Polycentricity and framing battles in the creation of regional norms on violence against women

Abstract

In Latin America and Southern Africa, norms on violence against women have developed with ups and downs, not simply in reaction to global norms, but sometimes even preceding global norm diffusion or surpassing it in terms of scope, framing and binding character. The classic global-to-local account with a single source of norm creation cannot capture these dynamics. Including the regional level in a dynamic model of norm diffusion enables us to understand the changing contents of a norm and to acknowledge transregional agency. We show (1) how norm contestation is an ongoing, multidirectional and polycentric process; (2) how the regional level opens up opportunities for feminists and femocrats; and (3) under which conditions regional norms can be both more progressive than global ones and more adapted to regional needs, and, in turn, are thus able to strengthen the ‘global’ norm.

Introduction

Violence against women (VAW) is recognized as a pervasive and intractable problem worldwide. These days it is even called a ‘shadow pandemic’, as the lockdown measures imposed to fight the spread of COVID-19 have led to an alarming increase of (emergency calls of) VAW. ¹

In June 1993, at the UN International Conference on Human Rights in Vienna, VAW was defined as a violation of human rights for the very first time. Governments agreed that action should be taken, and that same December, the UN General Assembly adopted the Declaration on the Elimination of Violence against Women. Although the Declaration was not binding, its formal recognition of the problem was a major victory for the women’s movement and sped up the norm’s spread around the world. ² It also generated academic interest in how international norms on gender equality develop. ³ This body of research provides us with important insights into international norm diffusion and into the role of transnational feminist advocacy in creating such norms and monitoring implementation. ⁴ Yet, many of these studies implicitly assume that norms are created at a single moment and spread from a single source. Evolving simultaneously with the global norm, however, Latin American and Southern African norms on VAW in different ways surpassed this global norm in terms of scope, framing and binding character. Classic global-to-local accounts therefore cannot adequately capture the regional dynamics of norm (re)creation.

In the Organization of American States (OAS), for instance, the Inter-American Commission of Women (known by its Spanish acronym, CIM, Comisión Interamericana de Mujeres) ⁵ had already taken up the issue of VAW in 1989 in response to the strong mobilization of women’s organizations, particularly in Latin America. ⁶ On 6 June 1994, the OAS General Assembly adopted the Inter-American Convention on the Prevention, Punishment, and Eradication of Violence Against Women (OAS Convention, also called Belém do Pará Convention) in Belém do Pará, Brazil. ⁷ It established the member states’ obligation to prevent, investigate and punish VAW, regardless of whether it took place in the home, community or public sphere. It was the world’s first legally binding treaty which exclusively focused on eliminating VAW. In Southern Africa, women’s movements demanded that the Southern African Development Community (SADC), a newly founded regional organization, take action on VAW in 1995. Initially, this led to the adoption of a non-binding instrument, but in August 2008, the SADC adopted the Protocol on Gender and Development. A Protocol is the most binding instrument available to the SADC, and this one not only reiterated women’s rights as they were laid down in the UN Declaration, it also responded to context-specific issues relevant to the Southern African region. ⁸ In addition, it established institutional mechanisms to monitor implementation and set targets and a deadline. Southern Africa became the world’s first region where all existing continental and international instruments were united in a single, legally binding Protocol. ⁹

The development of these two strong regional norms on VAW thus raises two questions which cannot be answered by global-to-local norm-diffusion models. First, how are regional norms (re)created in multilevel processes? And second, under which conditions does a strong (feminist) regional norm on VAW develop? To answer these, we propose a dynamic model of norm diffusion which includes the regional level, enabling us to understand what happens to the norms’ content and properly recognize the agency at that level. Our model shows (1) how norm contestation is an ongoing, multidirectional and polycentric process; (2) how the regional level opens up opportunities for regional advocacy networks of feminists and femocrats and strengthens the agency of such regionally operating actors; and (3) under which conditions regional norms can both be more progressive than global norms and more adapted to regional needs, and, in turn, are thus able to strengthen the ‘global’ norm.

The outline of the article is as follows. It starts with a brief overview of the major discussions on norm diffusion and the development of our conceptual framework. Next, the methodology section briefly presents critical frame analysis as the preferred method to unpack a norm at different stages of its trajectory and gauge its ‘quality’. In the empirical sections, we assess our theoretical arguments in two steps. First, we trace the fluctuating ‘quality’ of the norms on VAW being articulated in Latin America and Southern Africa. Second, we identify the regional landscapes’ drivers of, and opportunities for, norm diffusion. In the discussion, we compare the findings for Latin America and Southern Africa. We conclude that the story of these VAW norms shows the added value of including the regional level as a particular site of norm production in norm-dynamics models.

