Frames in Contestation

The article looks at the translation of international norms on domestic violence to the national level in five Central and Eastern European countries. It argues that translation brings a concept of domestic violence, which stretches gender equality ideas underpinning international norms so as to be easier to endorse by mainstream policy actors, and results in policies framed in degendered individual rights terms. The potential for keeping gender equality in focus is then guaranteed by gendering policy processes through empowerment of gender equality actors at all stages. Absence of ownership of the policy by gender equality actors risks co-optation by frames contesting gender equality.


Introduction
Violence against women has been the flagship policy issue of the global feminist mobilization efforts to sensitize human rights to gender equality (Keck & Sikkink, 1998;Kelly, 2005;Merry, 2003). As a result of transnational feminist cooperation and advocacy, the concept of human rights was reformed not just by extending men's rights to women but also by changing the meaning of human rights toward a more gender-sensitive understanding (Kelly, 2005) through the inclusion of the private sphere as a realm where violations, particularly violence in all its forms, can take place. While the "women's rights are human rights" movement is widely considered a success in changing the meaning of human rights (Keck & Sikkink, 1998), doubts have been formulated about the payoffs of this strategy for connecting violence against women to gender inequality, particularly in the context of national policymaking (Kelly, 2005).
Researchers such as anthropologist Merry (2003) noted a discrepancy between the international and national framings in that "the international movement takes a far more structural approach" to violence against women in intimate relationships than individual countries (p. 944). But does transformative gender equality content disappear domestically?
This article sets out to understand the outcome of women's advocates' mobilization to translate international human rights norms on violence against women into their national legislation and policies in five Central and Eastern European (CEE) countries: Bulgaria, Croatia, Hungary, Poland, and Romania. The article focuses on just one form of violence against women, domestic violence-the issue that dominated policy debates in the field in these countries. In our attempt to capture gender equality content, we understand gender equality as a dynamic, contested concept that takes on different meanings among different contexts and different stakeholders (Lombardo, Meier, & Verloo, 2009). Lombardo et al. (2009) show that gender equality can be stretched, shrunk, or bent so that it can be filled with a variety of meanings that arise from different political histories, contexts, struggles, and debates. Within this framework, we assess the extent to which transformative gender equality content emerges in domestic violence policy debates and their outputs in the five countries of our study by tracing the contestation of gender equality in these processes.
Based on theories by Ferree and Gamson (2003), Verloo (2005aVerloo ( , 2005b, Krizsan and Lombardo (2013) and Johnson and Brunell (2006), the article develops a twopronged conceptual framework to study the resonance of policies with gender equality. We argue for the importance of both in securing gender equality outputs in policies: on one hand, policy content and its framing; on the other, the participation of gender equality advocates in policymaking and implementation. The latter becomes even more important in the case of policies framed in gender indifferent ways. We argue that in CEE countries, the translation of international norms into domestic policy (Merry, 2006b) brings a concept of domestic violence framed dominantly in individual rights terms, in which gender equality norms are stretched to meet the endorsement of mainstream policy stakeholders. This opens up avenues for co-optation by those approaches that contest gender equality. The key to securing policies against co-optation and nevertheless realizing the potential for transformative gender equality content is gendering policymaking and gendering implementation through the sustained empowerment of gender equality advocates throughout the process.

Method
Our study uses a discursive neo-institutionalist approach (Schmidt, 2010), which combines discursive policy analysis with the analysis of structural and contextual data on civil society actors, state institutions, and preliminary implementation data in the five countries to trace processes and to explain variation in the gender equality content of domestic violence policies currently in place. The discursive analysis is done by critical frame analysis (Verloo, 2005b) to develop a conceptual framework for understanding the variation of policy frames in domestic violence policies and their relation to gender equality and to map elements of contestation between the different policy frames present in the debates. Starting from social movement theorists' work on frame analysis (Snow & Benford, 1988) and Bacchi's (1999) "what is the problem?" approach, critical frame analysis defines policy frames as "organizing principles that transform fragmentary or incidental information into a structured and meaningful problem, in which a solution is implicitly or explicitly included" (Verloo, 2005b, p. 20). Policy frames can be distinguished by their diagnosis and prognosis and the normative approaches underpinning them. Diagnosis is about the definition of the policy problem and main actors affected, whereas prognosis is about proposed policy solutions, their target groups, and responsible actors. We conducted in-depth qualitative analysis of policy documents, parliamentary debates, and interviews with stakeholders in the five countries, and collected contextual and structural data about them.
The initial data for this study were collected within the framework of the European research project Quality of Gender + Equality Policies in Europe QUING. For each country in our study, the history of the policy area was mapped, and based on that history a sample of documents related to important policy shifts and in particular to the adoption of legislation and policies on domestic violence were selected and coded. The texts included bills, the passed laws and their explanatory material, the parliamentary debates around the legislative proposal, and interventions from civil society organizations in the debates. 1 Data collected in QUING were supplemented by additional document reviews in each country to study the implementation documents, as well as interviews with main civil society actors involved in the initial adoption of the legislation in each country. Data collection for purposes of this analysis was closed at the end of 2009. Therefore, important subsequent developments that occurred in some countries, such as the 2010 amendment of the Act on Counteracting Domestic Violence in Poland inter alia introducing the civil restraining order, or the 2012 amendments to the Law against Family Violence in Romania are not analyzed here. This is without detriment to the main arguments presented in the article.
The article first develops a conceptual framework for understanding the gender equality content of domestic violence policies. Next, it describes domestic violence policy processes and their outputs, highlighting the main frames used in contentions between reformers and their opponents and the role of civil society mobilization in the adoption of reforms in CEE. A discussion follows on the five countries in juxtaposition to international framework documents, which assesses the emergent transformative gender equality content in their policy outputs within the conceptual framework. The conclusions review the potential and pitfalls of the emerging pattern of gendering domestic violence policies.

