Do norms travel? The case of gender in Danish development cooperation

I Introduction

A large body of literature discusses how norms developed in one location affect norms and practices in other locations, and one area of particular concern is the influence norms established in international conventions and agreements have on national organizations and policies (Levitt and Merry, 2009; Risse et al., 1999, 2013). While much of this work focuses on the norms and how they ‘travel’, this article analyses a particular policy-making process in an organization and seeks to understand the role of global norms in the context of a wide variety of other issues influencing the process.

Two important approaches to the study of global norms and their effects concentrate on the diffusion and the translation of norms, respectively. Work on the diffusion of norms tends to focus on the mechanisms by which norms travel (Ansari et al., 2010; Bol et al., 2015; Finnemore and Sikkink, 1998; Strang and Meyer, 1993), while studies of translation are concerned with the processes by which norms are adapted to particular contexts (Lendvai and Stubbs, 2009; Mukhtarov, 2014; Stone, 2012; Zwingel, 2005, 2012). The former approach implies a tendency to suggest that social processes in different contexts become increasingly homogeneous as they are influenced by global norms. National organizations and policy-making processes are for various reasons assumed to be amenable to global norms which, once established, are relatively unaffected by social and political processes. The latter approach, on the other hand, emphasizes that travelling norms meet resistance if they are not translated and adapted to prevailing, contextual ideas and practices. Norms change when they travel and enter heterogeneous contexts. Thus, the two approaches emphasize forces of homogenization and of heterogeneity, respectively. However, both approaches take norms as their central point of analysis and study how norms travel or change in different organizational and political contexts.

This article seeks to provide an alternative perspective on the role of global norms in national organizations and policy-making processes. Instead of focusing on a particular norm, it analyses a policy-making process and the ideas, actors and structural conditions that influence that process. The purpose is to understand the role of global norms in the context of a variety of other issues which shape policymaking, without assuming that global norms are either indifferent to the specificities of the particular context or necessarily changed by them. In this respect, the article contributes to an understanding of how forces of homogenization and heterogeneity interact in specific political and organizational situations. It is argued that the way in which actors engage with global norms is not adequately explicated by referring to either diffusion or translation.

The article centres on the elaboration of the ‘Strategic Framework for Gender Equality, Rights and Diversity in Danish Development Cooperation’ (Danida, 2014). This strategic framework constitutes the fourth policy document expressing Denmark’s views on gender equality in the context of development cooperation, and is based on earlier policies published in 1987, 1993 and 2004. As such, gender equality has been, and continues to be, a long-standing interest in Danish development policy.

Global norms on gender equality have been discussed since the mid-1970s in international development cooperation. As noted in the ‘Introduction’ to this special issue, gender equality norms are not well defined or stable and nor do they constitute a homogeneous ‘package’ of ideas. The meaning and focus of specific norms change, given that every instance of international negotiation in the field of gender presents an opportunity for revising, challeng- ing or further developing their content. For instance, it has been argued that norms such as gender-balanced decision making and gender mainstreaming have developed in tension with each other, first converging and then diverging over time (Krook and True, 2012). This article concentrates on three norms related to gender equality, namely: gender mainstreaming, sexual and reproductive health and rights (SRHRs), and women’s empowerment. This is primarily due to their importance in Danish policy, which discusses mainstreaming at length, clearly supports SRHRs and employs the concept of empowerment in a variety of contexts. Mainstreaming and SRHRs were also salient issues in the former policy (Danida, 2004), while empowerment was emphasized to a lesser extent. The three norms have been selected because the policy specifically addresses them, although it also refers to other global norms including women’s political participation and prevention of violence against women.

Internationally, global gender equality norms have been the subject of much debate and the manner in which they have been interpreted by various development organizations has been heavily criticized in the academic literature (Cornwall, 2007; Cornwall and Rivas, 2015; Eyben, 2013; Moser and Moser, 2005; Mukhopadhyay, 2014; Washington and Tallis, 2012), which has also emphasized the undermining effects of a normative environment which focuses on market solutions to development problems (Cornwall et al., 2008; Perrons, 2005). The selected norms were codified in particular at the 1994 International Conference on Population and Development (ICPD) and in the 1995 Beijing Platform for Action, but they are not any more or less important than other norms related to gender equality, and what is more they are continuously challenged and reinterpreted by international actors. For instance, mainstreaming has been questioned by the African Development Bank (AfDB, 2012), SRHRs and violence against women have been highly contentious issues at the annual sessions of the Commission on the Status of Women (CSW; Kabeer, 2015), and the World Bank has significantly reinterpreted women’s empowerment relative to the way in which it is codified in the Beijing Platform for Action (Jones, the present issue). Accordingly, global norms on mainstreaming, SRHRs and women’s empowerment are amorphous ideas that, despite their tentative codification in various declarations, do not exist outside of the specific conception of particular actors.