Theorizing norm diffusion as polycentric and contentious

Since the end of the 1990s, the diffusion of international norms has been a key issue in international relations. Norm diffusion was theorized as a life cycle or a spiral model, showing how a norm would emerge and then spread from the global to the so-called local level. ¹⁰ These rationalist approaches defined diffusion as a ‘causal process in which a diffusion mechanism transmits a diffusion item from a point of origin to a point of adoption’ and expected a certain degree of convergence to take place. ¹¹ Convergence was also at the basis of John Meyer’s idea of a world society in which ‘world-cultural principles’ and ‘world-approved policies’ result in the structural isomorphism of nation states. ¹² These rather linear models proved controversial, and reviewing five different strands of criticism, we discuss what we take from them and where we add something new.

A first strand of criticism was spearheaded by Amitav Acharya, who challenged the black-box treatment of the adoption process. ¹³ Taking issue with the assumption that global norms fall on bare ground, he argued that the diffusion of a new ‘global’ norm interacts with prevailing ‘local’ norms, political institutions and economic structures. Next, he formulated conditions under which norms would be adopted, rejected or ‘localized’, bringing in the role of ‘local’ agents. Adoption thus came to be seen as a ‘dynamic process in which the medium or the context at the point of adoption shapes the way in which external practices are received and implemented’. ¹⁴

The notion of localization was important in recognizing the interpretation processes surrounding norm diffusion, but it did not address a second limitation of the so-called broadcasting model of norm diffusion, namely its reliance on a single point of origin. ¹⁵ We depart from this view by using two proposed concepts; multidirectionality, meaning that norms not only travel top-down, and polycentricity, indicating how norm diffusion may have multiple simultaneous sources. ¹⁶ These concepts enable us to dislodge the implicit assumption that ‘liberal’ norms only emerge in Western democracies. ¹⁷ Whereas prevailing models considered norm-diffusion processes to result in a fairly stable, albeit localized (or vernacularized or ‘stretched and bent’) norm, ¹⁸ this static view was criticized by a third strand of critics. Norms came to be seen as ongoing definition and renegotiation processes that can be reversed at any time. ¹⁹

A related and fourth strand of criticism further highlighted the importance of contestation. Initially, norm diffusion was conceptualized as a linear, progressive development, and contestation only appeared as a ‘local’ practice which developed once a global norm ‘hit the ground’ – at least in the literature on (non-)compliance. ²⁰ Yet contestation acquired a different standing when constructivists started unpacking norms as fluid concepts. Now, contestation no longer was understood as an instance of non-compliance, nor as a temporary phase during which a new, localized understanding developed, but as the practice through which norms are constituted. ²¹ In this way, contestation becomes a mode of discursive framing in the sense that ‘frames shape how people understand the world and, based on this understanding, what they perceive to be appropriate action’. ²² This resonates with our view that a norm’s content is not a fixed standard, but the contested and contingent outcome of social interaction which can only ever be partially fixed. ²³ In our view, as norms are produced in context, they have a dual quality because they both structure social practices and derive their meaning from them. Contestation may thus weaken but also strengthen a norm, its content, as well as its application. ²⁴

We argue that contestation by regional actors impacts the previously existing national norms as well as the emerging global one, potentially changing both in the process. It does so through what we call framing battles in which regional non-state actors actively use the new terrain of the regional level to contest existing national norms and the global norm simultaneously. Antje Wiener nicely distinguishes between norm contestation as a social practice of merely objecting to norms (‘reactive contestation’), as it is commonly used in compliance literature, and contestation as a critical engagement with the norm (‘proactive contestation’). ²⁵ The discursive processes of proactive contestation create a norm which is appropriate in a certain context and therefore potentially obtains a certain legitimacy. In turn, a norm can only be legitimate when produced by practices which allow for contestation during each stage of the process of norm constitution, from emergence to implementation to revision and beyond. Contestation is assumed to be more intense in more diverse societies, where a great deal of interaction is necessary to reach a legitimate compromise. Whereas global society has a high degree of diversity when it comes to gender norms, the regional level is expected to show less diversity, so a contextualized compromise might be reached with relatively more ease. Because of the geographical coherence and identity, the regional level is perceived to be ‘more attuned to local constraints’ than the global level. ²⁶ Yet, regional norm dynamics inevitably entail contestation as a productive practice because of the relatively open character of the regional landscape. It is an instance of what Wiener calls the broadening of a governance process beyond the usual interpretative community. ²⁷

We want to emphasize that adopting gender norms, and particularly norms against gender-based violence, entails a change of regime. ²⁸ After all, attempts to end VAW typically involve profound normative change because they directly engage with ideas about what is public and private, what is acceptable and unacceptable behaviour and how rules and procedures need to be altered to protect women, and men, from gender-based violence. Such regime change is likely to meet with resistance, or reactive contestation, in Wiener’s terms. Consequently, it requires dealing with power relations and processes, resistance and resilience, not only in norms’ emergence but also in their diffusion. ²⁹ The acknowledgement that norm-diffusion processes entail reactive and proactive contestation calls for more explicit attention to the role actors play in such processes. Who is doing the contestation?