Meaning Of Gender Equality in Domestic Violence Interventions
Domestic violence has become a public policy issue as a result of feminist advocacy, well-documented in the case of feminist shelter movements in the United States or United Kingdom (Dobash & Dobash, 1992), or that of transnational advocacy networks (Keck & Sikkink, 1998). Originally placed within the context of violence against women and gender inequality, the meaning of domestic violence as a policy problem and its solutions have been translated and adapted across the globe. Co-optation of the gender equality agenda has often been raised in the literature as an inherent danger when autonomous civil society activists engage the state and partner with state actors (Bumiller, 2008;Dobash & Dobash, 1992;Matthews, 1994;Merry, 2009). At the same time, in today's Europe, state policies form the core of responses to domestic violence, and partnerships between state and civil society are inevitable. This section develops a conceptual framework for analyzing gender equality in domestic violence policies.
Theories about gender and politics argue that to assess the long-term transformative gender equality potential of policies, scrutiny should move beyond only policy content. A dual approach is suggested: (a) an investigation of the gender-transformative content of policies, and (b) a component on the inclusiveness of policy processes. Along these lines, Verloo (2005a) distinguishes between two main contributions brought by gender mainstreaming: displacement and empowerment. Displacement refers to realizing the transformative potential of the mainstreaming strategy by policies that go beyond problematizing the exclusion of women to challenging the gender hierarchy of societies. By empowerment, Verloo refers to organizing political space so that nonhegemonic interests, women's interest in this case, have a chance to participate in the debates concerning the ongoing articulation of transformative gender equality content. Displacement makes a contribution in terms of policy content, whereas empowerment improves the process.
Similarly, in the context of abortion debates Ferree and Gamson (2003) differentiate between two aspects of women's empowerment: autonomy and authority. These dimensions are operationalized as the governance of gender (autonomy) and the gender of governance (authority), where the first means framing policies in ways that promote women's autonomy and the second means the participation of women in public authority and standing on issues of concern to them.
In their recent work, McBride and Mazur (2010) conceptualize the gender equality potential of state responses along two axes: the level at which policy content coincides with the women's movements' framing of policy problems and the level of inclusion of women's movement actors in policymaking. They point out the desirability of dual responses, characterized both by a policy content that falls along the lines of women's movement framing and by an inclusive policy subsystem integrating women's movement actors.
Our conceptual framework follows the lines of this two-pronged assessment of gender equality policies in terms of (a) gendering policy content, and (b) gendering policymaking and implementation processes.

Gendering Policy Content 2
The relationship between domestic violence policies and gender equality varies among approaches that explicitly link domestic violence to gender equality, approaches that resonate with gender equality objectives but remain implicit about its relevance to addressing domestic violence, and approaches that contest the premises of gender equality thinking in relation to domestic violence.
We conceptualize differences in the meanings of gender equality in policy contestations around domestic violence along a continuum from gendered to degendered policy frames. Variation along the continuum is characterized by the extent to which genderedness of domestic violence is spelled out by the policy, but it is also linked to differences in defining who is affected by domestic violence, particularly the extent to which domestic violence is seen as a group-based social problem (affecting women as a group or particular groups of women) or as a problem affecting individuals (many of whom happen to be women). The continuum of frames ranges from structural gender equality to rights of individuals subjected to domestic violence, with two sub-frames between them: women-centered frame and implicit gender equality frame. Differences among the four approaches are best marked in their diagnoses, while their prognoses overlap in many aspects.
Structural gender equality frames see domestic violence as a form of gender-based discrimination. This frame understands domestic violence as rooted in gender inequality. None of the other three frames on the continuum make this connection. Here, domestic violence is a universal phenomenon, occurring regardless of social status, education, culture, or deviancy. Women are explicitly identified as being disproportionately affected by domestic violence. The frame is best represented by international framework documents such as General Recommendation 19 of the Committee on the Elimination of Discrimination against Women (CEDAW) and the 1993 UN General Assembly Declaration on the Elimination of Violence Against Women. Very few countries follow this framing in their laws and policies. The two notable exceptions in Europe are Spain and Sweden.
At the other end of the continuum is the rights of individuals affected by domestic violence frame, labeled as degendered domestic violence elsewhere . This frame is centered on the individual and is focused on symptoms of the problem rather than social structural causes that perpetuate it . The mechanism seen to lead to domestic violence is not structural but rather individual: dependency, in economic or emotional terms, in which more powerful individuals victimize weaker, dependent individuals. The group aspect is neglected and a generic category of "victim" is used without any further specification. Although this generalized approach may be seen as inclusive of groups affected by domestic violence other than women (among them children, elderly, or disabled), the loss of specificity in naming the groups also implies neglecting the main social mechanisms behind its (re)production: gender inequality. Nevertheless, this frame resonates with the tenets of the structural gender equality frame in its understanding of domestic violence as a universal problem leading to the violation of human rights in the private sphere conditioned by dependency, a phenomenon taking place amid the tabooization of violence by society. The rights of individuals subjected to domestic violence frame is dominant in Europe, as the overwhelming trend in European Union member states is "convergence towards gender neutrality" in legal measures (European Commission, 2010, p. 62).
Two frames rest on the continuum between gender explicit structural gender equality and the gender indifferent rights of individuals subjected to domestic violence. Women-centered frames describe the phenomenon in gender-neutral individual rights terms, but specify statistically the heightened prevalence of domestic violence against women. However, this frame does not connect with social structural explanations. The victim group is defined as "victims including mainly women and children." Implicit gender equality frames remain gender neutral throughout and discuss "victims" in general terms, occasionally mentioning children but never women. They nevertheless place the policy issue within the context of gender equality policy by referencing important international framework documents on gender equality such as the Convention on the Elimination of All Forms of Discrimination against Women (CEDAW), or national gender equality laws and strategies.
