolving it by way of public policies was part of the historic struggle by feminist and women's movements in the country (Souza-Lobo, 1991;Silva, 2016).Movements demanded measures such as daycare centers and parental leave during the CNPM, which included them as actions, the aim being to promote the economic empowerment of women and their inclusion in the world of work.As child-care policy was not the direct responsibility of the SPM, however, it should, in a mainstream way, be re-oriented by gender equality perspectives, which was expressly provided for in the PNPM.But did this happen?
In this article we reflect on this question by analyzing gender mainstreaming in child-care policy (daycare centers and parental leave) in Brazil during PT government, considering the disputes and the resistances to the inclusion of gender equality in this policy (Marcondes, 2019;Marcondes, Farah & Alves, 2019).To do so we mobilized elements from one of the interpretative aspects of public policy analysis, which is frame analysis (Braun, 2015;Fischer, 2003;Rein & Schön, 1993) linked to the concept of gender mainstreaming (Bacchi, 2005;Papa, 2012;Verloo & Lombardo, 2007).Starting from this theoretical link, we analyze the official discourse of the (re)instituted policy, which is mainly derived from documents, but complemented by interviews.
Based on the analyzed data, we argue that during Labor Party federal governments there were advances in institutional conditions for structuring a process of gender mainstreaming, with the creation of the SPM, the publication of the PNPM and the realization of the CNPM.Child-care policy was covered by this process, which put it in a dispute between two frames: "The right to education and child care" and the "Promotion of women's economic empowerment".Different subjects and points of view came together around each of them, with both frames becoming part of official government discourse, although the "Right to education and child care" prevailed, which marginalized gender mainstreaming in child-care policy (Shaw, 2002).Our analysis shows that the structure of the frame that prevailed had patriarchal traits.This is because they are embedded in state action, and, also, the frame is shared by the actors who mobilized it.This result reflects the power and domination structures that characterize gender relations, and that were intersected by other relations that lead to inequality.Defenders of the "Promotion of women's economic empowerment" frame were also unable to construct a comprehensive narrative that would respond, in its entirety, to the gender or care relationship, being limited to the "question of women".This also contributed towards its marginality.This article is organized into five parts, including the introduction.In the second part, we outline the theoretical framework, and then detail the methodology used.The fourth section is dedicated to the results.Finally, in Final Considerations, we present a summary and point out some of the limitations and the possible contributions of this work.

Gender mainstreaming in care policy
The public policy frame and gender mainstreaming Recognizing that ideas and arguments are central has been a trend in public policy studies since the 1990s (Farah, 2018;Faria, 2003;Muller, 2000).One of the approaches to the theme is the argumentative turn, which suggests that the contributions of the linguistic turn in Western philosophy should be incorporated into studies in this field (Fischer & Forester, 1993).Argumentative turn encompasses a multiplicity of theoretical and methodological approaches, one of which is policy frame analysis.
Frame analysis first appeared with the micro-sociological studies of Goffman (1974), who emphasized that the frame was a form of identity representation.In later years, social movement studies started using it to understand the agency of social movements as an ability to "frame" problems and solutions using strategies that mobilize beliefs and values (Alves, 2014;Rosa & Mendonça, 2011;Snow & Benford, 1988;Snow, Rochford Jr., Worden, & Benford, 1986).Scholars of political discourse, such as van Dijk (1985), worked on frame analysis on a deeper, cognitive level, over which the capacity for individual intervention is small.Public policy studies, on the other hand, incorporated frame analysis on the borderline between full and limited agency capacity (Braun, 2015;Fischer, 2003;Rein & Schön, 1993).
Frame analysis was used in public policy studies to provide the values, expectations and beliefs that shape public policies with visibility, which it did by bringing together different approaches, as Braun (2015) observed.Among these approaches is the one we used in this study, which, in connection with an intersectional gender stance, values the relationship between frames, ideologies and discourses, and places power relations, social exclusion and domination at the center of its analysis (Braun, 2015).
In these terms, we understand frames to be structured narratives about public problems and the course of state action that are anchored in ideological patterns of meaning and are the basis of contextual interactions between subjects in a dynamic of disputes and alliances about these meanings, which can contribute towards establishing and legitimizing power and domination relationships, but that also challenge and transform them.These structured narratives, which are produced and mobilized by the subjects, encompass both the social construction of the public problem (diagnosis), as well as proposals as to what action the state should take in responding to it (prognosis) (Snow et al., 1986;Snow & Benford, 1988;Bacchi, 2005;Verloo & Lombardo, 2007;Rosa & Mendonça, 2011).
The definition of frames we adopt places power and domination relations at the center, building bridges between feminist and public policy studies.These bridges have been explored by feminist researchers who have used them, for example, to analyze gender mainstreaming (Bacchi, 2005;Verloo & Lombardo, 2007;Lombardo & Meier, 2008).Since the Fourth World Conference on Women, held in Beijing in 1995 gender mainstreaming has become a widespread strategy for introducing the widely publicized gender equality policies (Walby, 2005;Bandeira, 2005;Papa, 2012).By incorporating mainstreaming into the course of state action, the objective is to produce policies that are committed to gender equality and to reorient existing policies by way of the same commitment.