To link this back to the notion of proactive contestation, actors engage with norm dynamics by shaping the legitimacy of these norms. When we understand contestation as a normal and even necessary procedure for a norm to become considered appropriate and legitimate, it becomes critical to further explore actor’s access to the spaces through which norms are diffused. Which actors have access to these spaces and processes depends on the governance landscape in which norm diffusion takes place. That landscape first influences norm diffusion through the extent to which gender equality norms match, or face a mismatch with, the logic of regional governance. ³⁰ The governance landscape’s second effect pertains to which kinds of actors are included in norm-diffusion and policy-making processes. The logics underlying the regional landscape play a role in the kind of engagement that regional organizations have with civil society actors. In this regard, SADC’s roots in anti-colonial liberation struggles have translated into a relatively open organization in which civil society actors can seek space to engage with regional politics, while the strong human rights mechanisms within the OAS explain why the regional advocacy network is dominated by legal experts. ³¹ Regional governance has emerged as a new opportunity structure for actors to advocate for new norms in a space with less entrenched positions than at the national or global levels. It is a distinct level with its own dynamics, offering access to new, transnational societal actors. ³² Feminist activists have been successful in establishing regional gender equality machineries, albeit small and understaffed ones. Femocrats within these machineries maintain strong ties with women’s organizations in the region, draw inspiration to put new issues on the agenda and, in return, provide support in terms of funding, expertise and access. An important characteristic of this cooperation between regional activists and femocrats is the generation and exchange of expert context- and issue-specific knowledge, through research and workshops, as a key instrument in the gendering of regional governance. ³³

A fifth and last point remains to be addressed. As normative change clearly is not a one-shot game but an ongoing process or ‘life cycle’, the dynamics and roles involved cannot be assessed and interpreted at one point in time. In empirical analysis, the norm-diffusion journey tends to be reduced to a single phase – mostly that of adoption or implementation – during which localization and resistance take place. In order to get away from the linear perspective on norm dynamics, Lisbeth Zimmermann proposes a useful distinction between different steps. ³⁴ Each implies a new political process involving different actors and different structural constraints and opportunities, leading to the norm’s contestation and renegotiation. ³⁵ The three stages that she distinguishes in the translation mechanism between a norm’s emergence and its adoption are reinterpreting a global norm during the translation into domestic discourse, reshaping it during the translation into law and reshaping during implementation. ³⁶ Zimmermann’s perspective is relevant for our approach to norm dynamics because it does not depict norm diffusion as a simple transfer from origin to end point and, hence, can be extended to a multidirectional model. We do not follow her, however, in identifying a point of full adoption or implementation. ³⁷ Instead, we take the view that contestation processes never come to a full stop as they are constitutive of the norm, shaping its meaning in an everlasting process of interactions between actors at different governance levels. Zimmerman’s different norm-diffusion steps are also of value to our argument because they allow studying more specifically which actors are involved and in what way. Each next step and level in norm diffusion implies a new political process involving different actors and different structural constraints and opportunities in contesting and renegotiating the norm. Also, each step and level feature different institutional settings and actors. The many steps and directions of the norm-diffusion journey challenge the assumption of a sequential and hierarchical relation between international, regional and national norm-diffusion processes. This not only means that norm creation is geographically and temporally dispersed and may have multiple points of origin, but also that norms are diffused in multiple directions. Instead of taking international norms as the starting point, the important empirical questions are: Where do norms arise? Where do they go? And how does their content change from one place to the next?

Method

In order to scrutinize how a norm’s content continues to evolve, we use insights from critical frame analysis (CFA). As an analytical method distinct from the theoretical concepts of strategic framing and discursive framing (see Note 21), CFA is an interpretative method which understands and examines policy problems as constructions based on competing interpretations of those problems and the desired policy solutions. In mapping the different representations of VAW, CFA employs a deconstructionist approach, treating VAW as an empty signifier and studying it as an open concept that in practice is filled with a multitude of meanings. This makes CFA particularly useful to gauge the ‘quality’ of a norm. It can show whether a norm becomes more progressive – a transformative norm requiring a profound change to the existing normative framework – or whether it is (re-)shrunk by limiting its scope or weakening its mandatory character. The method entails the analysis of policy documents, asking a set of sensitizing questions that aim to uncover each document’s implicit and explicit interpretations of problems, solutions and roles. ³⁸ CFA enables us to compare the norm on VAW at different moments and in different regional contexts. We trace the ‘quality’ of the norm by asking the following: Is VAW explicitly attributed to unequal power relations between women and men? Are the sexes of VAW victim and perpetrator specified, or is VAW represented as problem which is not sex-specific? And finally, are the role of the state in the perpetuation of that violence and its responsibility for eradicating VAW explicitly recognized? ³⁹ When the questions are all affirmative, we consider the frame to be a transformative or feminist one. In order to understand why the quality of a norm changes and under which conditions it is strengthened or weakened, we combine CFA with qualitative process tracing. To that effect, we identify who were involved in proactive contestation (norm entrepreneurs). ⁴⁰ In terms of data, our study relies on key policy documents from regional organizations and statements from feminist groups.