Frames along the structural gender equality-individual rights continuum have a more solid common ground in their prognosis. All of them agree that a complex set of measures is required, including a number of elements that cannot be disregarded. They agree that domestic violence requires intervention into the sphere of privacy. They all propose a primary accountability focus on the perpetrator and not the victim. As a first pillar, a victim's safety has to be ensured through protection measures and empowerment strategies for her to step out from the cycle of violence. As a second pillar, the perpetrator should be sanctioned for the crime of domestic violence and restrained from continuing further violence. A third pillar of action is awareness raising and education to address domestic violence. The coordinated action of all relevant actors including civil society is seen as central by all four approaches.
The main specific prognosis element of the structural gender equality frame is consistent with its diagnosis: to address domestic violence at its roots in terms of gender inequality. This frame thus proposes the transformation of gender unequal societies and specifically aims to change gender stereotypes seen to be behind unequal social relations and ultimately behind domestic violence. Importantly, the frame also aims for the protection of victims through the feminist empowerment of women. Support for independent services for women, such as autonomous women's shelters, is seen as one crucial tool in this prognosis. This specific gender equality content is missing from the other three frames.
The prognosis of the women-centered frame is largely aligned with that of the individual rights frame. By making implicit references to gender equality instruments, implicit gender equality frames point to the embeddedness of solutions to domestic violence in the framework of wider gender equality policy, but do not spell out any specific measures (see Table 1 for an overview).
In policy debates on domestic violence, gender equality frames, whether explicit or resonant, are challenged by several contesting frames. One may assume that contestation would come from explicit anti-gender equality approaches that either support the inferiority of women, view domestic violence as their duty, or argue that the privacy of the family is absolutely beyond the remit of state interference. However, these extreme forms of contestation are rare in the current European debates . More subtle forms prevail that do not openly reject gender equality, but contest some of its key tenets as defined above in the structural gender inequality frame. Four groups of contesting frames can be differentiated: the family protection frame, the externalizing frame, perpetrators' rights frame, and a certain version of child protection. Family protection frames see families as the main victims of domestic violence and believe that families need to be protected in their unity. While this frame sees violence within the family as undesirable and something to be addressed by policy, its ultimate rationale is not to protect autonomous individuals but to cease violence, while saving the unity of the family. This frame stays away from diagnosing unequal power relations within the family and proposes as solutions family support or conflict mediation, as if all the partners in the conflict would be equal. This frame contests one of the most basic tenets of gender equality resonant frames: the autonomous individual in need of protection and empowerment.
Externalizing frames contest another tenet of the gender equality-individual rights continuum: that domestic violence is universal to every social group. These frames externalize the social problem by arguing that domestic violence is not a universal problem, but rather one that is typical for certain marginalized, culturally different, or deviant subgroups. Consequently, they link solutions to policies addressing these different patterns of social marginalization and difference without capturing the larger problem. For example, these frames may consider domestic violence to be exclusively a problem produced by alcoholism, or see it as a problem of ethnic minority groups.
A third contesting frame is perpetrators' rights, particularly articulated in policy debates discussing restraining and protection orders against perpetrators. It argues that the real victims in domestic violence are perpetrators whose basic rights to property, free movement, privacy, or religious freedom are curtailed without due process by forcibly preventing them from returning home, approaching the victim, or even entering church if that is a location where they could encounter the victim. In this argument, protection of the victims from domestic violence, even in severe cases that may endanger their life, is seen as a secondary priority to classic civil rights of perpetrators.
Finally, contestation of a gender equality approach to domestic violence may also come from child protection approaches. Child protection is brought into domestic violence debates in a variety of ways: Some are compatible and indeed complementary to gender equality resonant frames; others can be seen as contesting them (Hester, 2011). As mentioned above, gender equality resonant frames often discuss child victims along with women as the specific victim groups that are most vulnerable to violence. But importantly, gender equality resonant frames discuss child victims in inclusive ways, cognizant of the links between violence against children and violence against women. Gender equality contesting child protection frames argue in exclusionary ways that the real victims of domestic violence are children, and even if women are to be protected from violence this is not because of the violation of their rights as autonomous individuals, but due to their role as mothers and protectors of child victims. Policy responses in this approach only protect and support women with children but not other women. Yet another version of the child protection frame argues that the foremost interest of the child is to be raised and cared for in a traditional two-parent family. Therefore, the family has to be saved at any price; even a violent family is better than no family. In this version, the individual rights of the child are ultimately subordinated to family protection, and therefore the frame can be seen as a sub-type of the family protection frame.
The presented typology of domestic violence policy frames facilitates identifying the gender equality content of domestic violence policy texts. Although such content is straightforward in the case of policies framed in explicitly gendered terms or in terms that reject gender equality tenets, assessing gender equality content becomes more problematic in the case of texts framed in the degendered zone of the above continuum. It is particularly in those cases that we need to go beyond policy frames to assess the gender equality implications of domestic violence policies.

Gendering Policymaking and Implementation Processes
The importance of dual-track evaluation that pays specific attention to the inclusiveness of policy processes for assessing the gender equality implications of public policies has been specifically highlighted for addressing violence against women, and specifically domestic violence. The coordinated community response developed by the Duluth domestic violence intervention model 3 embodies the skepticism of domestic violence policy initiators about the feasibility of state action in the field without coordinated civil society partnerships. Martin (2007) in her endorsement of the coordinated community intervention model clearly emphasizes the crucial role women's rights advocates play in ensuring the gender equality quality of sexual abuse interventions through multidirectional coordination of action at every level of policy development, monitoring, review, and implementation. Dwelling on the coordinated community intervention model, Johnson and Brunell's (2006) domestic violence regimes emphasize the importance of both the level of state response (content, quality) and the coordinated action between civil society actors and the state for assessing the quality of state responses. Their analysis of domestic violence regimes in several Eastern European countries finds that the most reformed domestic violence regimes emerge where a high degree of state responses occur together with high levels of non-governmental organization (NGO) mobilization.