Gender mainstreaming includes a process of ideological dispute over the narrative that structures the definition of public problems and the course of state action from a feminist perspective (Bacchi, 2005;Lombardo & Meier, 2008).In this sense, and starting from the literature mentioned above, we define it as a process of structuring gender equality policies, in which feminist perspectives (or gender equality) are incorporated into public policy frames, both in constructing the public problem, and in defining the course of state action.
In order for the gender mainstreaming process to materialize, institutional conditions need to be constituted in the State so that public policies adhere to feminist political agendas, bearing in mind both the demands and the proposals forged by feminist female actors (Farah, 2004;Fujiwara, 2002).These institutional conditions may combine instances and mechanisms of management and social participation, such as women's policy bodies, public policy plans, and public councils, conferences, and audiences.
Although the government is committed to gender equality, which is expressed through the development of the institutional conditions needed for gender mainstreaming to occur, there are structural limits that prevent this commitment from materializing.This is because gender mainstreaming focuses on the roots of the power and domination relations that constitute the social reality.Consequently, making it effective tends to be non-linear and marked by contradictions and nuances (Walby, 2005).In fact, it can vary between a central and transforming incorporation on the one hand, and a more marginal result on the other, as Shaw (2002) observes.This gradation reflects the high degree of conflict involved in mainstreaming, to the extent that this process opposes both the patriarchal conceptions that are crystallized in the course of the state action, and those mobilized by subjects in the daily life of public policies.
Understanding gender mainstreaming as a process marked by contradictions and nuances is recognizing the complexity and multiplicity of institutions and of the actors who are involved in the course of the state action (Morgan & Orloff, 2014).If, as Morgan and Orloff (2014) observe, a good metaphor for symbolizing the State is that of Kali, the many-armed Hindu goddess, mainstreaming implies a dispute about each one these arms that does not necessarily succeed in reorienting all of them.

Gender mainstreaming in care policy
Although the concept of gender mainstreaming can be used to discuss all public policies (Farah, 2004;Bandeira, 2005), it is also possible to use it to investigate specific ones (Bacchi, 2005;Verloo & Lombardo, 2007).One of these policies is care, which we understand as being a type of social policy, whose purpose is to guarantee conditions for meeting social needs by way of faceto-face interactions between those who care and those who are cared for, linking work, emotions and ethical standards of conduct (Aguirre, 2009;Guimarães, Hirata, & Sugita, 2011).
At first glance a care policy is of benefit to those whose needs must be met, such as children, the elderly and the disabled.However, when we consider that, historically, women have been socially constructed as being responsible for providing such care, as a result of the sexual division of labor, these policies can also have an effect on gender relations.Thus, despite the fact that this is a policy whose relationship with feminism is not obvious, as policies for combating violence or guaranteeing sexual and reproductive rights are, when we add a gender equality perspective these interconnections begin to appear.This is, therefore, a type of social policy that can make the State co-responsible for the social reproduction of human life and, also, for facing up to the problem of the sexual division of labor.
The interest of gender studies in care policy is justified by the fact that women are the main caregivers, either in a paid (professional care) or unpaid (family care) capacity, which is one of the effects of the sexual division of labor (Batthyány, 2009;Martinez Franzoni, 2005;Tronto, 2009;Guimarães, Hirata, & Sugita, 2011).Consequently, the democratization of care relationships and making the State co-responsible for them is a fundamental challenge to gender equality in the world of work.This scenario gained new contours with the resurgence of what some feminist theoreticians called the care deficit (Aguirre, 2009;Batthyány, 2009), which is the result of an increase in the demand for care (such as the problem of population aging), coupled with a reduction in the availability of caregivers (women, most of whom are in the labor market).In response, the adoption of a care policy can improve living conditions for those who are cared for, but also for those who are the traditional caregivers (Pautassi, 2007;Tronto, 2009).In relation to child care, this reorganization of social responsibilities can help reverse the tendency of a drop in the fertility rate, which is also a component of the care deficit (Aguirre, 2009;Batthyány, 2009).One of the desired effects of adopting these policies is a reduction in the burden on women resulting from their exercise of motherhood.
A gender mainstreaming strategy in care policy initially points to the fact that the provision of services to meet social needs is based on a hegemonic patriarchal ideology, anchored in familism and maternalism (Aguirre, 2009;Martinez Franzoni, 2005).In these critical terms, the social construction of the idea of "care" is understood as being one of the primary responsibilities of the family, with state action being subsidiary; this is a private matter, not a public problem.As a result, women are disproportionately held responsible for providing care in families, a situation that stems from their association with motherhood.
Over and above denouncing the existence of patriarchal elements in public policies, in a propositional and transformative second movement gender mainstreaming seeks to redefine care as a public problem by attributing co-responsibility for it to the State and to men, in order to meet the needs of those who are cared for, and to promote equality with regard to those who care (Aguirre, 2009;Pautassi, 2007;Tronto, 2009).As we saw earlier, however, the mainstreaming process is marked by contradictions and nuances.It means that the "Kali's arms" (Morgan & Orloff, 2014) can move in different (and even conflicting) directions.