When it comes to our case selection, we consider the norm on VAW as a typical case. ⁴¹ As a global problem, it has a strong mobilizing force, as evidenced by the #MeToo campaigns which went viral in 2017. To empirically analyse norm dynamics, we selected Latin America and Southern Africa, two regions where a strong regional norm on VAW developed. We assess our theoretical arguments by first tracing the fluctuating quality of the norms on VAW in these regions, and then identifying the regional landscapes’ drivers of, and opportunities for, norm diffusion.

Framing battles and the fluctuating quality of norms

The first global norm on VAW was enshrined in the non-binding UN Declaration on the Elimination of Violence against Women. Its preamble combines an instrumental approach (arguing that VAW should be eliminated because it ‘is an obstacle to the achievement of equality, development and peace’) and a human rights approach (considering VAW as a violation of women’s human rights). ⁴² It defines VAW as the manifestation of the ‘historically unequal power relations between women and men, which have led to domination over and discrimination against women by men’. Article 2 of the Declaration offers a comprehensive definition of VAW, covering public and private life, and including violence ‘perpetrated or condoned by the State’. Against the background of this transformative global norm, the next section traces the quality of two regional norms on VAW.

The regional norm: OAS

Regional feminist networking was central to the development of a specifically regional framing of the problem. In the 1980s, Latin American feminists began to analyse the regional pattern of military dictatorship and authoritarianism in the public sphere and its relation to the prevalence of authoritarian and patriarchal relations in society and the family. From here, they developed a comprehensive framing that included physical, sexual and psychological harm to women, as well as more indirect acts of humiliation and discrimination keeping women in subordinate roles. Violence, they argued, not only constituted a violation of women’s human rights, but also a major obstacle to being free and equal citizens in the newly developing democracies. ⁴³ This understanding of VAW gained resonance in global conversations on VAW, with Latin American feminists playing a pioneering role. ⁴⁴ Simultaneously, this understanding entered the deliberations about a OAS Convention.

In order to prepare the Convention, the CIM engaged in an open process that included consulting a wide range of regional civil society organizations and experts. ⁴⁵ Taking these into account, the Convention adopted a broad definition defining VAW as a violation of their human rights and fundamental freedoms. ⁴⁶ VAW is seen as a ‘manifestation of the historically unequal power relations between women and men’, ⁴⁷ and the subordinate position of women in society as constituting part of the problem. Article 6 of Chapter II states that

the right of every woman to be free from violence includes, among others: a. The right of women to be free from all forms of discrimination; and b. The right of women to be valued and educated free of stereotyped patterns of behavior and social and cultural practices based on concepts of inferiority or subordination.

Women’s organizations had originally advocated for an even more comprehensive definition including ‘indirect acts that intimidated or humiliated women, maintained them in sex-stereotyped roles, or denied them human dignity, whether or not these acts caused physical or mental injuries’. ⁴⁸ Although this was not to be, the Convention still resulted in a ‘path-breaking international human rights norm’ that placed the responsibility for defending and protecting the rights of women on the state. ⁴⁹ In fact, the Convention defined VAW as ‘any act or conduct, based on gender, which causes death or physical, sexual or psychological harm or suffering to women, whether in the public or the private sphere’ (Convention, 1994: preamble). This is one of the Convention’s most remarkable features, inspired by the dictatorial past of the region: it radically eliminated the classic private/public distinction that relegated violence to the private sphere, beyond the concern of the state. ⁵⁰

The establishment and ratification of such a progressive Convention was astonishingly fast and smooth, with most member states quickly signing and ratifying it. Of the 35 parties, 24 had signed it by 1995, with 15 also having ratified it. ⁵¹ This relative ease is surprising considering the relatively strong monitoring mechanisms attached to the Convention. It contains binding legislation and is embedded in the existing Inter-American legal framework and enforcement mechanisms. Any individual or group can appeal to the Inter-American Commission on Human Rights (IACHR) when they feel the state has neglected its duty to protect women. The Inter-American Commission of Women may also request advisory opinions on interpreting the Convention from the Inter-American Court of Human Rights.

Yet national laws established after ratification only partially complied with the Convention. Only four countries (Ecuador, Guatemala, Honduras, and Venezuela) adopted the full concept of VAW in their national legislation. All others reframed certain aspects or left them out. First of all, the feminist framing of the problem was systematically ignored. Most national legislation chose a limited focus on the domestic realm and family violence, and did not address violence in the sphere of the community or the state. ⁵² The expert committee reviewing the implementation process in 2008 criticized the persistent use of the terms ‘intra-family violence’ and ‘domestic violence’ in national legislation, whereas the Convention

considers that violence against women includes not only what happens within the scope of the family, domestic unit, or interpersonal relationships, but also sexual violence, torture, trafficking, and what is tolerated or allowed by the State or its agents. ⁵³

Second, intra-family violence was seldom criminalized. Most countries prosecuted perpetrators in civil or family courts that often proposed conciliation or mediation between the victim and her aggressor, or even a pardon for the aggressor if he agreed to marry the victim. ⁵⁴ Third, the responsibility for providing emergency services (such as shelters, hotlines and psychological and legal assistance) to victims of violence was largely left to private actors and the volunteer sector. CIM’s 2002 evaluation concluded that ‘the spirit of the Convention, which is to protect women’s human rights, has been altered’ because national legislation continues to ‘give priority to the family unit and not to women’s human rights; women’s rights are falsely contrasted with the rights of children, adolescents, the disabled, senior citizens, and men; and other forms of violence against women outside the family have been ignored’. ⁵⁵ As a result, the feminist framing was lost.