Inclusive policymaking, monitoring, and review where gender equality advocates play a crucial role can be seen as a pivotal aspect to ensure gender equality resonant policy outputs on a sustainable basis. Active participation of these groups and their solid standing in policy processes may also secure an implicit gender equality reading of policy outputs by political, policy, and public stakeholders. In addition, as recognized by earlier research on domestic violence, gendering implementation by sustainable coordinated partnerships between state and civil society actors advocating for women's rights ensures that gender indifferent policies are gendered in their implementation.
The dual conceptual framework made up of gendering policy content and gendering policymaking and implementation becomes relevant when the framing of domestic violence is gender indifferent, which is typical for the majority of domestic violence policies across Europe. As it has been pointed out in the context of strategic framing used for the reconciliation of work and family life in the EU (Stratigaki, 2005), cooptation by gender equality contesting frames becomes a problem during implementation. Gendering implementation may then become a durable guarantee against such processes.
Following this conceptual framework, we turn now to an analysis of the gender equality embeddedness of the policy outputs of NGO mobilization against domestic violence in the five countries in CEE in our study. We examine the cases separately and then together.

Gendering Domestic Violence Policies: Five Cases in Central and Eastern Europe
CEE countries joined the processes of transnational mobilization on violence against women in the mid-1990s around the time of the 1995 Fourth World Conference on Women in Beijing. Amid strong NGO mobilization, within a few years many countries of the region moved from minimal or no policy awareness of domestic violence to the adoption of laws and policies specifically targeting it (Krizsan & Popa, 2007). These reforms came about due to concerted efforts of NGOs, who led the mobilization processes, their allies among the transnational women's rights advocates, 4 international organizations (particularly the United Nations and the European Commission), and key supporters in the countries, including among women politicians and the media. While the patterns of mobilization differ in the five countries, the involvement of these key actors is similarly important. The comment by leading advocate Genoveva Tisheva on the adoption of reforms in Bulgaria is applicable to the other cases studies in the analysis. She noted that the changes in the legislation came about due to "the persistent efforts of women NGOs, as well as foreign pressure and the challenge of EU accession" (Tisheva, 2005, para. 5). This section summarizes the development of legislation and policies in each country included in the comparison, placing emphasis on framing processes and the role of gender equality advocates.

Bulgaria
In Bulgaria, governmental and civil society interventions in domestic violence developed after the 1995 Beijing Conference. The National Action Plan for the Implementation of the Beijing Platform for Action (BPfA) adopted in 1996 described the measures to be taken to prevent and eventually eliminate all forms of violence against women. It was also in the aftermath of Beijing (1997) that the Center Nadja Foundation in Sofia started to operate a shelter offering support and protection to women who were escaping abusive partners and to their children. Besides providing them with immediate but temporary alternatives to violence, women's organizations also tried from early on in their interventions to advance legal and policy reforms to secure a comprehensive institutional response to domestic violence. The Bulgarian Gender Research Foundation (BGRF), in cooperation with other civil society organizations and transnational feminist advocacy networks, particularly the Women's Program of the Minnesota Advocates for Human Rights (MAHR, currently Advocates for Human Rights), were forerunners of advocating for the criminalization of domestic violence and protective measures for victims. BGRF and MAHR initiated a process of drafting a law on domestic violence as early as 1997. Between 1997 and 2000, a network of civil society women's advocates developed among organizations and groups that were offering direct services to women in response to domestic violence.
The persistent advocacy efforts of BGRF met favorable political opportunities in 2001, when a critical mass of women (26%) was elected to the Bulgarian Parliament. BGRF and their civil society allies gained several important supporters among the newly elected women Members of Parliament (MPs). With their support, the Law on Protection Against Domestic Violence was finally adopted in 2005. The core group of experts who drafted the law included lawyers from BGRF and other civil society organizations.
BGRF and wider coalitions of NGOs also contributed to the development of a number of implementation documents: the Program for Prevention of and Protection Against Domestic Violence (2007)(2008)  In 2009, the Alliance for Protection Against Domestic Violence, a coalition of 10 NGOs, successfully advocated for further improvements in the legislation, including the introduction of a special line in the budget of the Ministry of Justice for NGO projects contributing to the implementation of the law starting in 2010. With some delay and due to sustained pressure from civil society organizations, the Ministry of Justice allocated funds (BGN 500,000) and opened a call for proposals for NGOs in 2011. 5 Bulgarian policy debates were relatively modest on gender aspects of domestic violence. The Law on Protection Against Domestic Violence uses the rights of individuals frame. The law protects "individuals having suffered from domestic violence" (Article 1). The prognosis includes programs for domestic violence prevention and protection measures of assistance to victims and restraint of perpetrators. The main pillar of the law is the court order for immediate protection, which should be issued within 24 hr, if the court has information about an imminent threat to the life or health of the victim. The rights of individuals frame (with focus on children) also provided the dominant frame for the first implementation document in Bulgaria, the Program for Prevention of and Protection Against Domestic Violence (2007)(2008). The document uses gender-neutral language and talks about "victims" and "children." Although it proposes complex intervention including prevention (awareness raising) and protection (SOS hotline and recovery programs for victims), none is targeted at women, but rather generically at "victims." At the same time, specific services (including a special SOS hotline) are designed for children. Meanwhile, the program does refer to major gender equality documents, particularly the Convention on the Elimination of All Forms of Discrimination Against Women and, specifically, its Article 5 on challenging gender stereotypes, thus giving an implicit gender equality meaning to its genderneutral individual rights framing.