Finally, because of gender mainstreaming's procedural character, it admits not only nuances, but also the possibility of multiple feminist perspectives (Bacchi, 2005;Verloo & Lombardo, 2007).This means that mainstreaming in care policy can focus on: women and/or the gender relationship; caregivers, those who are cared for, and/or the care relationship (considering all those in this relationship); the gender relationship and/or its intersection with other forms of inequality (e.g.class, racial or generational); and the economic empowerment of women, protecting motherhood, and/or male co-responsibility (Martinez Franzoni, 2005;Pautassi, 2007;Aguirre, 2009;Batthyány, 2009;Guimarães, Hirata, & Sugita, 2011).This is the theoretical framework on which we build our methodological pathway.

Methodological pathway
This article presents some of the results of research into gender mainstreaming in child-care policy in Brazil, Argentina, and Uruguay during these countries' left turn at the beginning of the 21st century (Marcondes, 2019).To carry out the research we used a qualitative case study strategy (Gerring, 2010;Stake, 1998), which enabled us to conduct a number of in-depth observations in a contextual and complex way, and to refine the theoretical framework that we developed to integrate gender studies and public policy studies.
In this work we focus on that part in the research results that referred to Brazil and to PT's federal government administrations (2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016).During this period, child-care policy not only expanded, but was transformed.In fact, the rules of the leave regime (maternity and paternity leave) were modified (Brazil, 2008;2016) and the policy of daycare centers was completely reorganized in the wake of the transition from it being funded by social assistance to being funded by education (Marcondes, 2013;Cruz, 2017).
Policies for women were instituted in the federal government as of 2003, with the creation of the SPM, which then started coordinating these policies (Bandeira, 2005).To do so it adopted gender mainstreaming, with social participation as a strategy.This materialized through initiatives such as three editions of the PNPM, which in its turn started with guidelines that had been agreed upon with the feminist movements during the CNPM.The PNPM brought together actions that made government agencies responsible for policies for women, which were then managed by the PNPM's Articulation & Monitoring Committee (Brazil, SPM, 2004;2008;2013).In these plans, daycare centers and leave were adopted as actions for promoting women's economic empowerment and their equality in the world of work.
Regarding the analyzed policy, we chose the initiatives that deal with the care of children from birth to three years old.Since the re-democratization process in Brazil, demand for such policies has been on the agendas of both feminist and women's movements, and those dealing with the defense of children's rights, and this was a decisive factor when it came to daycare centers and leave being included as rights in the Federal Constitution of 1988 (CF-88) (Marcondes, 2013;Rosemberg, 1984).This age group also accounts for a high degree of demand for care, considering the level to which children of this age are dependent on those who care for them.It is also at this stage that it is least socially acceptable for such care to be provided outside the family, or by people other than the mother (Batthyány, 2009).The heavy gender load in the construction of visions and points of view on care for young children enriches the discussion about mainstreaming in care policy.
In our investigation we focused on two initiatives: daycare center policy (education and care services for children up to 3 years old) and paid leave (maternity, paternity and parental).Daycare centers enable us to reflect on the issue of "(de)familization" by establishing the co-responsibility of the State (Martinez Franzoni, 2005;Pautassi, 2007).Parental leave does not encourage "defamilization"; it just protects the care provided by families (Pautassi, 2007).It can mean ensuring rights and protecting maternity, but also reproducing the maternal logic of caring when it is maternity leave.Paternity and parental leave, on the other hand, can help increase coresponsibility 2 .The choice of these two initiatives also enabled a discussion on both direct and indirect provision, on regulation and on the transfer of public funds.
Construction of the analyzed corpus focused on official discourse with regard to the (re)instituted policy, and was based on the documentary analysis of: binding normative acts (e.g.laws and decrees) and non-binding acts (e.g.legal opinions); proposals for normative acts (e.g.bills); and official records (e.g.shorthand notes).We complemented this analysis by interviewing two government representatives.
We used a methodology for data analysis that was developed by the Mainstreaming Gender Equality in Europe (MAGEEQ) research project, which investigated gender mainstreaming in the frame of public policies in the European Union (Bacchi, 2005;Verloo & Lombardo, 2007) 3 .We combined this theoretical-methodological framework with contributions from discourse analysis (AD).By discourse we understand the contextual interactions of subjects who, through language (spoken and written), (re)construct representations of the world, and (re)produce effects of meaning (Fairclough, 1989;Orlandi, 2000).These representations of the world are permeated by ideologies, which give patterns of meaning to these interactions.In this sense, frame analysis is largely an analysis of discourse and ideologies (van Dijk, 1985).
In adapting the MAGEEQ methodology to fit our research, we reconstructed the structured narratives of the policies we selected for analysis.We defined a script of questions for diagnosing the problem and the prognosis of the course of state action considering the gender perspectives mobilized (Bacchi, 2005) and their intersectionality with other (re)producing dimensions of inequality, such as class and race (Hill Collins, 2015).Each of the variables in the questions was transformed into code.The table below identifies each of the ten codes used in this step.Bacchi (2005) and Verloo & Lombardo (2007) We used these codes to analyze the selected corpus by way of the Atlas.tiqualitative research software.Based on our full reading of the texts, fragments ("subtexts") were selected, the gender meanings of which were explained ("super-texts") using the above-mentioned codes (Verloo & Lombardo, 2007).By associating blocks of super-texts, we were able to reconstruct the frames.In presenting the results of our analysis, and in line with the AD tradition, we reproduce some of these fragments verbatim.