In a next stage, reflecting Zimmermann’s notion of reshaping, the norm was partially strengthened again. A stronger monitoring mechanism introduced in 2004 enabled women’s organizations to draft shadow reports and closely monitor implementation (more on that in the next section). In 2006, pushed by litigation before the IACHR, Brazil adopted a new law on VAW. Although it did not address violence by the state and only focused on violence in intimate relationships, it expanded the Convention by redefining the scope of family to include extended families, cohabiting partners and same-sex relations. ⁵⁶ In 2012, CIM’s second hemispheric implementation report observed a ‘positive trend toward mainstreaming the definition of violence against women articulated in the Belém do Pará Convention into domestic law’. ⁵⁷ On a more critical note, it also stated that many states still failed to address violence in the public sphere and violence perpetrated by the state. ⁵⁸

The regional norm: SADC

In Southern Africa, the 1993 UN Declaration on the Elimination of Violence against Women and the preparations for the Fourth World Conference on Women in Beijing coincided with the rebirth of SADC. Despite pressure from women’s organizations, the 1992 SADC treaty did not mention VAW. As a follow-up to Beijing, SADC adopted the Declaration on Gender and Development in 1997. It called for urgent measures to be taken ‘to prevent and deal with the increasing level of violence against women and children’ (Section H). ⁵⁹ GenderLinks, the regional advocacy network, protested against the assimilation of VAW and violence against children, arguing it risked downplaying the gendered power aspects in the problem’s diagnosis and prognosis: ‘they are not the same thing at all. Women are adults with agency. Children are minors that need to be protected’. ⁶⁰ The norm was strengthened in the 1998 Addendum to the Declaration, which introduced a feminist framing identifying the unequal power relations VAW implied. ⁶¹ It reiterated and even exceeded the global norm by proposing measures to tackle systemic gender bias through the gender sensitization and training of ‘judicial officers, prosecutors, police, prison, welfare and health officials’ (Art. 20). Its regional contextualization is visible in its reference to ‘promoting the eradication of elements in traditional norms and religious beliefs, practices and stereotypes which legitimize and exacerbate the persistence and tolerance of violence against women and children’ (Art. 13). The Addendum also adopted a feminist frame in its explicit references to the ‘empowerment’ of women who are ‘survivors’, not victims, of VAW (Art. 19 and 25). Ten years later, the SADC norm further evolved into the Gender Protocol. To understand its significance, it is useful to look at the 2003 Maputo Protocol of the African Union (AU) and the SADC Audit of 2005.

In 2003, the AU elaborated the human rights of women in the Protocol to the African Charter on Human and People’s Rights on the Rights of Women in Africa, better known as the Maputo Protocol. ⁶² It articulated a strong, encompassing norm of VAW as ‘all acts perpetrated against women which cause or could cause them physical, sexual, psychological, and economic harm [. . .] in private or public life in peace time and during situations of armed conflicts or of war’ (Art. 1j). It commits states to eliminate all practices ‘based on the idea of the inferiority or the superiority of either of the sexes, or on stereotyped roles for women and men’ (Art. 2.2) and to ‘enact and enforce laws to prohibit all forms of violence against women [. . .] whether the violence takes place in private or public’ (Article 4.2). Reflecting the high prevalence of violent intrastate conflicts in the region, it calls for ‘peace education [. . .] in order to eradicate elements in traditional and cultural beliefs, practices and stereotypes which legitimise and exacerbate the persistence and tolerance of violence against women’ (Article 4.2). Other aspects where the African norm deviates from the global one can be found in its attention to trafficking in women and its prohibition of ‘medical or scientific experiments on women without their informed consent’ (Article 4.2), as African women had been subjected to unsafe clinical trials by foreign pharmaceutical companies. ⁶³ Compared with the 1998 SADC Addendum, the Maputo Protocol is less feminist in that it refers to women as victims and does not contain the concept of empowerment; yet by separating VAW from violence against children, it is also more feminist than the Addendum.

Shortly after the Protocol’s adoption, a SADC Audit took place. This 2005 Audit included new proposals such as ‘rehabilitation or counselling programmes for perpetrators of violence’ (p. 7) and introduced new region-specific concerns by arguing that VAW was ‘one of the leading factors for HIV infection’ and asking for post-exposure prophylaxis (PEP) ‘to reduce the likelihood of HIV infection after a sexual assault’ (p. 2). It problematized marital rape and argued that married women suffer by being held to double standards:

Abolishing marriage as a defence for sexual offenses is an important barometer of the level of gender justice discourse in a country, and is a critical factor in the era of HIV/AIDS, where one of the largest categories of those newly infected is married women who are faithful in their relationships while their husbands are not. (p. 2)

Following the Maputo Protocol, the Audit also noted the growth in sex trafficking and the absence of a regional approach to tackle this issue.