Although the main legislative and policy frame is gender indifferent, the womencentered and gender equality frames are articulated in the implementation of the Law for Protection Against Domestic Violence. A central implementation document, the Handbook With Advice for Persons Who Have Been Victims in Cases of Domestic Violence for Receiving Quick and Effective Protection uses a women-centered frame in which all references to victims are female (e.g., "she" or "her") and all references to perpetrators are male (e.g., "he pushes you"). A structural gender equality frame is articulated in documents that map NGO actions in support of the implementation of the law. 6 With respect to gendering policy content, we found that domestic violence policy development in Bulgaria was gender indifferent, and consequently unfolded without strong processes of contestation of gender equality frames. The family protection and the perpetrators' rights frames only marginally occurred in the parliamentary debates (the first reading of the Bill on Protection Against Domestic Violence, June 2004). But policymaking on domestic violence and subsequent implementation processes were inclusive of gender equality groups, especially from 2001 onwards. Bulgarian gender equality advocates developed a strong sense of ownership over domestic violence reforms, celebrated as major feminist successes. The Director of BGRF, Genoveva Tisheva (2005), described the adoption of the law as "a great victory for Bulgarian women and women's NGOs" (para. 1).

Croatia
Croatia had a strong feminist movement working on gender-based violence, including domestic violence, relatively early (Jalusic & Dedic, 2007). In 1988, the first SOS hotline for women who suffered violence was launched in Zagreb; by 1990, the first women's shelter, Autonomous Women's House, was set up in Zagreb and registered in 1992. Croatian feminists were also part of East-West exchanges with other feminists in Europe by the end of the 1980s. The mobilization of women's NGOs in Croatia was nearly continuous with annual campaigns since 1997 aiming at both awareness raising and policy change. Several NGOs, such as the Autonomous Women's House, Center for War Victims ROSA, and Be Active Be Emancipated (B.a.B.e.), focused on advocating for legislative changes toward adequate legal treatment of domestic violence since 1995. Civil society organizations came together in the strong coalition Women's Network of Croatia, which focused its efforts on creating a specific law on domestic violence from 2000 onwards, when larger democratization processes boosted the mobilization efforts of women's groups.
This mobilization, in concert with international influence from the Beijing process and the process of European integration, resulted in the piecemeal amendment of all relevant laws. Crucial policy changes that mark the development of a state response to  (2005-2007 and 2008-2010, respectively). The main legislative act-the Law on Protection From Domestic Violence-was initiated by a group of female politicians (following the 2000 elections), and it was drafted by a broad governmental working group in close cooperation with feminist NGOs, among them B.a.B.e. After the adoption of the law, feminist NGOs continued to press for more improvements. In 2008, autonomous women's shelters marked an important victory when they were included in public budgets in a ratio of 30% municipality, 30% county, and 30% Ministry of Family. 7 In terms of gendering policy content, the main legislative and policy documents are gender indifferent. The initial agenda-setting voice of feminist NGOs in Croatia used a structural gender equality frame. As early as 1997, a coalition of women's NGOs in Croatia called for passage of a law on violence against women that would address violence as a "form of discrimination that prevents women from exercising their rights and liberties on an equal basis with men." The 24 organizations that signed a public letter on November 25, 1997, also demanded that "women victims of violence be offered protection with the assistance of autonomous, independent and state-funded counseling centers, SOS hotlines, organizations offering shelter, and self-help groups." 8 In contrast to these initial claims, the 2003 Law on Protection Against Domestic Violence largely uses the rights of individuals frame, with some fragmented elements of the family protection frame. Children are the only victims specifically mentioned in the law. The prognosis, besides its reliance on elements of the rights frame, also emphasizes the need for protection of the family, and refers to the correlation of domestic violence with alcohol and drug problems and the need for specific therapy for these cases, but does not make such problems the exclusive causes of violence. Thus, the dominant frame of the law remains the individual rights frame, which makes the text of the law agreeable (if not ideal) for feminist NGOs. These NGOs focus their criticism on the inefficient implementation of the law, rather than its content. The first National Strategy on Protection From Family Violence, 2005-2007 used a women-centered approach. The document mentions women as highly affected by domestic violence and refers to international obligations under CEDAW and the Beijing Platform, but does not link domestic violence to structural gender inequality. Meanwhile, the prognosis uses a gender equality framing that integrates elements of both complex intervention and social transformation with policy actions explicitly targeted toward changing dominant gender norms in society. Implicit gender equality frames and women-centered domestic violence frames are relatively widespread in the Croatian debate, whereas the family protection frame is more strongly articulated in more recent debates (2008)(2009)(2010), particularly in the creation of Family Centers (2006) charged with addressing domestic violence and the implementation of psychosocial treatment programs for (sentenced) perpetrators (introduced in the second National Strategy on Protection from Family Violence, 2008-2010).
Although the dominant frames for domestic violence policies are closer to the gender indifferent end of our continuum, policymaking and implementation processes in Croatia have been inclusive of gender equality groups. Policy changes in Croatia occurred on the initiative of feminist advocates in cooperation with women politicians and other state officials. Alongside continuous NGO mobilization, good relations with women politicians on both the left and the right stand out as exceptional in the Croatian case. This may explain good norm resonance between feminist groups and formal politics, and the absence of strong contestation in adoption of legislative reforms. Women's rights NGOs in Croatia perceived the process of policy and legal development as a success of the women's movement in as far as enactment of the law was The predominant framing of policy outputs is gender indifferent, with the rights of individuals providing the main frame for all adopted policy documents; women victims are never mentioned specifically in any of the Hungarian official documents. The Law on Restraining Orders follows this frame as well, although not consistently. The law has a remarkably weak protection pillar and it also includes some marginal family protection frame elements (the possibility for mediation). The Parliamentary Strategy uses an implicit gender equality approach; its text is entirely gender neutral, but its preamble refers to the CEDAW. Police bylaws and documents use the women-centered domestic violence frame in that they speak about the disproportionate number of women victims. 9 Yet, debates over the gender equality content of domestic violence policies were very polarized in Hungary. Although feminist NGOs used a structural gender equality framing of their claims for state responses to domestic violence, their strategic use of individual rights or women-centered frames was also notable in periods of more intensive cooperation with state actors and the policy process (Safrany, 2003). Contestations to the framing proposed by feminist civil society notably came from the family protection-child protection frame strongly present in Hungarian public debate since the early 2000s but remained outside formal political documents until the recent adoption of the Law on Restraining Orders for Domestic Violence in 2009. The perpetrators' rights frame also was a strong contesting frame in the debates on restraining orders, often voiced by representatives of the Ministry of Justice. In addition, externalizing frames (domestic violence typical for deviant, marginalized groups) were present in different documents related to crime prevention, such as the 2003 Crime Prevention Strategy, but never became dominant frames.