The discrepancy between policy and practice is well established. Street-level bureaucrats face particular challenges which policies rarely take into account, for which reason policies are typically adapted into something manageable by practitioners (Lipsky, 2010; Schön and Rein, 1994). Why, then, focus on policies? First, policies in development cooperation are often used by governments to signal values and ideas to domestic audiences (Lancaster, 2007; van der Veen, 2011), and governments may use global norms to legitimize particular views in this process. Second, policies may be used by stakeholders to hold governments accountable. As foreign aid often serves many different purposes, including commercial, security and foreign policy interests, specific development policies constitute an instrument with which to influence the allocation and use of development assistance, and stakeholders may draw on global norms to legitimize their concerns. Third, policies belong to the realm of ideational struggles (Béland, 2009; Padamsee, 2009). While material concerns may underlie these struggles, policy ideas have to appeal to widely accepted notions, values and meanings in order to gain credibility. Global norms may play a critical role in this context, and all the more so in development cooperation because responses from the field rarely have a significant influence on the policy-making process. As a rule, the distance between taxpayers and aid recipients largely isolates policy ideas from being questioned on the basis of development challenges in poor countries. For these reasons, while not necessarily decisive for concrete development cooperation, development policies are likely to be influenced by global norms—assuming these have any significance.

The analytical approach used in the article is based on critical frame analysis (Verloo and Lombardo, 2007), frame-critical policy analysis (Rein and Schön, 1996; Schön and Rein, 1994) and the concept of framing processes (Benford and Snow, 2000; van Hulst and Yanow, 2016). The basic idea is to identify the different dimensions and competing frames in a policy discourse and to specify the deep-rooted understandings underlying different ways of framing the policy problem. In the present case, this requires an in-depth analysis of the policy document, clarifying what the policy problem and solutions are represented as being. As policy documents often contain ‘oddities, contradictions and ambiguities’ (Eyben and Napier-Moore, 2009: 287) reflecting different frames and ideas, one and the same policy can be variously interpreted as presenting a range of different views. Such incoherence may have both advantages and disadvantages. Ambiguity may be intentional and used strategically so as to appeal to a range of audiences (Eisenberg, 1984), may serve the purpose of protecting particular concerns (Sørensen, 2016) or may be the unintended consequence of a policymaking process wherein different parts of the policy are drafted by different people (Eyben and Napier-Moore, 2009). As such, a policy may reflect global norms in some sections, while disregarding or even contradicting them in others.

The next analytical step is to attempt to link different framings of the policy problem to different possible explanations. As was discussed in the introduction to the special issue, different factors—internal and external—to the policy-making organization may explain development policies, and the challenge here is to assess the explanatory power of these possible factors and of the global gender norms. A particular aspect of the Danish case is that gender equality is not a new phenomenon in Danish development cooperation. Accordingly, rather than explaining how a new issue gains political traction, the task here is to explain why a salient political issue is modified. These are two different tasks, given that an idea often goes through a life cycle: ‘At the beginning of the rule of a paradigm, it is its power to excite, to mobilize and to energize that is most noticeable; toward the end, it is its unquestionability, obviousness and taken-for-granted explanatory power’ (Czarniawska and Joerges, 1996: 37). Likewise, it is argued that ideas move from conscious (cognitive and/or normative) debate in the foreground to a taken-for-granted status in the background (Campbell, 2004, 2008). If gender equality has acquired such a taken-for-granted status in Danish development cooperation, an element of path dependency may be important compared to a situation in which a new issue is developed in a policy document.