The Audit’s call for a binding instrument resulted in the SADC Gender Protocol, adopted in 2008. The ‘regionalized’ frames of the Addendum, Maputo Protocol and Audit found their way into this, making it a more progressive document than the 1997 Declaration. The Gender Protocol featured specific targets on VAW: enacting and enforcing legislation that prohibits all forms of gender-based violence (GBV); ensuring that this legislation provides specific measures directed at survivors and perpetrators of sexual assault; and adopting integrated approaches to reduce the incidence of GBV. ⁶⁴ A positive achievement in the negotiations was the change of language from ‘victims’ to ‘survivors’, and from ‘protecting women’ to ‘the empowerment of women’ (Article 3). The Protocol specified what action should be taken on human trafficking and also called for the provision of ‘comprehensive testing, treatment and care of survivors’ (Article 20). Nevertheless, if we compare the final version of the 2008 Protocol with its draft, some weakening becomes apparent. Replacing the term ‘violence against women’ with ‘gender-based violence’ (Art. 1.2) constituted a major shift. While GBV clearly reflects the gendered nature of VAW and is more inclusive in terms of acknowledging violence experienced by men and sexual minorities, the term ‘violence against women’ was preferred by proponents to underline that almost all GBV is committed against women and better reflects how gender inequality affects women. ⁶⁵ The replacement of ‘marital rape’ with ‘domestic violence and rape’ (Art. 20.1) also was a clear loss. Several member states refused to include marital rape, arguing it did not exist in their country. The measures addressing traditional norms were also weakened, as they should now be ‘discouraged’ instead of ‘eradicated’. Moreover, though ‘religious beliefs’ and ‘polygamy’ were mentioned in the draft, they were omitted from the final version. Weakening manifested itself most clearly in the overall goal: the draft protocol stated that GBV was to be eliminated by 2015, while the final version only aimed to reduce it by half. Overall, the language of the final text was less binding than feminists and femocrats had hoped during the drafting of the protocol. ⁶⁶ That said, the Protocol was a huge step forward in terms of implementation. It contained six targets, including the exhortation for states to review their criminal laws to ‘ensure that justice and fairness are accorded to survivors of GBV in a manner that ensures dignity, protection and respect’ (Article 20.3) by 2015. In contrast, at the global level, the UN Millennium Development Goals (MDGs) from 2000 had not a single target on VAW. ⁶⁷

The SADC Gender Protocol was further revised and updated in the Post-2015 Protocol. Although it largely reproduced the 2008 version (and did not mention marital rape), the section on GBV was extended by a clause which called for member states to ‘develop strategies to prevent and eliminate all harmful social and cultural practices, such as child marriage, forced marriage, teenage pregnancies, slavery and female genital mutilation’ (Art. 20.1). This extension was significant because it was proactive and sought to address social norms, ‘the bedrock upon which harmful practices such as GBV are rooted’. ⁶⁸ A huge step forward was the move from ‘halving’ to ‘eliminating’ GBV (Art. 25).

Drivers and opportunities

After tracing the fluctuating quality of the regional norms on VAW, we identified the drivers of regional norm diffusion and the opportunities created by the regional landscapes. In this section, we elaborate how women’s movements and femocrats were the key promoters of regional norms on VAW during different phases of the norms’ shaping and translation.

Feminists and femocrats in the context of OAS

Latin American feminists already brought up the issue of violence in the 1970s, while under military rule and in situations of armed conflict. However, the context of closed states and nascent, fragile democracies offered them only limited opportunities to set the political agenda. Through regional networking, women’s organizations started to seek access to international organizations open to women’s rights issues in the 1970s. The first UN conference on women that took place in Mexico City in July 1975 and the UN Decade for Women (1975–1985) thus became important arenas for regional activists. Latin American feminists played a very active, pioneering role in global conversations on VAW. ⁶⁹ In response to this strong regional mobilization, femocrats at the CIM took up the issue and declared VAW a key issue of concern in 1986. ⁷⁰ During a meeting of the UN Committee on the Elimination of Discrimination against Women in 1989, CIM president Milagro Azcúnaga de Meléndez denounced the lack of attention to the issue of VAW in the UN Convention on the Elimination of all Forms of Discrimination against Women and announced that the CIM would henceforth start a regional campaign to promote international legislation on VAW. ⁷¹ In July 1990, the CIM convened the Inter-American Consultation on Women and Violence, inviting regional activists to draft the proposal that ultimately became the 1994 OAS Convention.

