The myth of Sweden’s success: A deconstructive reading of the discourses in gender mainstreaming texts

Introduction

The question of how to achieve gender equality has been on the public agenda for decades, taking various forms in different times and places. One challenge has concerned how to develop and impose policy recommendations for societal change and achieve effective implementation. In light of the severe economic and democratic crises of the last decade (Walby, 2015) and the ambitious scale of government initiatives, it is of particular importance to challenge perceived successes. Gender mainstreaming is often considered a new, modern and innovative way of organizing gender equality work (Callerstig, 2014), and it has given rise to extensive research and numerous policies. Embedded in the discourse of gender mainstreaming, it has been argued, lies a story of progress, development, innovation and future success (Schmidt, 2005), and its ambition of ‘doing better’ fits neatly into the practice of ‘best case’ scenarios and the spread of ‘good practice’. On the other hand, conventional research about gender mainstreaming has largely focused on criticizing the strategy by identifying and addressing the possible negative effects of this way of promoting gender equality (Bacchi and Eveline, 2003; Benschop and Verloo, 2006; Keisu and Carbin, 2014; Martinsson et al., 2016; Prügl, 2009; Walby, 2005).

Sweden is of special interest in this regard, as it is ascribed a unique position as a successful case in the discourses of gender equality (Martinsson et al., 2016) and gender mainstreaming (Sandberg and Rönnblom, 2013). However, as I will argue, the discourse about Sweden’s success in introducing and implementing gender mainstreaming at national, regional and local levels is composed of myths. In this article I will present an alternative reading of the Swedish success story. Within research on gender mainstreaming there has been a lack of recognition of the construction of discourses in different kinds of texts. The present article aims to address this lack by contributing an empirical study to the field of critical studies of gender mainstreaming (Magnusson et al., 2008) and analysing how the discourses of gender mainstreaming in Sweden are constructed.

As will be shown, I have identified in the academic gender mainstreaming literature a specific discourse that displays Sweden’s success. However, there are also subtle and complex indications, which can be found in domestic evaluations and in some Nordic research, that the story of gender mainstreaming can be told differently. I will also argue that the discourse of successful gender mainstreaming contains an element of adaptation from gender equality policy. Thus, three different discourses will be identified: (1) a discourse portraying Sweden as a successful case of gender mainstreaming; (2) a counter-discourse depicting Sweden as less successful in this regard; and (3) a conflation between gender mainstreaming and gender equality policy.

The theoretical framework employed in this article is inspired by Della Sala’s research on political myths (2010). In the following section, I will present the concept of myths and expand on my analytical framework. Following this is a description of the empirical material and how it was selected. The analysis of the material is then presented. My findings and discussion conclude the article.

The concept of myths

Inspired by Della Sala (2010), I apply the concept of political myth as a theoretical tool. However while Della Sala studies formal institutions such as EU, nations or states, I use the concept of political myth to capture more informal ways of governing, such as gender mainstreaming. I draw from two main areas of Della Sala’s argument. First, discussing the function of myths, he states that myths are stories with a very specific function in society. Political myths form a history with the purpose of creating visions where better times lie in the future. He concludes that myths can be born from facts, but that they then take on a life of their own (2010). Second, Della Sala argues that myths are valuable for determining normative positions in governing, and that myths do not necessarily stand in opposition to knowledge and reason (Della Sala, 2010). He tackles this tension between facts/knowledge and myths when arguing that it is not the accuracy of myths that should be studied, but rather ‘their capacity to make us believe in the normative vision they are presenting’ (Della Sala, 2010: 7).

To summarize, Della Sala asserts that the concept of myth has been employed to distinguish between statements based on rigorous analysis and statements that are simplistic and distorted. In making such a distinction, one must recognize the complex relationship between ‘knowledge’ and ‘discourse’. My position in this article, when pointing to different types of knowledge production, is that knowledge can be more or less well-grounded in systematic empirical research. This also concerns the extent to which different forms of knowledge are integrated into, or excluded from the dominant discourse. As Della Sala finally points out, in political myths, ‘knowledge’ is put to use as facts to make sense of political history.