In order to explore these issues, I have analysed the policy document, interviewed central staff members involved in the drafting of the document and attempted to gain access to written inputs into the policy-making process. As a researcher and a practitioner with a long-standing involvement in Danish development cooperation, in policy studies commissioned by Danida, and in public debates on the subject of aid, I contacted the head of Danida to request permission to access staff and unpublished material. (Danida and the Danish Ministry of Foreign Affairs are one and the same organization. Danida is a term used to describe the ministry in the context of development cooperation.) In granting permission for this, the ministry organized a meeting to facilitate my work, and on this basis, I subsequently contacted staff members individually. Despite a widespread willingness to share information with me—no doubt facilitated by my different relations of cooperation with the ministry—I was sometimes treated as an outsider to whom unofficial material should not readily be disclosed, and it was not possible for me to attend internal meetings concerning the implementation of the policy. Accordingly, the article is not based on an ethnographic study of the ministry or the policy-making process. The main sources used comprise: the policy document itself (Danida, 2014); two studies conducted as inputs into the policy-making process (Kristiansen, 2013b, 2013c); the minutes of an expert workshop organized during the drafting of the policy (Kristiansen, 2013a); two synopses of the strategic frame- work presented to an internal Programme Committee and to the Council for Development Policy, respectively (Danida, 2013b, 2013c), and the minutes of the Council’s meeting (Danida, 2013a); seven interviews with central participants in the policy-making process; and participant observation at the meeting on 19 September 2014 at which the strategic framework was presented to the public.

Following a brief historical review of the approach to gender equality in Denmark’s development cooperation and an outline of the policy-making process, the article analyses how the policy frames gender equality as a problem, and what policy solutions this framing gives rise to. Based on this understanding of gender equality, the subsequent section examines how the document engages with the three aforementioned norms. This is followed by a discussion of the contingency of policymaking, and the conclusion, in which the arguments presented are wrapped up.

II Danish Aid and Gender Equality

In the early years of Danish development cooperation, a central mantra was not to influence political processes in decolonized societies (Bach et al., 2008). However, in the 1980s politicians began to propose that Denmark should pursue particular political priorities through its development assistance. The first attempt to outline a general policy was published in 1988 (Danida, 1988), and gender equality featured as one of six qualitative goals in this policy. At the same time, an action plan for assistance to women was published (Danida, 1987), partly in response to a parliamentary request. The action plan was based on a Women in Development (WID) approach, seeking to raise awareness of the fact that development is not exclusively about men. A mere six years later, a new and more radically-phrased policy was published (Danida, 1993). Maintaining a WID approach, the policy called for mainstreaming women’s needs and priorities in development cooperation as well as transforming dominant social and economic structures so as to secure women’s basic human and economic rights. In the following year, a white paper identified poverty reduction as the overall objective, with gender equality, environment and democratization as the three cross-cutting concerns (Danida, 1994).

In hindsight, the 1990s constituted the heyday of Danish foreign aid. In this period, foreign aid received strong support of the Danish government, the budget allocated to foreign aid was large and constant, the Danish Ministry of Foreign Affairs developed an ‘active multilateralism’, and a professional and long-term commitment to programme support increasingly characterized bilateral cooperation. With the change of government in 2001, political support for development cooperation decreased significantly, the budget was cut and the military became Denmark’s main foreign policy instrument. Development cooperation was decreasingly seen as an important element in Denmark’s international engagement and increasingly as a playground for domestic politics. A new policy on gender equality was launched in 2004, this time adopting a Gender and Development (GAD) approach. By and large, it consisted of a consolidation of existing practices rather than an attempt to invest new energy into gender-related activities (Engberg-Pedersen, 2016). Administratively, the ministry’s work on gender continued, most notably with the publication of a Gender Equality Toolbox consisting of booklets on different aspects of GAD (Danida, 2008) and the development of an e-learning course on gender equality in 2009. At the political level, the then development minister, Ulla Tørnæs, launched an international MDG3 campaign ‘to do something extra’ to promote gender equality and women’s empowerment. The main element in this campaign was targeted at high-level representatives of governments, the private sector, civil society and international organizations, committing them to work in favour of the MDG3 (Danida, 2009: 25–29).

III The Policy-making Process

In May 2012, a new general policy for Danish development cooperation was adopted, highlighting a human rights-based approach to development. Subsequently, the so-called strategic frameworks on various issues were elaborated to turn the new general policy into reality. The policy on gender equality was one of these strategic frameworks and it was, accordingly, drafted as part of the minister’s attempt to introduce a human rights-based approach in Danida’s work. Moreover, the existing policy on gender equality (Danida, 2004) was by then 10 years old and, in the opinion of ministry staff, was in need of revision. The first step in the elaboration of the policy was to commission a study of the most important trends in international gender equality efforts since 2005 (Kristiansen, 2013b) and a study of the lessons learned with respect to gender in Danida’s 2004–14 evaluations (Kristiansen, 2013c). The former study was commenced in early 2013 and included a review of policies and programme documentation of other aid agencies, interviews with staff in bilateral and multilateral institutions, and three gatherings: an expert workshop, a meeting with staff assigned as gender focal points at the Danish embassies and a public meeting. On this basis, the report was submitted in June. The latter study was published in September.