In our investigation we focused on frames that mobilized the gender perspectives in dispute (patriarchal and feminist).In doing so we restricted ourselves to elements that are based on the data, which does not mean that other frames might not have been present in the Brazilian reality in the period.

Contextualization
An important milestone in defining daycare centers and leave as rights, and which resulted from the struggle of organized civil society, such as feminist and women's movements, in defense of children and education, was the Federal Constitution of 1988 (Rosemberg, 1984;Marcondes, 2013).Daycare centers were provided for in it to guarantee the right to work and to education in early childhood.Paid parental leave also acquired constitutional status, and covered the public, private, urban, and rural sectors, being 120 days for maternity leave and 5 days for paternity leave.
Although the constitutional text represented progress, it also had its limits.Leave, which is restricted in its scope since it mainly covers the formal sector, is emblematic of this 4 .It is also important to point out the notable discrepancy of the length of time that is guaranteed in each of these modalities (maternity and paternity) within the CF-88 frame.Implementation of the constitutional text also encountered obstacles in the subsequent decade.In the case of daycare centers, they continued to be linked to the social assistance policy, mainly because of the way they were funded.
During PT governments there were changes in daycare centers and entitlement to leave regulations.Daycare centers were fully incorporated into the educational system when the Fund for the Maintenance and Development of Basic Education and for Valuing Education Professionals (FUNDEB) was instituted.The federal government also created initiatives for expanding the funding of daycare centers -a stage in teaching that is the responsibility of the municipality -like the inclusion of the early childhood education system in the second edition of the Growth Acceleration Program (PAC-2), and the creation of Brasil Carinhoso [Loving Brazil] (BC) (Marcondes, 2013;Cruz, 2017).By including the early childhood education system in PAC-2, funding was secured for the building work needed to expand the public municipal network, and for purchasing furniture.The BC, in its turn, provided for supplementary funds from Fundeb as encouragement so that children from the low-income families that were beneficiaries of the Bolsa Família [Family Income] Program could be enrolled in municipal public daycare centers (including the affiliated network) 5 .
During the process of incorporating daycare centers into formal education, their identity was debated by the National Education Council (CNE).Linked to the Ministry of Education (MEC), the CNE is responsible for regulating the National Education Policy.The Council comprises specialists from the education area in its composition.It was up to the CNE to define general guidelines for the operation of daycare centers (Brazil, CNE, 2009), in addition to resolving specific issues, such as the period when the service would function during the year (Brazil, CNE, 2012) and the age of children to be included in the service (Brazil, CNE, 2016).
The main alteration in the leave regime was restricted to highly formalized sectors of work (Brazil, 2008;2016).This was because the main legal changes that occurred in the period guaranteed that leave (maternity and paternity) was extended to benefit categories of public service and formal employment in large companies 6 .Maternity leave was extended to 180 days as a result of the advocacy of the Brazilian Pediatrics Society (Tominaga, 2015), while paternity leave under the "Legal Framework for Early Childhood" was extended to 20 days (Brazil, National Congress [CN], 2016).This project, which was drawn up by a multiparty group, was articulated by the Parliamentary Front for Early Childhood, made up of more than 200 parliamentarians in dialogue with representatives from government and civil society, especially those from the childhood area, such as the Maria Cecília Souto Vidigal Foundation (Brazil, CN, 2016).
Daycare and leave were also provided for in the policy for women.Both were actions that made up the PNPM, which, as we have seen, was based on CNPM resolutions.The transversal management of the PNPM was coordinated by the SPM, which relied on help for doing so from articulation and monitoring instances, like the PNPM's Articulation and Monitoring Committee, which brought together the government agencies that were responsible for implementing the Plan's actions, and representatives of the National Council of Women's Rights (CNDM) (Brazil, SPM, 2004;2008;2013).In fact, the above-mentioned Committee was made up of those agencies that were In this context, two main frames emerged to signify this policy from a gender perspective.On the one hand was what we call the "Right to education and child care", which is mainly shared by civil society organizations that defend the rights of children, parliamentarians, representatives of government and education sectors and members of the CNE, while on the other was what coexisted with it in the course of the state action, which we call "Promoting women's economic empowerment", as mobilized by government representatives from the SPM and by feminist and women's movements that were active in the CNPM and CNDM.Over the following pages we present the narrative underlying each of them.

Predominant frame: The right to education and childcare
We introduce our diagnosis of this frame's issue by way of a fragment taken from a speech made by a parliamentarian during the debate about the "Legal Framework for Early Childhood": (1) Today we know that all the basic competences, abilities, intelligence, socioemotional skills, impulse control, and peaceful behavior, or the opposite, violent behavior, are developed and shaped in the first two or three years of life.Human beings are then going to live with this for the rest of their lives.The pattern of stress, the way in which a person is going to react when faced with difficulties, the personal and emotional anxiety and difficulties they are going to have when they suffer some setback, the way in which they react in relation to this is put together and organized in the first two years of life on the back of the relationship they have with their caregiver, and the way in which their needs are attended to (Brazil, CD, 2014a, p. 1).