Scholars have applied the concept of myths in slightly different ways and to different policy areas. Empirically it has been applied to the area of entrepreneurship (Gibb, 2000) and in policy studies on the EU (Manners, 2010). Gibb used the concept to argue that within the myths lie different sets of assumptions, partly about how the world works, and partly about how to build policy actions and thereby to (re)produce discourse (Gibb, 2000). I understand Gibb as contributing to the understanding of how to study the effects of myths.

To expand my theoretical framework I need to include tools on how political myths can be operationalized, for which I turn to contributions from discourse theory. The post-structuralist theorist Lévi-Strauss argues that myths deal with unwelcome contradictions (1968), and Edelman suggests that myths are both societal and individual coping strategies to handle issues that are uncertain, controversial and filled with ambiguities and anxiety (Edelman, 1977). Building on this, Laclau uses myths to deconstruct the dominant structural objectivity, and to deconstruct what is seemingly natural and objective (1990). According to Laclau, myths have a social function when they smooth over dislocations in events, enabling discourses to be upheld (Laclau, 1990: 60ff.). According to Laclau, dislocations occur all the time within any discourse. Dislocation can be explained as occurring when an articulation destabilizes or ‘shakes’ an event, shifting it out of place. Following this, (re)articulation can occur, enabling the dominant discourse to be upheld. Societal imaginaries therefore ‘resolve’ those dislocations. Because dislocated events cannot be integrated into existing discourse, myths ‘cover’ them over so that stability is maintained. The function of myths is therefore to repair dislocated structures by offering a transformed collective social imaginary (Laclau, 1990); the point being that by taking the form of myths, dislocations become neatly embedded in the dominant discourse.

To conclude, in my analytical framework I follow Della Sala’s understanding of the concept of myths, while adding Laclau’s concepts of societal imaginaries and dislocations. I do this because I want to understand the construction of discourses in gender mainstreaming and I want to be able to discuss societal and political practices and future consequences of these constructions. Gender mainstreaming is a political strategy to address the complex social challenges of modern life. In this article I argue that myths are embedded in gender mainstreaming to generate legitimacy for that particular political rule/strategy.

Understanding gender mainstreaming and gender equality

Gender mainstreaming and gender equality are two different concepts. The relationship between them cannot be treated as clear cut, as they are interconnected in a discursive field. A multitude of different operational variations have been used in both policy areas, and the concepts have been used despite their lack of any clear definition. For decades, this extensive, and not particular careful use of the concepts has resulted in a problematic confusion.

In the case of gender mainstreaming, both the vagueness of the concept and its theoretical diffusion are core topics of debate in gender mainstreaming studies (Andersson and Hedlund, 2012; Stratigaki, 2005). Gender mainstreaming is often described as a main strategy for implementing gender equality, but as a strategy it lacks lucidity. Gender mainstreaming can be described as a means to make states, organizations or businesses more gender aware, and hence to make gender visible and to promote gender equality. Some scholars argue that gender mainstreaming has become a goal in its own right (Rönnblom, 2011). The spread of gender mainstreaming as a policy idea has resulted in great variation of execution in practice. For example, the extent of involvement of the women’s movement in gender mainstreaming and the degree of its institutionalization differ widely, ranging from non-existent (in Sweden) to very high (in Taiwan) (Andersson and Hedlund, 2012; Peng, 2015). Using a concept from discourse analysis, one could categorize gender mainstreaming as a floating signifier (Bergström and Boréus, 2005).

In this article I define gender equality as a political project with the objective of achieving a gender-equal society. Turning to the concept of gender equality, one can see variations in different policy texts. These differ to such an extent that some scholars argue that the concept should be considered a process rather than a distinct definition or policy (Magnusson et al., 2008). Aligning with this idea, but from a discursive viewpoint, I understand the concept of gender equality as a floating signifier (Bergström and Boréus, 2005).