The inclusion of women's rights groups in policymaking and implementation processes is uneven in Hungary. Although they played a crucial role in bringing the issue onto the policy agenda and gained insider status for a short period around the adoption of the 2003 Parliamentary Strategy, thereafter women's groups were pushed to the outside of the process and were gradually replaced in consultation processes by civil society groups representing alternative approaches. The main role in policy processes and implementation of anti-domestic violence policy has since been played by family and child protection experts and groups. Women's groups are only marginally recognized by the state to have a role in this process. The gender indifferent framing of the policy documents was co-opted by implementation processes that place the emphasis on family and child protection.
NGOs in Hungary perceive the domestic violence policies in place as failures in terms of gender equality content. Discontent revolves around a variety of issues, including the lack of adequate involvement of women's NGOs in policymaking and implementation; framing that does not see victims' safety as a central objective, but rather focuses on family protection; and serious problems with the implementability of the law. 10

Poland
Domestic violence was on the Polish policy agenda relatively early as a specific public health issue. The State Agency for Resolving Alcohol Related Problems (PARPA), created in 1992, set the main initial and enduring framing of domestic violence as an alcohol-related problem (externalizing/deviance frame). PARPA's programs and policies, such as Safety in the Family-Preventing Violence in Families With Alcoholic Problems, 1992-2002 were consistent with this frame. Post-Beijing, feminist NGOs in Poland asserted a stronger voice and resolutely advocated for a gender equality and human rights approach to domestic violence. NGOs, such as the Center for Women's Rights, occasionally (depending on the political context) in coalition with the Plenipotentiary for Women and the Parliamentary Group for Women, followed the transnationally diffused feminist practice of Tribunals on Violence Against Women. From 1995 onwards, tribunals were organized biannually to provide testimonies of violence against women, particularly domestic violence, and promote social and policy change. After repeated failed attempts of women's groups to initiate legislation on domestic violence (1998,2001), the draft Law Against Domestic Violence was prepared by the Plenipotentiary with some involvement from NGOs, including women's groups. The Law was adopted in 2005, amid criticism from women's groups about its lack of gender equality explicit content and the weakness of the perpetrator-restraining pillar.
The dominant frame for domestic violence law and policies in Poland is the rights of individuals. It provides the main frame for the 2005 Law Against Domestic Violence and the 2006 National Action Plan Against Domestic Violence. The dominance of the gender indifferent frame is the result of contestations between gender equality explicit and gender equality rejecting frames. The structural gender equality frame was represented by feminist groups, but remained marginal in the official debates, despite a 2001 law proposal prepared by NGOs. 12 However, the implicit gender equality frame and the women-centered approach are largely present in NGO texts within the policy debate and the first also surfaces in the explanation of the Law Against Domestic Violence. 13 The major contesting frame is deviance, which was the initial state frame on domestic violence, for documents such as the Program "Safety in the Family," and it is still dominant in implementation processes. Contestation to gender equality also came from the frames of family protection and perpetrators' rights, prevalent among conservative political forces in the Ministry of Interior and Ministry of Justice and advisory bodies, throughout the entire debate about restraining orders.
Although the law and main policy documents are gender indifferent but not rejecting gender equality approaches, the implementation of the legislative framework is not inclusive of women's groups. A National Action Plan Against Domestic Violence implementing the law was adopted in 2006 giving the core implementation role to the anti-alcoholism institutional network. Women's groups are given limited recognition for their work. In Poland, gender equality groups never really became insiders to the process of policymaking, but maintained the role of outside critics all along. The Plenipotentiary rarely stood up as their state feminist representative within the government (Stetson & Mazur, 1995). Polish feminist NGOs perceive the policy outputs of their mobilization as failures in terms of gender equality. Ursula Nowakovska, one of the leading gender equality activists, argues there was no effort to link work on gender equality with that on ending violence against women. Even in the better draft [of the law] that was sent to Parliament, there was still no reference to gender equality. During discussions, the draft law became weaker and weaker; in the end what was left was a total disaster. (Authors' interview on February 12, 2008)

Romania
The 1995 Beijing Conference provided the occasion for the first significant effort to raise awareness of the high incidence of domestic violence and lack of any policy response in Romania. The first emergency helpline and a small government-funded shelter were established in 1996. Between 1997 and 2000, advocates of legal reforms, including women's NGOs as well as politicians, UN agencies, and development assistance agencies such as the United States Agency for International Development (USAID), focused their efforts on amending the Penal Code of Romania to specifically include acts of violence against family members. In November 2001, a proposal for a special civil law against domestic violence was drafted by the NGO Association for the Promotion of Women in Romania and registered with the Romanian Parliament by a supportive MP. As a result of advocacy efforts of another NGO, the Center for Mediation and Community Security, two other legislative proposals on violence in the family were also registered. The Center Partnership for Equality, a Bucharest-based gender equality NGO took the initiative of brokering a consensus among the different proposals and bringing together a coalition of several NGOs in 2003. Women's groups found themselves in the minority within the coalition that ended up supporting a family protection frame, rather than a women-centered, women's rights perspective in the draft law. The Center Partnership for Equality eventually withdrew from the final stages of drafting the legislation, as only two other organizations in the coalition were supporting a feminist understanding of domestic violence. The structural gender equality frame never became a public frame for understanding the phenomenon.