From this extract it seems that many of our cognitive and emotional faculties are shaped in the early years of life.Even violent behavior in adulthood may reflects care practices that are incompatible with the complex web of needs of the first years of a child's life.Early childhood, therefore, constitutes a critical phase in becoming human, which can become a problem and have a detrimental effect on the entire cycle of life.The critical point, therefore, is the relationship that the child "has with the caregiver, in the way their needs are attended to".Regarding to this relationship, the CNE's formal opinion, which analyzes the age at which children in daycare centers receive care, explains that: (2) Defense of immediate enrollment is normally based on the mother of the child being unable to do this work, since she has no one with whom to leave her child safely.Everybody wants to guarantee that children have the care necessary for their full development.This wish, however, reveals one of the great paradoxes of the daily lives of needy people.On the one hand there is the legal and moral duty of the parents to support their children, keep them safe and ensure they develop fully, while on the other they are obliged not to leave them with strangers, or even abandon them (Brazil, CNE, 2016, p. 6).
This second extract identifies a conflict, in which "mothers" are unable to enter the labor market as they do not have an arrangement available to them that allows their children to be looked after safely.As a result, a paradox arises between the duty of "the parent" as provider and as caregiver.In other words, what makes this phase critical is the way in which care is given and the concrete conditions that allow this to happen.This leads us to an investigation of the causes that were attributed to the production of this problem when this frame was constructed.The speech made by a deputy during a discussion about the framework of early childhood introduces us to these causes: (3) We pay a price for modernity, which requires that all members of the family, and here I am talking especially about women, who make a super-human effort for survival.Women assume new roles day by day because they are in the labor market.Their children, however, still need care.They must reconcile apparently opposing demands (Brazil CD, 2014b, p. 1).
The new roles played by women (such as being included in the labor market) are features of modernity, but they do no the fact that children still need care, which results in a tension between opposing demands (children and care vs, women, and work).This tension is central to the narrative: (4) Companies cannot forget the work of women, whose contribution is so important, and at the same time, they have to take into account that, since the natural role of women is to bear children, this role has to be respected and supported for the happiness and well-being of everyone (Brazil, CD, 2007, p. 2).
The fourth extract was taken from the speech of a deputy in the debate about maternity leave.In it he identified a conflict that arises between the contribution made by paid female work and the natural role of the "mother", on whom everyone's happiness and well-being depend.According to this argument, there is a mother-child conflict, which it is the woman's responsibility to reconcile.It would be legitimate to ask what role the State plays in resolving this conflict, which brings us to the following fragment: (5) First of all, the State needs to assume its responsibility for the collective education of children as a complement to family action.Second, daycare centers and nurseries are a strategy for promoting equal opportunities for men and women, since they allow women to realize themselves beyond the domestic context (Brazil, CNE, 2009, p. 5).
The objectives of state action are hierarchical.First, early childhood education and then promoting equal opportunities for men and women so that the latter can realize themselves beyond the bounds of the home.It is important to note that the family is identified as being the main agent in the collective education of children, with the State being responsible for complementing it; it is not co-responsible for it.
The logic of the functioning of daycare center services is shaped by an understanding that, in the pursuit of hierarchical objectives, the State's responsibility is complementary.This is what we understand from the discussion about the times and periods when such centers should function.In fact, the CNE was instigated into analyzing this issue, because questions were raised about services being closed at weekends and during holidays and school breaks, which did not meet the needs of families.According to the Council: (6) Many families need help with their children in periods and at times that do not coincide with those of the regular functioning of these educational institutions, such as at night, on weekends and during vacation and school-break periods.This type of service, however, which is a legitimate demand by the population, falls within the scope of "Early Childhood Policies", and must be funded, guided and supervised by other areas, such as social assistance, health, culture, sport and social protection (Brazil, CINE, 2012, p. 4).
The discourse in Section 6 recognizes the demands of families as legitimate.The way they are served, however, is the scope of another policy (that of childhood), which is not structured in Brazil to provide such a service; only exceptionally has it been allowed to use the physical space of daycare centers (Brazil, CNE, 2012).In the prognosis the argument does not recognize the paradox of social care practices as a problem.Therefore, no solution is offered.Diagnosis, therefore, is separate from prognosis in the structure of this narrative.
The meaning in this frame is layered.Care, associated with education, is assumed to be a right, a hypothesis in which there is a partial and legitimate "defamilization".It is the educational function that defines the boundaries between the care that must be exercised by families and that which it is the State's responsibility to provide in a complementary way.In other words, educating and caring are inseparable, but the latter is only recognized as being part of the State's legitimate sphere of action when the former also is.When caring falls outside the limits of educating it is not the State's responsibility.On what lies beyond these limits, this frame "is silent", which is also a way of producing meaning (Orlandi, 2000), because silence ends up reinforcing the attribution of the role of care to families and, in families, to women.Consequently, silence limits the potential of gender mainstreaming, which in its fullness implies making the state co-responsible for care, and ensuring it assumes an active stance in relation to the issue.