So, what makes Sweden such a significant and important case with regard to gender equality? Sweden’s national gender equality objectives are unique. The overall policies and discourse include articulations of both radical and feminist viewpoints on power relations between women and men (Government Bill 2005/06:155). The radical policies are reflected in international indexes, as Sweden often enjoys high rankings for gender equality. Sweden comes in fourth place in Gender Gap 2014 (Hausmann, 2014) and first place in the European Institute for Gender Equality’s Index in 2013 (EIGE, 2013), but in the UN’s Index GII (UNDP, 2014) Sweden is ranked only in 12th place. ¹ These rankings indicate that the vision of the women-friendly state (Hernes, 1988) is a national policy being implemented in areas such as employment, life expectancy, maternal mortality, political representation and monetary resources. Sweden’s relatively high position in these rankings can be understood as a combination of the effects of the social-democratic welfare regime (Esping-Andersen, 1990) and the women-friendly state (Hernes, 1988). Feminist criticism has concerned the lack of focus on reproductive work in Esping-Andersen’s theory as well as the role of the state in the progression towards gender equality (Bergqvist et al., 2007; Borchorst and Siim, 2008; Hedlund and Lindberg, 2012; Hernes, 1988; Hobson, 2004; Kantola and Squires, 2012; Sainsbury, 1999; Svensson and Gunnarsson, 2012). I agree with the feminist critique of Esping-Andersen, but still argue that features of the social-democratic welfare regime are what enable the discourses articulating Sweden’s success as an egalitarian state. Finally it is important to stress that in the international rankings, the evaluations measure gender equality, and not the process of applying gender mainstreaming.

Materials

Like other policy research areas, such as studies on entrepreneurship, which also have seen an explosion of published articles, the academic field of gender mainstreaming is now extensive. The political interest in gender mainstreaming has also generated official reports and documents, as well as an extensive grey literature. Grey literature is defined as texts produced by non-governmental organizations or private actors. The time frame of the selected texts stretches from the 1980s to 2014.

The criteria for selecting academic texts for analysis were as follows: the text should be written by an academic scholar in the field of gender mainstreaming, and should contain mention of achieved successes or good examples, or references to Sweden. The method for selection was both systematic and strategic, in the sense that I screened Web of Science, Google Scholar and Örebro University’s library database (SUMMON) for significant texts. The search strings included the words gender mainstreaming, implementation, case, best case, best practice, progress and effects, in combination with each other and with different names of countries.

The category grey literature and mixed texts, henceforth called grey literature, are defined as non-peer reviewed literature. The difference between mixed texts and grey literature is the partial participation of academics in the former. Mixed texts can be co-authored or edited by academics in collaboration with non-academics. In this kind of text, the division of authorship/editorship tasks is not always explicit. The criterion for inclusion of these texts is that they refer to the case of Sweden.

The third category comprises texts from Swedish governmental agencies, such as official reports and documents. I have exhaustively searched the national corpus of evaluations and reports targeting gender mainstreaming and gender equality politics.

Using these categories and criteria, the texts serving as empirical materials in this article are as follows:

It is important to emphasize that the majority of academic articles on gender mainstreaming do not characterize countries as good examples or success stories. But when screening the texts for such cases, a majority of the ones found point to Sweden. I have also found examples where Finland or the Netherlands are put forward as forerunners of gender mainstreaming, but not to the same extent as Sweden. Larger research programmes that focus on gender equality and/or gender mainstreaming, such as MAGEEO and QUING, have not been selected for the empirical material. This is mainly because Sweden was not one of the countries that was empirically investigated. In the case of QUING, I excluded the programme reports, but individual articles related to the programme that were identified during the selection process were included. The more recent work done by EIGE has not been included.

In order to identify and understand how the story of gender mainstreaming in Sweden is constructed, it is vital to explore the themes and descriptions that are embedded in the myths. After identifying the relevant texts, the analytic process continued. The articles that I cite most frequently are those that most clearly and distinctly articulate discourses, however such discourses are present in almost the entire sample. When reading the selected material I searched for recurring themes, descriptions of Sweden and descriptions where different kinds of texts pointed to different empirical results. The themes identified in the readings include: adaptation, integration, volume and initiatives.