The Law Against Violence in the Family was adopted in May 2003. It was followed in 2004 by the creation of a National Agency for the Protection of the Family as an implementing body and by the adoption of the National Strategy for Preventing and Combating Violence in the Family and its implementation plan in 2005. Given that the legislation and subsequent policies left significant gaps, particularly in restraining perpetrators, women's organizations continued to mobilize support for legal reforms. The National Coalition of NGOs involved in Addressing Violence against Women, which emerged in the context of advocating for legislative changes in 2003, continued to formulate proposals for legislative amendments, given that the legislative framework was considered inadequate.
The dominant frame for the main legislative and policy documents is the family protection frame, which is used by the Law Against Domestic Violence and the National Strategy. The frame was supported by most state actors, such as governmental representatives and MPs. Feminist civil society actors, in particular the Center Partnership for Equality and supporting MPs (among them, Mona Musca), struggled for an understanding of domestic violence as a violation of women's rights and framed their claims in terms of a women-centered approach to domestic violence. 14 The policy outputs of NGO mobilization in Romania are seen to be running against gender equality. The policy gives the key role in implementation to the family protection institutional network (family assistants). One of the lead women's rights activists in the NGO mobilization period, Roxana Tesiu, judges that they had good momentum in 2002-2003, which they then missed and which has subsequently been nearly impossible to regain. She sees the Law Against Family Violence as a complete failure, in the sense that it does not talk about women as individuals but rather about families, and even families are interpreted in a narrow reproductive sense. In an interview with one of the authors on the process of adopting legislative reforms in Romania, she described the 2003 law as "completely useless." She went on to explain that the law "reproduces a conception of the family that we have inherited from the communist past and which stresses the reproductive function of the family." She sees the law also failing altogether in targeting funds allocated by the state for resolving the problem of domestic violence; in fact, she thinks "nothing has been solved" with the adoption of the 2003 law (Authors' interview with Roxana Tesiu, June 14, 2007). Given the very limited impact in gendering the content of the domestic violence policy, civil society mobilization has continued to focus on demanding inclusiveness in the process.Pressure and efforts were upscaled from 2010 onwards, when new organizations and cohorts of activists joined the mobilization.

Comparing Policy Outputs of NGO Mobilization Against Domestic Violence in Five Central and Eastern European Countries
The five stories are remarkably similar with respect to the centrality of gender equality groups in placing the issue of domestic violence on the policy agendas, as well as the simultaneity of comprehensive policy responses adopted between 2003 and 2005. But how do the policy outputs of the sweeping civil society mobilization compare in the five countries with respect to gender equality?
The story lines showed that the structural gender equality frame for understanding domestic violence proposed by international framework documents has not become dominant in any of the five countries. Nevertheless, this was the starting point for feminist NGOs across CEE. On returning from international events such as the Beijing Conference, or as a result of interaction with transnational feminist groups, "local" feminists brought international norms on domestic violence framed in terms of structural gender equality to their countries. However, in the process of bringing their agenda to fruition, feminist NGOs often chose frames that were not explicit on the structural gender inequality roots of domestic violence.
The women-centered frame, for example, is fairly consistently used in statements entered directly into the policy process by NGOs in Bulgaria, Hungary, and Romania. In Hungary, feminist NGOs have used the women-centered and implicit gender equality frames during their short-lived periods of intense involvement in the policy process. Polish and Croatian NGOs used more explicit gender equality language all along, the Polish movement being the only one from among the five that explicitly challenges the absence of the explicit gender perspective from the adopted domestic violence law ( Table 2).
The resulting policy outputs are ambivalent. The first element we identified as important for assessing domestic violence policies from the point of view of gender equality is the content of the laws and policy texts. Laws on domestic violence in four out of the five countries use a gender indifferent frame-rights of individuals subjected to domestic violence-which remains individualist and largely silent about the social structural roots of the problem and possible responses, and even about connections to gender equality issues. In Romania, the contesting family protection frame is used throughout.
Although gender inequality aspects seem to remain entirely hidden in the laws, more gender equality explicit approaches occur in documents closer to implementation in some countries. The Croatian and the Bulgarian implementation strategies, as well as the 2003 Hungarian parliamentary strategy, link domestic violence interventions implicitly to gender equality by making references to international framework documents. The Croatian strategy also makes reference to the gender equality strategy in place in the country. They all use the implicit gender equality frame. The Polish implementation strategy does not add the gender component to the law, but remains consistent with the individual rights frame of the law. The Romanian strategy does the same: It remains consistent to the law by using the same contesting family protection frame. The patterns of framing implementation are further supported by the involvement and empowerment of independent women's rights advocates in the implementation process.
This brings us to the second element in assessing gender equality resonance of domestic violence policies: the inclusion of independent women's rights advocates in policymaking and policy implementation. While in Croatia and Bulgaria women's groups are explicitly (Croatia) or implicitly (Bulgaria) seen to be the main actors in the implementation, and accordingly financed by the state, in the other three countries the representatives of contesting voices are given key roles in implementation. In Hungary, these are the family and child protection agencies protecting only women with children; in Poland, the anti-alcoholism institutional network; whereas in Romania, it is the family protection network (family assistants; central and local governmental agencies and civil servants).