The OAS presented two institutional openings for developing an instrument on VAW, as long as it was framed as a human rights issue. The first was its system of human rights mechanisms, which consisted of the Inter-American Commission on Human Rights (IACHR) and the Inter-American Court of Human Rights, responsible for upholding the 1948 American Declaration of the Rights and Duties of Man and the 1969 American Convention on Human Rights. The second lay in CIM’s strong historical legacy, as it had been previously successful in establishing women’s rights and was particularly open to women’s rights activists. Its pressure on governments was also boosted by the almost simultaneous global developments. The CIM’s regional campaign overlapped with the preparation of the 1993 UN Conference on Human Rights in Vienna, where after two decades of mobilization VAW finally was recognized as a human rights violation. ⁷²

After the norm-establishment phase, feminist activists took on a new role of pushing national compliance. This proved to be particularly complicated compared to the relatively smooth and rapid process of norm development. Dissatisfied with the low level of compliance, CIM femocrats and regional activists together started to search for new transnational strategies to get national legislation in line with the convention. In 1998, Brazilian feminists and human rights activists filed a complaint with IACHR against the Brazilian government. It became a precedent-setting case, as the IACHR applied the OAS Convention for the first time to find a state responsible for not preventing and punishing domestic violence. The IACHR asked the Brazilian government for ‘the adoption of measures at the national level to eliminate tolerance by the State of domestic violence against women’. ⁷³ As the Brazilian government remained silent and did not file an answer, feminist activists continued to lobby. In 2004, the newly elected left-wing President Lula da Silva took up the issue and 2 years later promulgated a new law (see also previous section). The CIM, in close consultation with women’s organizations, developed a ‘Mechanism to Follow-up on Implementation of the Inter-American Convention on the Prevention, Punishment and Eradication of Violence Against Women’ (MESECVI). This new monitoring mechanism was adopted by the OAS General Assembly in 2004, when several national governments in the region had experienced a left turn. ⁷⁴ The MESECVI mechanism allows CIM to monitor state compliance by a Committee of Experts that consists of experts and activists who played a key role in getting the issue of VAW on the regional and national agendas. Women’s organizations and other civil society actors draft shadow reports and closely monitor the follow-up to the recommendations of MESECVI. Since the monitoring instrument has been introduced, many states have passed amended and new legislation. Still, not all legislation is properly implemented, and often insufficient budgets are allocated to make laws and policies function properly. ⁷⁵ In 2015, state parties to the convention agreed to strengthen the monitoring mechanism by appointing more experts to continue and expand its work. ⁷⁶

In sum, over the decades, national women’s rights activists have organized regionally and together with the CIM have put double pressure on governments to adopt a strong norm on VAW. They collaborated in systematically mapping the non-compliance of states and reporting this back to OAS. Their pressure resulted in the establishment of the follow-up monitoring mechanism MESECVI that provided the CIM with a stronger instrument to monitor state compliance. It also gave a key role to feminist activists in the Committee of Experts that monitors compliance.

Feminists and femocrats in the context of SADC

In Southern Africa, meanwhile, women’s movements and activists played a crucial role in the different stages that led to the 2008 SADC Gender Protocol. They had gained access to decision-making at the state level through their active role in the anti-Apartheid and Liberation struggles in the 1980s. Yet as a newly founded regional organization SADC had no established structure to address human rights violations, unlike the OAS. It also lacked an institutional framework for litigation. ⁷⁷ Women’s groups and women from government gender units had set up a Southern African taskforce to prepare for the Beijing conference. In August 1995, it called upon the SADC Summit of Heads of State and Government to ‘seriously address the upliftment of women of the region, and to eradicate [. . .] women and child abuse’. ⁷⁸ The Summit decided to transform the taskforce into a regional Advisory Committee on Gender composed of representatives from governments and women’s groups. These femocrats and feminists successfully lobbied for the establishment of a Gender Unit at the SADC Secretariat. In 1997, the Summit adopted the Declaration on Gender and Development. As feminist activists and SADC femocrats felt that VAW was still not properly addressed, they started to lobby for specific measures to combat VAW. ⁷⁹ Together with the Ministers of Justice and of Gender/Women’s Affairs they convened a Conference on the prevention of VAW in March 1998, where a number of specific measures were proposed. These recommendations were collected in the Addendum on VAW, approved as an integral part of the 1997 Declaration, and signed by the SADC heads of state and government in September 1998. ⁸⁰

In early 2005, the year of the 10th anniversary of the Beijing Conference, women’s activists started lobbying to upgrade the status of the Declaration and Addendum to a legally binding instrument. They established the Southern Africa Gender Protocol Alliance, a lobby of civil society actors, the SADC Gender Unit and parliamentarians concerned with gender equality. As a first step, the Alliance urged SADC to assess the level of compliance with the promises made in Beijing. With a view to Beijing+10, SADC Summit gave in to the pressure. To this end, the SADC Gender Unit and the SADC Parliamentary Forum initiated an Audit to monitor member states’ progress in formulating policies and laws in line with the Declaration and Addendum. The key finding of this 2005 Audit was that ‘while gender violence is now firmly on the political agenda in Southern Africa, laws, services and resources to address this scourge are patchy’. ⁸¹ It also signalled that NGOs were stepping into the void left by governments, playing a critical role in the training of service providers and providing safe places and legal aid to survivors of VAW, often financed by foreign donors. In response to the Audit, the SADC Heads of State Summit mandated the SADC Secretariat to develop a legally binding instrument. The Gender Unit formed a task force, including two representatives of women’s organizations, which was invited to draft the Gender Protocol. ⁸² The Protocol was adopted in 2008 and eventually entered into force in 2013. The SADC Gender Protocol Alliance, which developed into a regional network of transnational and national networks of civil society organizations and activists, compiles a yearly Barometer that measures progress towards the targets set in the Protocol.