Adaptation

In this section I will investigate how the material ascribes to Sweden a leading position in gender mainstreaming. The first element, adaptation, illustrates how gender equality (as a policy objective) is conflated with gender mainstreaming (as a strategy). I identify this as the dominant ‘success discourse’ (Jutta and Schneiker, 2012; Woodford-Berger, 2004; Zebrachi, 2014).

Articulations of Sweden as being special are found in the following statement from an academic text: ‘In effect, the cases studied are, apart from Sweden, all “hybrid” cases of gender mainstreaming’ (Daly, 2005, quoted in Bendl and Schmidt, 2012). This quotation may be characterizing gender mainstreaming as more evolved in Sweden than in other countries, but it also may be expressing the idea that the Swedish version is the correct and true form of gender mainstreaming. This description is also found in grey literature (Silvell, 2000). The articulation of gender mainstreaming in Sweden as being something special, as opposed to other ‘hybrid cases’, will be investigated as a possible myth.

Taking countries as a whole, at least three varieties of a gender mainstreaming approach can be identified. The first, with Sweden as the locus classicus, can be styled an ‘integrated approach’. Gender mainstreaming is employed in a global fashion, whereby responsibility for gender equality is extended to most, if not all, actors involved in public policy and is embedded across institutions in society. (Daly, 2005: 438, emphasis mine)

Sweden is described as the locus classicus of gender mainstreaming by this scholar, who identifies shared responsibility and embeddedness in public institutions as key articulations. This description is reiterated in Wittbom (2011) and in Meier and Celis (2011).

The next example of how Sweden is articulated as special is illustrated by this quotation from an academic text: ‘these countries [Sweden and the Netherlands] can be said to be in the relatively luxurious situation of knowing what gender mainstreaming is, having sufficient gender data and having developed a broad concept of gender equality’ (Woodward, 2003a: 81). This assertion highlights two things: first, the taken-for-granted connection between gender equality and gender mainstreaming (as presented in the following section); and second, the articulation of Sweden (and the Netherlands) as having special insight into the nature, or essence, of gender mainstreaming (Verloo, 2005). The articulation ‘knowing what gender mainstreaming is’ should be understood as a floating signifier displaying the lack of coherent definitions of gender mainstreaming. The construction of Sweden’s special nature is thus articulated in vague and non-specific ways, which is in line with Della Sala’s concept of political myths.

In the international arena of GM, Sweden has always had special relevance. GM has been practised since the 1970s (formally adopted as a central strategy in 1994) and there is strong institutionalization of gender equality policies … . Sweden is seen as having particularly favourable preconditions for GM, including a wide diffusion of egalitarian values, historic dominance of leftist parties, and over 50 percent female representation in the Riksdag (Parliament). (Hankivsky, 2013: 633)

This excerpt from an academic text first classifies Sweden as being of special relevance, then continues with what is presented as a description of gender mainstreaming history. This illustrates the conflation of gender mainstreaming and gender equality discourses, where the institutionalization of gender equality policies is adapted to the implementation of gender mainstreaming in Sweden. This adaptation is also evident in True and Mintrom’s (2001) academic article about transnational diffusion of gender mainstreaming. In their article, Sweden is described as already having been an early- and high-adoption country of gender mainstreaming in 1980–1984 (True and Mintrom, 2001). This confusion about which policy has been institutionalized (gender mainstreaming or gender equality politics) is also present in the second category of texts, grey literature, and is found in an academically edited book published on behalf of the United Nations. The chapter on Sweden is written by a former femocrat with experience from the Swedish government, the European Commission and the Nordic Council of Ministers. She states that ‘it is important to realize that this “model” [gender mainstreaming] is based on a long democratic tradition’ (Åseskog, 2003: 160).