The resulting policy outputs thus have varying levels of transformative gender equality content, and therefore meet the approval of feminist NGOs differently in each country. Only Romania adopts a gender equality contesting framework for domestic violence, and no country adopts a framework explicitly treating domestic violence as a gender inequality problem. In Bulgaria and Croatia, outputs are celebrated as feminist successes that bring transformative outcomes for women, while in the other three countries, key feminist advocates see the policies as failures in capturing the main points of the feminist agenda. These different evaluations of similar outputs in terms of policy content are explained by the different gendering policymaking and implementation processes. Croatia and Bulgaria have implementation documents that link domestic violence implicitly to gender equality policy. In addition, both of them also open a window of opportunity to empower feminist groups to participate in the policymaking process, and also empower them to keep ownership of the implementation process. Women and feminist NGOs gain standing in the debates and authority in the implementation process, which accounts for key elements of empowerment in Ferree's theory of the gender of governance (Ferree & Gamson, 2003). But in Hungary and Poland, implementation documents and mechanisms open the way for co-optation of the seemingly gender-neutral rights agenda of their laws. Poland has an implementation strategy framed in individual rights terms without even implicit references to gender equality, whereas Hungary in effect has no up-to-date implementation plan. In addition, in both countries feminist NGOs are displaced from policymaking and implementation by contesting actors.
Thus, it is only in Bulgaria and Croatia that the ownership of women's groups over domestic violence policy processes gives a gender equality and women's rights discursive position to domestic violence, while in the other two countries with predominantly gender indifferent policy frames this discursive position is shaped by alternative approaches. For these reasons, it is only the Croatian and Bulgarian policy outputs that are seen as gender transformative in their content and celebrated as feminist successes, while the other three cases are seen as lost battles in terms of bringing genuine advancement to the gender equality agenda.

The Potential and Limits of International Norms Adaptation
International norms on violence against women are translated to national arenas in framings that are far less gender equality explicit than proposed by international documents. In line with earlier global (Merry, 2006a) and European Union (European Commission, 2010) research, we found that a gender-neutral frame centered on the rights of all victims of domestic violence is the dominant frame in most of the countries analyzed, with only implicit gender equality elements included. Although women's movements in CEE have been highly successful in drawing attention, setting agendas, influencing discursive positions (Keck & Sikkink, 1998) on domestic violence, and moving the debate from a total taboo in the early 1990s to being a normal part of policy discussions by the early 2000s, the policy outputs of their mobilization are often distant from feminist movement claims. The gender-transformative content of the international human rights norms that emphasize the rootedness of domestic violence in historically unequal power relations between women and men may be lost in the process of adapting these norms to national contexts.
Strategically or incidentally seeking resonance with mainstream frameworks, reformers, including feminist NGOs and experts, stay away from the explicit structural gender equality approach and from the claim that women are the main group affected by domestic violence to be more inclusive of other groups at risk, such as children, elderly, disabled, and ethnic minorities. Individual rights frames resonate well with mainstream policy agendas (Keck & Sikkink, 1998), which are shaped in this context by the core issues of post-communist democratizing societies, such as protecting human dignity, autonomy, life, and bodily integrity, especially for the vulnerable, or children's rights. Meanwhile, the human rights strategy as translated into legal texts loses women as a group and the structural gender equality elements, focusing instead on genderless individuals (Keck & Sikkink, 1998). But can this strategy serve gender equality goals? Or, is co-optation inevitable as has been the case of strategic framing used for reconciliation of work and family life in the EU (Stratigaki, 2005)? In domestic violence, as the analysis shows, degendering does not necessarily bring co-optation. The key to "saving" the gender equality content, illustrated by the better faring cases of Bulgaria and Croatia in our sample, is the empowerment of gender equality voices to assume ownership of policy development as well as policy implementation processes.
Whether the gender-neutral framing realizes the potential of bringing in transformative gender equality content or it becomes a path to co-optation by gender equality contesting policy objectives depends on factors that fall outside the framing of laws and the content of policies. The continued ownership and standing of civil society actors using gendered frames in the policy process influences the reading of policies framed in gender indifferent ways. Gendered implementation, understood as both gender equality resonant framing of implementation documents and empowerment of feminist autonomous groups to take part in the implementation process, may channel transformative gender equality content back into the process of implementing gender indifferent laws and policies. Meanwhile, it is important to note the vulnerability and instability of such solutions. Sustaining gender equality measures in the implementation stage of domestic violence policies, including by providing funding to independent women's rights advocates, may be much more vulnerable to budget cuts and political cycles than the legislative entrenchment of gender transformation. 15 Coming back to the dual-track approach for assessing gender equality content suggested in our conceptual framework, our analysis shows that the cases that meet better gender-transformative standards are the ones that, along with good performance in governance of gender, also perform better in terms of the gender of governance (Ferree & Gamson, 2003). In Croatia and Bulgaria, the two elements work together: policy frameworks on privacy and classic human rights are challenged and transformed in gender equality sensitive ways and NGOs are empowered to be the guarantors of its implementation and improvement. Meanwhile, and quite importantly, in the cases of Hungary and Poland, in the absence of insider NGO participation in policymaking and implementation, we see the vulnerability of the gender indifferently framed domestic violence agenda to co-optation and the risk of losing women victims from view.
The translation of norms for addressing violence against women from the international to the national level brings to countries of Central and Eastern Europe a concept of gender equality that is stretched (Lombardo et al., 2009) to be more inclusive and to meet strategic interests that make it easier to "sell." The concept certainly has the advantage of leading to the adoption of gender equality resonant laws and policies, but where support and ownership by gender equality actors is absent, it runs the risk of co-optation by frames that contest the tenets of gender equality. The focus on a combined assessment of the legislative and policy outputs in terms of their transformative gender equality content, which uses the standards of empowerment just as much as displacement (Verloo, 2005a), autonomy just as much as authority (Ferree & Gamson, 2003), becomes imperative in this context.