The following fragment adds another layer to this texture: (7) According to the constitutional imposition (Art.227), these intervals allow children to enjoy family and community life.That is why the duty of parents is to help, raise and educate their children (Brazil, CNE, 2012, p. 3).
According to Section 7, the service provision interval is a condition that allows mothers and fathers, in addition to living with their children, to exercise their duties to help, raise and educate, which are imposed by the Constitution.This passage introduces us to the implicit postulates of this policy, which bear traces of patriarchal ideology, in conjunction with other forms of inequality.It is assumed that families are available in these intervals to live with their children, regardless of the material conditions of survival of each family unit.This is even a constitutional obligation.There is a second obliteration, here, referring to rights enshrined in the CF-88 not being upheld (the right of caregivers to work); in other words, the right to work and education are not satisfactorily met by daycare centers, but families are nevertheless required to meet their related obligations.
In addition to family traits, there are maternal elements in this narrative.The abstract categories "families", "parents" and "caregivers" refer to women, actually.This is because their "natural" role is that of bearing children (Section 4), and they must reconcile their demand to be included in the labor market with their children's demands for care (Section 3).Only recently have parents been incorporated as subjects.As a female deputy said in the discussion on the Legal Framework for Early Childhood: (8) Mr. President, I wish to congratulate you and all men who yesterday received an exceptional gift.With the Legal Framework for Early Childhood, which was sanctioned by President Dilma, you extended to 20 days it should be more the right to paternity leave so fathers can be with their children.This is a contribution so fatherhood can be assumed as something that is a responsibility of society and not just of women.Congratulations to the men, women, and children of this country!(Brazil, CD, 2016, p. 1).
For men, extended paternity leave meant "receiving" an "exceptional gift".Because of it the problem of paternity and the solution of sharing were legitimized, albeit in a limited way ("it should be more"), and although in a secondary way, male co-responsibility in this frame was directly related to the standard meaning attributed to motherhood.
It is also important to note that the measures dealing with leave, although they were discussed in a generic sense as being a guarantee of rights, are restricted to formalized sectors, an experience that is marked by class, gender, and race relations.As we have seen, the extension of leave in the period, benefited only the most formalized sectors of the labor market (public service and large companies, and it was optional for the latter).In these terms, specific situations of subjects that are permeated by inequalities are taken as being universal in the narrative surrounding these policies.
In the case of the policy dealing with daycare centers, there is a level of concern with the social (class, gender, race and ethnic) and territorial (regional, urban/rural) inequalities of its guidelines and diagnosis, because these are factors that shape the asymmetries of access to and permanence in the service (Brazil, CNE, 2009).These references to inequalities, however, are generally generic, and do not result in a systemic reorientation of this frame, even though occasional advances have occurred, like the BC, whose aim was to benefit poor children.
In short, this frame, in which constructing the meanings of child-care policy of the period predominated, focused on the rights of children to education and care.There are traces of patriarchal ideology in the construction of this narrative, which combines family and maternal elements.These elements are largely crystallized in the course of the state action, which the frame reproduces by reaffirming and legitimizing the relationships of power and of gender domination.This ideology, however, is not only present because of the reproduction of discursive memory; it is brought up to date by the subjects who mobilize it discursively, and it crosses between political positions and brings together multiple party and organizational affiliations, as well as voices on the left and the right.

Marginal frame: Promoting women's economic empowerment
A second narrative on child-care policy emerged from policies for women.Their meanings refer to the question of what women's work is.In fact, especially after the second edition of the PNPM (Brazil, SPM, 2008), daycare centers and maternity leave were incorporated as actions on the axis dedicated to women's economic autonomy and equality in the world of work.In its third edition (PNPM 2013(PNPM -2015)), paternity and parental leave were also provided for (Brazil, SPM, 2013).
Fundamentally, the diagnosis of this frame was that: (9) Despite increases in the activity level of women, it is still far below that of men (...) in this process, the places occupied by black females tend to be more precarious than those occupied by white men (Brazil, SPM, 2008, p. 32) In this narrative of the problem, there were fewer women in the labor market (activity level) than men, and they also found themselves in worse conditions (precarious positions).An understanding of the causes of this problem can be extracted from the following: (10) The economic empowerment and equality of men and women in the world of work are based on specific actions that aim to eliminate the unequal sexual division of labor, the emphasis being on eradicating poverty and guaranteeing the participation of women in the development of Brazil (Brazil, SPM, 2013, p. 14).
The cause of the inequalities between women and men that are identified in this frame is the sexual division of labor, which separates and hierarchizes the experience of the world of work according to gender relations (Kergoat, 2009).Women perform less valued functions, or those that are not recognized as work (such as unpaid domestic work), while the spaces occupied by men are more valued.Promoting women's economic empowerment in the world of work implies attacking the sexual division of labor, including eradicating poverty and ensuring that women become part of the country's development.To achieve this, it is necessary to: (11) Expand the supply of public equipment and policies that favor an increase in the time available to women, thus promoting their autonomy, including their insertion in the labor market (Brazil, SPM, 2013, p. 17).