The articulations of this gender mainstreaming history are to be understood as a dislocation of meaning. Gender mainstreaming does not have a long democratic tradition; however gender equality policy does. The earliest mention of gender mainstreaming in Sweden can be traced back to a document from 1987, where the Swedish government argues for the development of a gender-impact assessment tool to be used before making decisions (Government Bill 1987/88:105). On the other hand, however, Sweden has had a well-established gender equality policy in place since the 1960s (Florin and Nilsson, 2000). Thus, the reference in the cited texts refers to a non-specific past, and the myth that gender mainstreaming and gender equality are the same thing is supported and upheld (see e.g. Woodford-Berger, 2004).

Integration

As previously shown, myths are embedded within the discourse of gender mainstreaming when Sweden is discussed in the investigated articles. I now continue by illustrating the perceived success of the institutionalization of gender mainstreaming in state administration at national, regional and local level, as it is articulated in the selected material (Guenther, 2008; Rubery et al., 2004; Zebrachi, 2014). One of the themes in the success discourse is that gender mainstreaming strategies have been established on all levels. This can be illustrated by the following quotation:

Sweden is the exception in that it has in place an entire ‘package’ in the sense of an acceptance of the analysis of gender equality, as well as the integration of the full spectrum of relevant procedures for gender mainstreaming across levels of administration. (Daly, 2005: 440, emphasis mine)

This description underlines the fact that an important indicator of Sweden’s gender mainstreaming success would be for it to be integrated at all levels. However, I have not identified any articulations supporting this in the third category of selected texts, official reports. In contrast to the confident picture in the quotation above, the other categories of texts appear to contain conflicting descriptions. In two governmental reports, a national audit report and a national evaluation report, it is argued that gender mainstreaming in state agencies at the national level in Sweden is not a success story (see below) (Riksrevisionsverket 2000:17; Statskontoret 2005:1). One of these, an audit of how the government has applied its governance mechanisms to gender equality issues, states that only half of all state government agencies are required to report on how gender mainstreaming is being implemented in the agency with reference to gender equality objectives (Riksrevisionsverket 2000:17: 6). The audit further states that government agencies have not received adequate direction regarding the implementation of national gender equality policies from the government (Riksrevisionsverket 2000:17). The government has substantial influence over the agencies, influence which has not been used in the case of gender equality. The same report from the Swedish national audit office shows that of the 55 investigated government agencies, 35 answered that they have some kind of programme to implement gender mainstreaming. This implies that 20 agencies have not made any organized efforts to realize the national strategy of gender mainstreaming (Riksrevisionsverket 2000:17: 45). This lack of successful implementation should be understood as a dislocating articulation that has not been incorporated into the discourse of success. A counter-discourse can be identified in Sweden’s success story about gender mainstreaming.

The articulations above concern how Sweden is governed at national, regional and local level. All levels are involved in the implementation of gender mainstreaming, though in different ways. Another conflict between the success discourse and the counter-discourse concerns the different levels and their associated duties. In Sweden, there are 21 county administrative boards, which function as an extension of the central government authority at the regional level (Bergqvist et al., 2007). A gender equality official serves on each of these boards,. There are also elected regional councils, which govern the policy areas of health care and regional development. When the academic texts state that the gender mainstreaming commitments have been undertaken at the regional level, they are referring to the county administrative boards, not to the regional councils. Neither the audit nor the evaluations (referred to above) describe any tangible development in the regional councils.

The quotations above exemplify how the discourse of ‘success’ disregards counter-discourse articulations in the national audits. This enables the dislocation in the official reports to be smoothed over, and the dominant discourse of success can instead articulate a social imaginary with shared responsibilities and embeddedness in governmental institutions; in other words, the myth can be upheld. Only one single articulation of a counter-discourse occurs in the academic category of texts: ‘Equally sobering is that the task of gender mainstreaming the central administrative agencies and the regional and local administration is still in its infancy’ (Sainsbury and Bergqvist, 2009: 229). This academic articulation is not in line with the other academic texts, in that the authors articulate a discursive perspective that is consistent with the more complex counter-discourse.

Volume

Tuning to the third administrative level in Sweden, the municipalities, a description in the national audit evaluation of gender equality politics and gender mainstreaming states that insufficient information is available about the municipalities’ undertakings. Instead, they argue, there is a lack of information about what the municipalities actually do to improve gender equality (Statskontoret 2005:1). At the same time, the municipalities are described as a cornerstone of the Swedish welfare state.