In early 2013, the ministry established a project group tasked with drafting the policy. The group was coordinated by the Department for Development Policy and Global Cooperation and, according to an interviewee, comprised representatives of seven other departments including the Technical Advisory Services and the three departments with country desks. The group drafted a synopsis of the policy and, according to one participant, the work consisted of a focused process based on an exchange of written input, in contrast to the previous approach, with formal meetings at which the various departments could express their views. The group endorsed a synopsis, ‘Strategy for Promoting Gender Equality in Danish Development Cooperation’, which was submitted to the internal programme committee and was the subject of a public hearing in June 2013. Three organizations (AWID, LGBT Danmark and FORDI—Forum for Rights and Diversity) commented on the draft, and a revised synopsis was submitted to the Council for Development Policy in August 2013. The council consists of representatives of civil society organizations, trade unions, employers’ associations and academia, and discusses Danida’s new policy and strategic initiatives. The synopsis was also presented to the Foreign Affairs Committee of the Danish Parliament. In this process, Danish embassies working with development cooperation had the opportunity to comment on the policy. Following these various consultations, the strategic framework was finalized, but it was not published until August 2014, seemingly delayed by the minister’s busy schedule. Ultimately, the minister was unable to attend the meeting at which the framework was presented to the public.

IV Framing Gender Equality

How does the policy frame gender equality? As a basis for the subsequent analysis of the role of specific global gender equality norms in the policy, this section discusses the policy problem and the associated solution that the document identifies as central. While gender equality is presented as a relatively clear goal in the Millennium and Sustainable Development Goals, it can take on very different meanings. One way of understanding these various meanings is to distinguish between gender equality as inclusion implying equal opportunities for women and men, as reversal implying a contestation of dominant male norms or as displacement implying a deconstruction of gender in general (Squires, 2005; Verloo and Lombardo, 2007; Walby, 2005). These three interpretations differ substantially, but are not necessarily incongruent with one another. The first view stipulates that women have been excluded from the ‘normal’, with the goal being for women to have the same access to economic assets and political influence as men enjoy. It is a supposedly objective approach emphasizing that people are autonomous. The contemporary social order is accepted, yet women’s lack of access and the obstacles that women face are not viewed as structural phenomena integrated into the social order. The second view criticizes dominant norms and hierarchies, and seeks to accommodate female norms that differ from the so-called ‘normal’. The perspective here is interpretative and emphasizes that fundamental differences between women and men are decisive for social development. Gendered norms are taken for granted in this approach, which calls for a change of the existing social order. The third view questions the common categories used to understand gender, and seeks to accommodate the diversity of peoples’ experiences with respect to gender, age, race, ethnic identity, etc. The point is that neither women nor men are homogenous categories, that women’s (and men’s) social positions are highly diverse and that processes of engendering are central, while fixed categories of gender are inadequate when we seek to explain social change. Generalizations should be avoided, contextualization should be emphasized and gender stereotypes should be transformed. While clearly distinct, these three views on gender equality should not be seen as incompatible (Squires, 2005). It is widely accepted that the diversity of groups and individuals should be recognized, and that identities cannot and should not be homogenized. This does not mean that equality is impossible, but merely that it should be limited to those areas where marginalized people seek equality. Thus, inclusion may be highly pertinent in particular situations, whereas reversal and acknowledgement of various gendered concerns are important in others within a framework which recognizes the diversity of individual actors.

As noted earlier, policy documents frequently contain ambiguities, and Danish gender policy is no exception in this respect. The policy features phrases that reflect all of the three aforementioned views on gender equality. Referring to the human rights-based approach, the policy focuses on ‘women’s rights and equal access to decision-making, resources and opportunities’ (Danida, 2014: 7). Here, the policy seeks to address the discrimination and marginalization of women and girls. Perhaps due to the fact that the policy is also intended to make the case for support for lesbians, gays, bisexual, transsexual and intersex (LGBTI) persons, achieving equal rights and opportunities is emphasized in the policy. The minister argues in his foreword that

Danish development cooperation supports efforts to increase women’s participation in decision-making at all levels. We will assist women and girls to access resources and seize opportunities that enable them to take control over their own lives. And we will support efforts to fight discrimination and allow all citizens to play an active role in forming the societies they live in. (Danida, 2014: 3)

The last sentence conveys the message of inclusion. Inclusion in the category of citizens, with the associated rights and duties, is a central goal of the policy. The phrasing of these statements fails to highlight the diverse spectrum of gender identities, and the policy problem identified here is that women and girls do not have the same rights, access and opportunities as men and boys. The framing of gender equality under the umbrella of a human rights-based approach contributes to the focus on inclusion, for instance, when the policy promotes a ‘strategic focus on: Enabling rights holders—women, men, boys and girls—to claim their basic and fundamental human rights’ (Danida, 2014: 6).