Expanding the supply of equipment and policies that provide women with free time, including their insertion in the labor market, is a condition for promoting women's economic empowerment, which is a fundamental objective of this frame.There is, here, a distinction in relation to the frame that was previously analyzed.The insertion of women in the labor market is not what causes the problem of child care.It is the unequal sexual division of labor, which makes women disproportionately responsible for caring.Insertion in the labor market is part of the solution to the problem in this second narrative.This frame mobilizes a specific gender equality perspective, in which the focus is on women, and not on the gender relationship or on the care relationship.The women who were interviewed made this clear: (12) (...) Because discussing parental leave -'I don't want to because it's going to upset men.We don't upset men here, do we' (...) I've already discussed this several times in meetings... 'we're here to discuss more rights for women; longer maternity leave for women'.At the end of the day, I don't think the Department was ever very concerned with changing gender relations.Their concern was: 'let's guarantee rights for women' (...) (Interviewee 1).
(13) There was general criticism of this matter in the SPM.In fact, it was frequently said that the focus of the SPM was not on children (Interviewee 2).
Concern for children was not neglected; guaranteeing their rights was a means of guaranteeing women's rights by way of initiatives such as daycare centers.The priority, however, was for women to have access to the formal labor market, thereby overcoming poverty and social exclusion.Thus, the emphasis was on women's economic empowerment.Although this was the perspective that prevailed in this frame, others also emerged, as we can see from some of the actions of the PNPM: (14) Promoting unpaid domestic work being valued and helping overcome the current sexual division of work (Brazil, SPM, 2008, p. 39).
In addition to the importance of valuing unpaid domestic work, the second PNPM (2008)(2009)(2010)(2011) extended maternity leave to six months.However, it made no provision for paternity leave.In this sense, not only did the economic empowerment of women become more relevant, but so did the protection of motherhood.Although these perspectives are not necessarily contradictory, they have their peculiarities.In the first case, women's equality with men is valued, while in the second what women traditionally do (caring for others) is valued.
Another perspective of gender equality emerged in this frame from the PNPM (2013)(2014)(2015) in an even more erratic way.It can be identified in an extract taken from the Plan; whose diagnosis predicted that: (15) Reducing the time women spend on housework is a task for public policies for economic empowerment.This is how they can become part of this plan of actions for facing up to this particular form of inequality, with the setting up of public equipment and a wider debate about sharing the use of time and of coresponsibility for housework (Brazil, SPM, 2013, p. 14).
As a result, the actions of the plan were: (16) Promoting a culture of housework being shared between men and women by way of campaigns, extending paternity leave and engaging in debate about parental leave (Brazil, SPM, 2013, p. 18).
Based on the PNPM (2013)(2014)(2015), male responsibility became more prominent than in previous editions, even though the forecast actions were generic (extending paternity leave, but without proposing an extension period; discussing parental leave; etc.).Even with these limitations, elements that signal a concern not only with women's rights, but with gender relations, were incorporated at the time.This means not only ensuring that women have equal conditions for entering the labor market, and that the social roles they have traditionally assumed are valued, but that men should be encouraged to engage in these care relationships.
In summary, women were prioritized in this frame, as evidenced in its main mechanism of managing mainstreaming (PNPM).In fact, this was an assumed choice for "mainstreaming" gender in policies, since it was understood that by guaranteeing women's rights and achieving the social inclusion of women gender relations would also be modified (Fujiwara, 2002;Farah, 2004;Bandeira, 2005).Government nomenclatures explain this conception: Department of Policies for Women, National Plan of Policies for Women, etc.Even though this decision to focus on women (instead of on gender relations) may be important, it also has a limiting factor, especially in cases such as the policy we have analyzed in this article, which does not just deal with the rights and needs of women.
Finally, there were also some efforts to promote intersection between gender inequality and other social relations that produce inequalities, such as race and class (Hill Collins, 2015).Section 9, extracted from the Second PNPM, is emblematic of the attention paid to black women in diagnosing inequalities in the labor market.In this Plan, the need to consider "the ethnic-racial, generational, regional and disability dimensions" was still explicit (Brazil, SPM, 2008, p. 39).As a rule, however, this more general guideline did not develop into specific child-care actions.Furthermore, the gender perspective adopted in this frame was marked by hetero normativity, which implies assuming heterosexual affective relationships between women and men as being the rule, which fails to afford visibility for other arrangements, such as those of LGBTQ + families.
This second frame shared some aspects with the first.Both made use of a grammar of guaranteeing rights, when they set out actions for ensuring improvements in the living conditions of historically discriminated subjects (children and women), which was in line with the discourse of expanding policies and social rights and gained strength during the left turn.Children's centrality in the first frame, and women's in the second, however, made integration difficult, also because of the gender perspectives underlying each of them.In both, there was also a growing concern about a third person (the father), but one who remained as an extra in the care relationship.A more comprehensive frame did not emerge, therefore (Snow et al., 1986;Snow & Benford, 1988;Rosa & Mendonça, 2011;Alves, 2014), to encompass the multiplicity of subjects and points of view that disputed the meaning of the child-care policy.For this to happen it would be necessary to bring together the needs of those who traditionally care (women), those who are cared for (children), and those who should care (the State and men).