The relationship between the state and the municipality is rather complex, in that there is extensive self-government at local level. The municipality is independent of the state, but at the same time it is the executor of national objectives as well the main provider of welfare services. The municipalities have great influence over citizens’ lives – when organizing child care and elderly care, as a large employer, as a tax collector, and when setting political priorities – and their activities affect women and men differently. Because Sweden lacks strict laws concerning gender equality, there is quite extensive variation between the 290 different municipalities in Sweden (Jutterdal, 2008; Larsson and Bäck, 2008). There is no agreement about the extent to which municipalities should be compelled to comply with state interventions. Because of their central position, the municipalities have been subjected to ‘soft’ governance through the Swedish Association of Local Authorities and Regions (SALAR). Being an independent employer and lobbying organization for Sweden’s municipalities and regions, and partly dependent on financial support from the government, SALAR has entanglements in the national political arenas. It has played a leading role both in discourse production and in pushing gender mainstreaming into local politics, in two waves, one in the 1990s (JämKom) and one in the late 2000s (PSGE; Programme for Sustainable Gender Equality).

Turning to the next theme in the success discourse, we will engage with the construction of volume, which describes Sweden as having passed the critical point of implementation in municipalities. This can be illustrated by the following articulation from an academic text:

For many years, Sweden has had, in its association of municipalities, large-scale projects on gender equality in local authorities. In connection with mainstreaming, a program called 3-Rs was launched and carried out by municipalities themselves. (Woodward, 2003a: 79)

In the second category of texts (grey literature) we find statements that engage with this theme. A policy process was started in 1995, when the government created expert groups and project groups and gave SALAR funding to improve gender equality policy in Swedish municipalities/cities by implementing the programme JämKom (Svenska Kommunförbundet, 1998). The grey literature presents the programme JämKom as a pilot project that took place in one or two selected boards/committees in six municipalities. The aim was to implement the so-called 3Rs (representation, resources and realia) method of gender mainstreaming (Åseskog, 2003). The articulations found in the final report from the projects (Svenska Kommunförbundet, 1998) are repeated with some variations in several grey texts (Åseskog, 2003; Åström, 2007; Jutterdal, 2008; Lundkvist, 2010). An obvious counter-discourse to the theme of volume would be that the efforts were small scale, only occurring in a few municipalities. Nevertheless, the programme is described in the academic literature as a large-scale municipality project (Woodward, 2003a, 2003b). Possible positive effects and achievements aside, a project involving six out of Sweden’s 290 municipalities cannot really be counted as large-scale. This is another example of how articulations that can be seen as counter-discourses are not incorporated into the dominant discourse. The myth of volume can thereby be preserved.

Initiatives

The next theme in my analysis is categorized as initiatives. I conclude that when Sweden is described as a ‘model country’ in gender mainstreaming, it is the starting-up process that is being referred to. Examples of this can be found in the different categories of texts. Below is an example from the academic literature:

What makes Sweden distinct among all the cases is that in September 2011, the government reconfirmed its commitment to strengthening GM by developing a specific platform designed to cut across central, regional, and local levels. It is composed of: a strategy for GM in the government offices; a development program for government agencies; support for GM at the regional level; quality assurance of the development of GM in municipalities and county councils; and initiatives to gather and share experience and knowledge concerning the practical implementation of GM. (Hankivsky, 2013: 633)

As shown in the quotation above, it is development in the form of initiatives that is articulated in the dominating discourse about Sweden in academic texts. In the category of grey literature, we can find another articulation relating to this theme. In the evaluations following the PSGE (see above section on volume) it is concluded that the gender mainstreaming programme has focused on the initiation of the project and the process of the programme, rather than on its possible effects and contributions towards reaching the national gender equality objectives (Halldén et al., 2011; Svensson et al., 2013). And because the PSGE is located at SALAR, rather than a government agency, the evaluations are limited to the specific project. This is visible in one of the evaluations of the PSGE where the outcomes are expressed in management terms, such as running an ‘orderly and well managed’ project (Svensson et al., 2013). The recurring theme of initiatives can also be identified in a consultant report about the perceived potential of gender mainstreaming at the local and regional level of governmental administration (Åström, 2007). The report mainly deals with how to establish or initiate gender mainstreaming in Swedish municipalities. The focus is on the initial phase, including how to plan activities and strategies.