The document also contains statements pointing towards the second view on gender equality. It repeatedly calls attention to the ‘root’ or ‘fundamental’ causes of inequality and discrimination to ‘power relations’ and ‘redistribution’. These notions suggest a critical view of the existing social order, a need for social change and a confrontational approach to actors who discriminate against and marginalize women and girls. The document notes that ‘[t]raditionally, […] efforts have focused on addressing barriers to women’s participation and access to resources […], but future focus will emphasize redistributive initiatives’ (Danida, 2014: 6). As such, the policy frames its newness in part as a move towards a more political orientation. How deep this more confrontational approach goes is, however, not obvious when it is stated that ‘men and boys are indispensable partners in bringing about change’ (Danida, 2014: 7).

In line with the third view focusing on dis- placement, the policy argues that ‘[w]omen’s, as well as men’s, situations are also defined by the intersection between different forms of identity (i.e., gender, class, caste, ethnicity, sexual orientation and age)’ (Danida, 2014: 7). This emphasizes that women and men are heterogeneous groups and that Danish support will focus on people with multiple disadvantages. Moreover, the policy states:

Gender should be understood as a cultural and social construction associated with specific expectations and norms which construct certain stereotypes that may result in limited tolerance of diversity in identity, appearance and behaviour. (Danida, 2014: 4)

The term diversity is included in the title of the policy in order to recognize the rights of LGBTI persons. This promotes an appreciation of different gender identities, and in a few sentences, equal rights and access to resources and opportunities are presented as a means ‘to make individual life choices regarding […] gender roles and identity, sexuality, education and profession’ (Danida, 2014: 6). Thus, the policy may be interpreted as recognizing that equal rights and opportunities for women and men are a condition for diversity at the individual level.

One further aspect of the understanding of gender equality is that it is not simply presented as a goal in itself, but also as a means of achieving poverty reduction and sustainable development:

Women’s economic empowerment is first and foremost a right’s issue—but it is also ‘smart economics’ as ensuring women’s equal access to resources increases overall productivity and growth as well as improved access to nutrition, health care and education for her family. (Danida, 2014: 5)

This argument is in line with the view emphasizing inclusion. Despite the detours into reversal and diversity, providing women with the same rights, political influence and access to resources and opportunities as men appears to be the central concern of the policy. The minister writes in the foreword that: ‘Women and girls remain the focus of the strategy, because they are disproportionally affected by poverty, discrimination and marginalization’ (Danida, 2014: 3). Nonetheless, the document is far from unambiguous, and has the potential to be used to legitimize various concrete activities.

Looking at the documents and discussions which feed into the policy-making process, it does not seem that the understanding of gender equality has been much debated. The main concern of the commissioned study of international trends is to clarify the implications of the human rights-based approach for gender equality, and it is argued that the ‘fundamental shift in development approach is from focus on needs to rights ’ (Kristiansen, 2013b: 11). A gendered analysis of rights-holders and duty-bearers is stated as being a central task, and this implies that efforts to develop gender equality will chiefly concentrate on the policies and practices of state institutions.

Discussing the policy’s understanding of gender equality with the various interviewees, two general points became apparent. Gender and gender equality have not been subject to significant debate at the conceptual level, and the policy is mainly based on the view which emphasizes inclusion. Some of the interviewees identified elements reflecting the two other views, and two staff members indicated that equal rights and access are policy-level concerns, while pragmatic development cooperation on the ground has to take the other views into account. Thus, the policy’s use of different understandings of gender equality may be a consequence of the perceived need to send policy signals to a broad domestic and international audience while also providing some room for manoeuvre to staff at the embassy level. Several respondents from the ministry argued independently that a pragmatic, experience-based, target group-focused approach has generally characterized Danida’s aid practices, in contrast to the more policy- focused practices of, for instance, Swedish Sida. It is possible that pragmatic organizational concerns have influenced the policy, since the ministry has initiated a large number of development policy changes over the years (Martinussen, 1989; Olsen, 2005) and an extensive decentralization of aid management to embassies took place in 2003 (Engberg-Pedersen, 2014), necessitating a certain degree of pragmatism.