The option for a less comprehensive narrative in this second frame (regarding both the gender relation and the care) may have contributed to the marginality of gender mainstreaming.However, the main reason for the marginality of mainstreaming, as we have seen, lies in the patriarchal ideological elements that the prevailing frame mobilized, and that reaffirmed and updated the familial and maternal logic of care provision.The table below summarizes the results.

Gender perspective
Patriarchal traits (familism/maternalism) in policy assumptions.Social and territorial inequalities are mentioned, but there is no intersectional approach; some of the measures benefit more formal sectors (e.g.changes in leave) Women's economic empowerment (main) and protection features and male co-responsibility (secondary).Class, racial, ethnic, regional, and other perspectives are guidelines, but there is no intersectional approach.Hetero normativity is not problematized Prepared by the author, based on research data

Final considerations
In this paper we analyze the child-care policy frame (daycare centers and leave) in relation to gender mainstreaming.Our investigation focused on the meanings that discursively structured public problems and the course of state action on child-care policy in Brazil during the country's turn to the left.
We identified that the problem of child care enjoyed legitimacy in the period, but what this meant was disputed.In the prevailing frame, which we call "Right to education and child care", the problem was constructed from an understanding that the first years of life are a critical phase that requires attention, even though issues of responsibility for care subsequently also emerged as part of the diagnosis.The related solution would be to promote partial "defamilization" by way of education services, from which care cannot be separated, but without changing the primary responsibility of families (and, within them, women) for caring.It was mainly government representatives, parliamentarians, and members of civil society from the fields of childhood and education who gave voice to the elements of this frame.
The predominant frame, which we call "Promotion of women's economic empowerment" coexisted with another in the course of the state action.Its landmark moment was when public policies for women were structured, a process that was coordinated by the SPM and engaged governmental and non-governmental female actors.In it, we identified nuances in relation to the gender perspectives mobilized in constructing a gender mainstreaming process.Greater emphasis was placed on women's inequalities in the labor market compared to men.We also identified other perspectives, such as valuing motherhood and male co-responsibility, although they were less central.Finally, some efforts were made to link gender relations with other forms of inequality.This was done, however, in a more generic way (such as general guidelines), or one-off way (poverty).There was, therefore, no intersectional approach.
In summary, the two frames we described were not integrated, but coexisted asymmetrically within the scope of the course of state action, which produced marginal mainstreaming in care policy (Shaw, 2002).Reorienting child-care policy by a gender equality perspective was limited to policies for women.Therefore, it was not effective, because it did not deal with the constructive meanings of daycare centers and leave in the period.
The results indicate that institutional conditions developed during PT governments for structuring a gender mainstreaming process, with the creation and integration of instances (e.g.SPM, CNPM, CNDM, and the Articulation and Monitoring Committee) and mechanisms (PNPM).The effects on child-care policy, however, were limited.This is because the frame that prevailed in this policy showed traits of patriarchal ideology, because it reproduced discourse that was crystallized in the course of the state action, and because it was present in the ideas mobilized by the subjects who gathered around this frame.This evidence seems to suggest that the creation of institutional conditions for gender mainstreaming is important, but insufficient for promoting substantive changes in the gender perspectives that guide child-care policy (in particular) and public policies (in general).To this end, the process of gender mainstreaming should be accompanied by a more forceful commitment by the State to this objective.Also, by a cultural change towards the whole society.Gender relations, intersected by other social relationships that reproduce inequalities, are power and domination relations that not only construct the mentality of subjects, or outline institutional dynamics, but structure social reality.Consequently, any attempts at a more profound change tend to face strong resistance.Finally, these results underline the importance of understanding the institution of gender equality policy as a complex and contradictory process, because of the multiple hands of the State (Morgan & Orloff, 2014), and the structural and intersubjective dynamics that reproduce and update inequalities.
It is important to note that neither the frame incorporated by the instances nor the policy mechanisms for women were able to offer a comprehensive narrative, which would respond to the gender or care relationship as a whole, and not only to the situation of women.Indeed, when the concern about who should care and who should be cared for was mobilized it was a secondary matter, which may also have contributed to the marginalization of gender mainstreaming.
With our analysis, we hope to contribute to the convergence efforts of gender studies and public policy, and support gender equality policy analysis and practice.We believe that the interface between gender mainstreaming and frame analysis in public policies can support more critical and less idealized reflections of gender mainstreaming, in which the radicality of the political dispute involved in this process is not lost sight of.By investigating these policies during the turn to the left, we also intend to provide empirical evidence on the potential and limits of left-wing governments in Brazil for addressing gender inequalities linked with other forms of inequality.
There are limitations in our analysis.In this article, we focus only on the Brazilian case, which did not allow a comparative approach, which enriches the understanding of the phenomenon of the left turn in Latin America.Our investigation did not cover the implementation of the policy we analyzed, which could provide more elements for reflection, nor did it include paid domestic work or other forms of paid care, such as the professions involved in early childhood education.We understand, however, that these issues do not weaken our work.By recognizing the limits of this work, it is possible to understand it as part of an ongoing process, which is the constitution of mediation between gender studies and public policy studies.It is with this effort that we intend this article should make its contribution to scholarship.
then responsible for education, social development, and labor policies (MEC, the Ministry of Social Development & the Fight against Hunger [MDS] and the Ministry of Labor and Employment [MTE]).