The 2015 deadline of the Gender Protocol deliberately coincided with the revision of the UN MDGs. As the Alliance proclaimed: ‘Globally, we are moving from the MDGs to the SDGs. We cannot afford to be caught flatfooted. The SADC Protocol on Gender and Development needs to move apace’. ⁸³ Pushed by the Alliance, the Protocol was amended to align it to the Sustainable Development Goals, including a new deadline for 2030. ⁸⁴ As regards implementation, the latest SADC Gender Barometer notes that many problems remain. The Alliance pushes member states to conduct GBV-prevalence studies that can be used to track progress. ⁸⁵

Discussion and conclusion

The empirical sections have shown how the OAS Convention was unprecedented in its feminist and context-specific definition of VAW. At the national level, its content was in a first step watered down by de-gendering VAW and excluding the state as perpetrator, and at a later stage in a different aspect expanded (definition of family). The regional norm was strengthened again through a stronger implementation mechanism. Time and again, regional and national feminist activists connected with CIM femocrats engaged in framing battles through political pressure and lobbying, and litigation before the IACHR. The norm on VAW in SADC reflects regional specificities as well, referring for instance to harmful social and cultural practices, and medical experiments. Over time, its content has fluctuated. It has borrowed from the global level, but mainly from the continental AU level, resulting in a strong norm (although persistently silent on marital rape) and a set of specific targets. Regional feminist activists and femocrats established the Southern Africa Gender Protocol Alliance, a regional network of networks as a key player in this process.

These two cases of regional norm setting clearly show the polycentric and contentious character of norm diffusion. In contrast to what global-to-local approaches to norm diffusion sometimes suggest, regional processes do not necessarily follow global processes. Regional processes can also precede and inform global processes, as is illustrated by the Latin American concept of the state as perpetrator of VAW, or the Southern African setting of targets on GBV. As the norm developed in a ping-pong fashion and had multiple starting points, its trajectory was clearly multidirectional. This implies that regional norms, while embedded in processes of global governance, should not be understood as simply adopted versions of global norms. Instead, both the OAS Convention on VAW and the section on VAW in the SADC Gender Protocol are more progressive in their scope, transformative framing and binding character than global norms. Another dimension of polycentricity can be found in the regional specificity of the norm, which means that issues can be identified, objectives formulated and measures and strategies developed that fit the specific context of Latin American and Southern African women. This regional character facilitates norm diffusion within the region. Whereas the global UN agenda is troubled by tensions between ‘the West’ and ‘the rest’, a regional norm cannot as easily be dismissed as ‘colonial’. We have shown how they instead respond to specific needs and experiences that differ for each region.

We found that the regional level offers new, additional opportunities for effective lobbying. Our reconstruction of the development and implementation of regional VAW norms points to the crucial role of regionally mobilized and organized feminist actors. Change occurred when activists operated in tandem with femocrats and enjoyed access at the domestic, regional and/or global level simultaneously. Within regional networks, feminists exchanged experiences, developed a contextualized understanding and framing of the issue and generated expert knowledge that was instrumental to the emergence of regional gender equality norms. In addition, feminist networks played a key role in developing policy instruments, coordinating implementation or monitoring regional compliance.

Access of feminist and non-feminist actors cannot be taken for granted and needs to be understood in terms of the opportunities that the regional governance level offers. Regional networking among state and non-state actors was facilitated by shared histories and identities (of post-colonialism, authoritarian regimes and conflict) and language (in Latin America). Central players within these networks, as our data makes clear, are the gender equality policy machineries established within regional organizations. Despite their often limited mandate and resources, these machineries maximized their effectiveness by using feminist allies outside the organizations, seizing international and regional political opportunities. To increase their political leverage in (previously) closed policy domains, feminist actors within and outside regional (and national) institutions have developed cooperative constellations. The creation of these machineries is a feminist success in itself. While working within the confines of dominant organizational and policy logics, they successfully expanded the narrow focus and mandate of regional organizations and policy fields by including VAW in the mandates of OAS and SADC.

The contentious character of norms on VAW became clear from their laborious trajectories after adoption. We found that neither region reached full adoption, and that the norm did not simply become stronger (or weaker) over time, but instead went through instances of partial strengthening and partial weakening. Our CFA has shown how the content of the norm was strengthened, weakened, stretched and bent; in both regions, it was fixed for only short periods of time. Notwithstanding the enormous effort that institutional and non-institutional feminist actors put into defining and ‘fixing’ a transformative norm, their interpretation was opened up to contestation as soon as the journey of diffusion continued.