It is clear from this analysis of how the policy addresses gender equality that the policy problem is framed primarily so as to make women’s inclusion in economic and political processes the solution (for a similar reading of Norwegian development policies, see Østebø and Haukanes, 2016). There is less emphasis on a confrontation with dominant norms and practices or a deconstruction of gender identities. However, it is also worth noting that the policy accommodates all three views, thereby allowing for a pragmatic, context-dependent approach to gender equality. It is in the context of this framing of gender equality that the policy engages with global norms.

V Global Gender Equality Norms and the Policy

As with the preceding three gender policies, and in common with other Nordic countries (see Selbervik and Østebø, 2013), the ‘international framework’ constitutes a significant point of reference and is presented and analysed in a separate section of the policy. Various relevant UN conventions, resolutions, conferences and organizations are cited, and the status of the promotion of gender equality at international level is briefly outlined. Three points stand out: first, the framework is presented as based on human rights. The Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW) is referred to as ‘the international bill of rights for women’ (Danida, 2014: 8), and it is emphasized that ‘women’s human rights instruments provide a specific legal framework in which gender equality and non-discrimination on the grounds of sex are explicitly applied to all categories of rights (civil, political, economic, social, and cultural)’ (Danida, 2014: 8). The policy suggests using CEDAW against states that uphold discriminatory legislation despite having ratified the convention. In later sections, legislation is repeatedly mentioned as an area of focus, such that the document’s framing of international norms brings laws and public reform processes to centre stage. While this is largely in agreement with CEDAW, the convention is broader in scope and addresses discriminatory customs and practices as well as ‘discrimination against women by any person, organization or enterprise’ (United Nations, 1979: Article 2 (e)).

Second, the international discussion of gender equality is described as a struggle between modernists and traditionalists; between those who seek to strengthen women’s rights and those who wish to limit them. The policy argues that

the pressure on the gender equality agenda has increased in recent years. This attack on fundamental rights mainly come from conservative groups that seek to undermine progress on gender equality and women’s rights […] Disagreements in this domain often emanate from religious, cultural and traditional beliefs and practices. (Danida, 2014: 9)

The document goes on to state that disagreements can best be resolved by ‘evidence-based factual arguments’, and that UN organizations such as UN Women play an important role in this context, ‘providing factual and impartial evidence’ (Danida, 2014: 9). This framing of the international discussion clearly situates Denmark as part of the struggle for gender equality.

Third, the policy seeks both to gain legitimacy by referring to international norms and to suggest that Denmark influences these norms. The positive terms in which the policy describes CEDAW and UN agencies integrates Danish gender policy thoroughly into the international framework. At the same time, Denmark has ambitions of influencing international discussions of gender equality. For instance, it is argued that ‘Denmark seeks an active dialogue with UN Women to ensure a human rights-based approach in UN Women strategies, policies and operational programmes’ (Danida, 2014: 9).

VI The Contingency of Policymaking

One interviewee remarked that what passes through the system is quite haphazard. Given that several peers and equal-ranking departments have contributed to authoring the policy, one may hypothesize that, without thorough editing at the management level, inconsistencies between different parts of the strategic framework may easily be the unintended result. However, it may be no accident that some points are deleted, while others remain in the text. And ambiguity, or at least diversity, may to some extent have been a conscious aim so as to render the strategic framework relevant in a variety of situations and contexts. As several members of staff emphasized, the policy should ideally be relevant both in widely differing national contexts and in international negotiations.

In social movement theory, it has been argued that actors who engage in framing activities are ‘actively engaged in the production and maintenance of meaning for constituents, antagonists, and bystanders or observers’ (Snow and Benford, 1992: 136). The resulting frame ‘simplifies and condenses the “world out there” by selectively punctuating and encoding objects, situations, events, experiences and sequences of actions within one’s present and past environment’ (Snow and Benford, 1992: 137). Transferring these observations to policymaking, actors select particular ideas within their environment in order to construct an understanding of the world which appeals to particular constituents. Moreover, the more consistent a frame is, the more credible it is (Benford and Snow, 2000). However, this creates a dilemma since ideas that appeal to different constituents may not form a consistent frame or policy.