The theme of initiatives has no counter-discourse, though hypothetically such a discourse could deal with effects, results, or possible even sustainability. But there are several examples in the grey literature category that connote the theme of initiatives and could be described as articulations of implementation critique. In a consultant report describing different local projects since the 1990s, it is said that the projects did not result in the establishment of a lasting method for implementation of gender mainstreaming on the local level. The report also states that there are no evaluations conducted by academic scholars that show how the 3Rs method (representation, resources and realia) has been implemented over time, including evaluations of its effects (Lundkvist, 2010: 11). This counter-discourse is also present in a report from SALAR, which states that many evaluations in Sweden focus on the size of projects and on variables such as participation and participants’ opinions about projects, rather than on the effects of projects. The report further concludes that this kind of evaluation has nothing to say about whether and how any increase in gender equality has been achieved (Jutterdal, 2008: 41). In a report from JÄMI (National Commission for Gender Mainstreaming), the authors instead assert that there has been almost no focus on the effects of gender mainstreaming (Norrbin and Olsson, 2010).

Conclusion

The description of Sweden in international research as a successful case when it comes to its progress in gender mainstreaming is not just a harmless story; rather it is of great significance. Although discourses may be viewed as only abstractions or at least as connoting something innocent, they have consequences for the possibilities to formulate concrete action. Research is authoritative, and abstractions and theories are important co-creators of policy change. The discourse of Sweden’s success in gender mainstreaming can therefore affect future (re)constructions of the welfare system and can influence, or even hinder, real political change. The normative power of gender mainstreaming as a political myth is significant, in that it not only answers questions of ‘how’, but also constitutes a normative map that helps create political legitimacy for the strategy of gender mainstreaming.

The aim of this article has been to investigate the discourse of Sweden in gender mainstreaming texts and to increase the understanding of how recurring themes and descriptions are articulated. The intention is not to ‘correct history’ or to ‘tell the truth’, but rather to point to and investigate the relation between the success discourse and other selected texts that seem to tell a different story. I argue that the conflation of gender mainstreaming (viewed as a strategy) with gender equality (as a policy objective) has been a vital part of the construction of Sweden as the best case of gender mainstreaming. The adaptation of the gender equality discourse to the discourse of gender mainstreaming has societal and political consequences. Rather than arguing for the spread and transfer of the Swedish way of doing gender mainstreaming, one should emphasize other aspects. One such aspect is the Swedish welfare state, which, in line with Esping-Andersen’s (1990) categorization of welfare regimes, can be recognized as a best case worthy of imitation. When the national evaluations create dislocation in regard to the theme of integration, it is evident that the dominant discourse disregards this information in order to uphold the social imaginary of Sweden as a success. I show in the section on volume how the counter-discourse articulated in the grey literature is missing from the academic texts. The articulation of large-scale implementation becomes part of the myth. In the theme of initiatives, articulation of possible effects and outcomes is missing from the texts. I believe that these findings are relevant to how to build and frame gender equality policy recommendations in the future and globally. If aspects of the social-democratic welfare state generate the most conducive conditions for increased gender equality, then public policy has to make that clear in order to deliver effective implementation.

The myth of gender mainstreaming in Sweden can contribute to egalitarian societies. What can be envisioned and conceived as possible is important, and the implications of possible imaginations should be further investigated. There are normative aspects in perceived successful cases. The argument is: Sweden did it! Others will, can and should follow! Building on this analysis, it should be investigated how Sweden is used as an illustration of how learning by example serves as guidance and evidence in the international spread of gender mainstreaming. The implications of such a political myth can be inspirational for others, using it as a political tool to push for progressive social change for greater gender equality.