It appears that earlier drafts of the Danish gender policy contained more extensive discussions of LGBTI issues, but that these were downplayed in the final version of the policy. According to one member of staff, this was done because Danish ambassadors indicated that it would be difficult to make use of the policy if LGBTI issues were strongly emphasized. Thus, pragmatic concerns in the field may here have influenced the policy- making process. Another member of staff recalled that the first draft of the former policy from 2004 contained a discussion of the third gender, but that the management deleted it for unknown reasons. These are examples of ideas that did not make their way through the policy-making process as intended by the staff drafting the policies. In the first case, an internal audience opposed the visibility given to a particular issue; in the second, the management may have considered the idea to be politically unfeasible. One member of staff noted that policymaking involves a constant consideration of what is acceptable for the minister and the government, and that the members of staff who draft policies may choose to leave out otherwise relevant issues if they are regarded as politically inconvenient. An example of a proposition that was retained in the policy, to the surprise of a member of staff, is a reference to transformative change. This notion, which is used by UN Women to denote a change of the structural impediments to gender equality (UN Women, 2013) and in the academic literature in contrast to instrumental change (e.g., Eyben, 2010), is used five times in the document to indicate profound social change, sometimes linked to women having political power or being in positions of influence. Another member of staff pointed to the references to transformative change and redistribution as some of the main changes from the previous policy, thus indicating that it was not by chance that they were included in the document.

This suggests that the drafting of a policy by a project group consisting of peers with diverse experience and responsibilities, which is then approved by management, easily results in a text in which differing—not strictly—coherent ideas, issues and events are selected for presentation and only the most politically inexpedient points are left out. Peers cannot easily censor each other and may instead ‘opt out’ of the drafting process if they disagree with others, something which, according to one member of staff, happened in the course of drafting an earlier policy. Another member of staff assessed that the gender policy was not believed to be of significant political importance, which indicates that management’s examination of the text may have been limited to the censoring of inexpedient points. While the policy should not appear incoherent, the policymaking process has resulted in a text with a set of not entirely consistent propositions that partly reflect the experiences of those drafting the policy and partly seek to appeal to different constituents.

Given that domestic constituents represented in the Council for Development Policy did not support the policy’s ambition to address power and structural causes for discrimination, maintaining this focus may be linked either to staff members’ positive experience with such an approach or to a perceived need to appeal to other constituents. One member of staff argued that specific country-level experience and international norms and signposts are the two main factors guiding policy-making processes. However, another respondent questioned the extent to which practical development cooperation feeds back into policymaking and an external observer considered that a norm-setting competition with other bilateral aid agencies plays an important role in policy- making. All in all, the interviews suggest that the policy’s somewhat inconsistent set of propositions can be explained partly by the way that the drafting of the policy was organized, partly by a pragmatic concern with enabling the policy to address qualitatively different contexts and situations, and partly by a consideration of three different groups of constituents, namely, the minister and the government, specific members of staff (in this case, in particular, the ambassadors) and peer aid agencies.

Another dimension framing the policy- making process concerns organizational pressures and priorities. Continuous administrative cuts have reduced the number of staff at Danida to an extent which significantly influences the scope of policy initiatives. According to some respondents, a central concern was to elaborate a policy which would require as few resources as possible. Being lean was the name of the game, and one member of staff noted that there is a great deal of freedom to shape new initiatives—as long as they are not costly. Thus, early in the policy-making process, the ambition was to get rid of the special interventions that, together with mainstreaming, were perceived by several informants as the main instrument of the 2004 policy. Special interventions incurred too great an administrative burden; and as the funds set aside for promoting gender equality in the 2004 policy had been abolished years ago, it was logical to develop a policy without such initiatives. Moreover, one member of staff argued that while special interventions may produce reasonable results, they rarely have a broader impact.

Thus, with lukewarm support for mainstreaming and an ambition to eliminate special interventions for administrative reasons, the policy risked ending up with no instruments. The invention of targeted measures was perhaps intended to remedy this, but as noted, the difference between these and the special interventions is minimal. However, targeted measures have no separate funds and should be integrated into sector programmes and other activities. Thus, they both appeased management’s concern with costs and provided an alternative to the discredited and diffused instrument of mainstreaming. Nevertheless, one member of staff suggested that one of the new points in the policy is that ‘targeted interventions have returned’—which corroborates the elusiveness of the distinction between special interventions and targeted measures.

Overall, global norms pertaining to gender equality have not been translated into the Danish gender policy in a straightforward manner. On the one hand, the international normative framework is important in that it legitimizes the Danish position, and it is framed to do so in the document. On the other hand, the framework is clearly regarded as a means of influencing partner states and as something to be influenced by Denmark. Thus, gender policy strives to make use of international norms and to influence them rather than to adopt them uncritically. However, the norm of women’s empowerment as ‘smart economics’ seems to have slipped relatively inadvertently into the document.

While the main motivating forces behind the policy appear to be domestic politics and agency dynamics, at a deeper level international norms exercise substantial influence on the policy. First, it is not possible to be at the cutting edge without relating it to current norms and practices. The advocacy of SRHRs is only possible on the basis of CEDAW, ICPD and the Universal Declaration of Human Rights. Second, notions such as transformative change, intersectionality and LGBTI persons refer to emerging international ideas and norms which have been translated into the Danish human rights-based approach to gender equality. They have not been developed through thorough analytical work, neither in the ministry nor by Danish constituents. Thus, the policy neither directly adopts international norms nor rejects them, but instead interacts with them in an intimate manner.

Conclusion

This article gives rise to a set of tentative propositions. It seems that appealing to particular constituents (the minister, particular staff members and peer agencies) has significantly influenced the policy-making process, resulting in a not entirely coherent policy. This is a challenge to the literature on framing, according to which coherence strengthens frame resonance. It may perhaps be useful to distinguish between degrees of coherence since an obviously incoherent policy is unlikely to receive support. At the public meeting at which the policy was presented, no one in the audience raised the issue of coherence.

Global norms have not deliberately been used as the starting point in framing the policy. They are important in terms of providing an international framework, they legitimize the policy and they are used to promote particular viewpoints, but they are not regarded as the reason why the policy was developed. However, the policy at all times relates to global norms; and some of the concepts introduced clearly refer to emerging international ideas that may develop into norms. As such, the policy has been developed with the ambition of being progressive, being at the cutting edge of current discussions and setting new standards for how to address gender equality. In a somewhat subconscious manner, global norms have heavily influenced this work, and it would be erroneous to argue that the international normative framework does not constitute an important foundation for the policy; it is integrated into the vocabulary of the staff members who drafted the policy, and on that basis they have consciously set out to develop a policy which ideally will break new ground and respond to the concerns of different constituents.

More specifically, the three gender norms discussed here appear to influence the gender policy in rather different ways. The mainstreaming norm has been challenged, although it has not officially been abandoned. Making reference to a thematic review, an external study and conclusions compiled from evaluations of Danish aid, the policy argues that mainstreaming has not produced the desired effects, and instead policy dialogue and targeted measures are presented as the way forward. Given that mainstreaming no longer appears to be a strong norm internationally, the change of policy could be an expression of this circumstance, but it may also reflect the difficulty of sustaining an idea for extended periods of time, notably if staff increasingly perceive it to be ineffective. In any case, the policy drafters have quite openly sought to challenge the mainstreaming norm as a decisive means of achieving gender equality.

The SRHRs norm is dealt with in a different manner. Although it has not been developed conceptually, it has been promoted forcefully. It fits nicely with the human rights-based approach, which politically has been adopted as the framework of all Danish development assistance, but it also helps Danida to develop a progressive profile among aid providers. Thus, Danida seeks to play the role of norm entrepreneur with respect to SRHRs. Finally, a particular normative interpretation of women’s empowerment has been adopted rather uncritically and without much consideration. While ‘empowerment’ is anything but well defined, the term is used repeatedly in relation to the economic sphere, social services and the countering of violence. The underlying message is that women are victims who would be strong producers and consumers to the benefit of the economy if they had power. This power is achieved through empowerment, which is not supposed to challenge structural conditions that prevent women from liberating themselves in whatever way they choose. The particular normative interpretation of women’s empowerment appears to have diffused into Danish gender policy relatively unnoticed, possibly because it fits well with contemporary ideas about the market as the main engine of development.

Accordingly, the case of Danish gender policy demonstrates a rather close interaction between homogenizing global norms and context-specific conditions. The three gender norms examined here do not simply diffuse and nor do they become the subject of significant struggle and translation beyond recognition. Rather, the key to understanding their role is to be found in a highly contingent policy-making process influenced by particular practices and the diversity of the staff members involved in drafting the policy, by certain stakeholders and not others, by organizational pressures and priorities, by a partly practice-based organizational culture, and by a particular Danish normative environment. An additional aspect is that some of these explanatory factors may change relatively quickly, possibly producing a different policy-making process whereby global gender norms may be addressed differently. Furthermore, as global norms and emerging ideas are also used for advocacy purposes internationally to create a particular image among peers, it becomes less useful to focus on norm diffusion or norm translation as opposed to the interaction between contingent policy processes and norms when analysing the role of global norms in development co- operation. This proposition is supported by the point that gender issues have been previously agreed in international conventions, without this having any impact on political processes or existing practices (Zwingel, 2005, 2012). Internationally agreed ideas only become norms when they interact with policy processes where they may then be supported, changed or discarded, thereby influencing their normative status. This is very much the implicit result of Danish gender policy as gender mainstreaming has been weakened, SRHRs reinforced and women’s empowerment strengthened as ‘smart economics’.