Twenty years of EU gender mainstreaming: rebirth out of the ashes?

Abstract Twenty Years of EU Gender Mainstreaming: Rebirth out of the Ashes? The article looks at the progressive development of gender mainstreaming as a strategy to achieve gender equality in the EU and the role played by policy makers to implement and/or instrumentalise this new concept. From the outset, bureaucracies seized it as an opportunity to dilute the focus on gender and cut resources, transforming its development into a permanent power struggle. The article starts with a brief overview of the major steps and research findings on GM during its launch in the early 1990s and discusses problems arising from related concepts. It then assesses the backlash directed against gender equality, and the demotion of gender mainstreaming, precipitated by the 2004 Eastern enlargement, and the sovereign-debt turned Euro-crisis after 2008. The article concludes by highlighting issues that suggest gender mainstreaming remains the only effective strategy for tackling gender inequalities in the EU, as long as it is understood and implemented in the “policy transformative way”. The article is based on the authors’ insider view stemming from extensive experience in gender equality policies at the European and national level. ----- Zwanzig Jahre EU Gender Mainstreaming: Wiedergeburt aus der Asche? Zusammenfassung Der Beitrag befasst sich mit der fortlaufenden Entwicklung von Gender Mainstreaming als gleichstellungspolitischer Strategie der EU und der Rolle die dabei politische Entscheidungstrager*innen haben, dieses neue Konzept zu implementieren und/oder zu instrumentalisieren. Von Beginn an wurde Gender Mainstreaming genutzt um den vorherigen Fokus auf Geschlecht auszudunnen und Ressourcen zu streichen, wahrend zeitgleich die Entwicklung von GM sich in einem permanenten Machtkampf transformierte. Der Artikel fasst eingangs die zentralen Schritte und Forschungsergebnisse zur Einfuhrung von Gender Mainstreaming in den fruhen 1990er-Jahren zusammen und diskutiert entstehende Probleme durch verwandte Konzepte. Dann wird die Gegenreaktion zur EU Gleichstellungspolitik und die Herabstufung von Gender Mainstreaming bewertet, die sich im Zuge der 2004er Osterweiterung und der Finanz- und Staatsschuldenkrise seit 2008 entwickelte. Abschliesend wird erlautert, warum Gender Mainstreaming dennoch die einzig effektive Strategie zum Abbau geschlechtsspezifischer Ungleichheiten in der EU bleibt – jedenfalls so lange die Strategie in ihrem ursprunglichen „transformativen Ideal“ verstanden und umgesetzt wird. Der Artikel basiert auf den Innenansichten der beiden Autorinnen als langjahrige Gleichstellungsexpertinnen auf EU und nationaler Ebene. ----- Bibliographie: Hubert, Agnes/Stratigaki, Maria: Twenty Years of EU Gender Mainstreaming: Rebirth out of the Ashes?, Femina Politica, 2-2016, S. 21-36. https://doi.org/10.3224/feminapolitica.v25i2.25350

M o k re , M o n ik a /B o rc h o rs t, A n e tte , 2013: The E U 's G e n d e r and D iv e rs ity P o lic ie s and th e E u ro pean P u b lic S p h e re s . In: S iim , B irte /M o k re , M o n ika (Hg.): N e g o tia tin g G e n d e r and D iv e rs ity in an E m e rg e n t E u ro p e a n P u b lic S p h e re . B a s in g s to k e , 141-160. M u s h a b e n , Joyce M., 2012: W o m e n on th e M ove. EU M ig ra tio n and C itiz e n s h ip P olicy. In: A bels, G a b rie le /M u s h a b e n , Joyce M. (H g.): G e n d e rin g th e E u ro p e a n U n io n . N e w A p p ro a c h e s to Old D e m o c ra tic D e fic its . H o u n d s m ill, B a s in g s to k e , 20 8-227. P o lla c k , M a rk A ./H a fn e r -B u r to n , E m ilie , 2000: M a in s tre a m in g G e n d e r in th e E u ro p e a n U nion.
H a rv a rd Jea n M o n n e t W o rk in g P a p e r 2. C a m b rid g e . P rü g l, E lis a b e th , 2012: The C o m m o n A g r ic u ltu r a l P olicy and G e n d e r E q u a lity. In: A b e ls, G a b rie le / M u s h a b e n , Jo yce M. (H g.): G e n d e rin g th e E u ro p e a n U nio n . S tie g le r, B a rb a ra , 2000: W ie G e n d e r in den M a in s tre a m k o m m t. K onze pte, A rg u m e n te und P ra x is b e is p ie le z u r E U -S tra te g ie des G e n d e r-M a in s tre a m in g . B onn.
V le u te n , A n n a van de r, 2007: The P ric e o f G e n d e r E q u a lity. M e m b e r S ta te s and G o ve rn a n ce in th e E u ro p e a n U nion. F a rn h a m . W e in e r, E la in e /M a cR a e , H e a th e r, 2014: The P e rs is te n t In v is ib ility o f G e n d e r in EU P olicy. In tr o d u c tio n . In: W e in e r, E la in e /M a c R a e , H e a th e r (Eds.): The P e rs is te n t In v is ib ility o f G e n d e r in EU P olicy. E u ro p e a n In te g ra tio n O n lin e P a p e rs. S p e c ia l is s u e 1, vol. 18, a r tic le 3, 1-20. W o o d w a rd , A lis o n E., 2003: E u ro p e a n G e n d e r M a in s tre a m in g . P ro m is e s and P itfa lls o f T r a n s fo r m a tiv e P olicy. In: R eview o f P olicy R e s e a rc h . 20 (1), 65 -88.

Twenty Years of EU Gender Mainstreaming: Rebirth out of the Ashes?1 A G N È S H U B E R T . M A R IA S T R A T IG A K I
In tro d u c tio n When gender mainstreaming (GM) first appeared in the text of the Third Medium Term Community Action Programme for equal opportunities between women and men 1991-19952, it was seen as an instrument for expanding and strengthening the reach of EU gender equality policies. GM gradually became recognised as a central, innovative policy tool, particularly in the wake of the United Nation's Fourth World Conference on Women in Beijing in 1995 where the EU delegation had insisted on adding a sentence concerning its importance in all twelve areas of concern found in the Platform for Action.3 Support for its formal inclusion in the EU's 1997 Amster dam Treaty (Articles 2 and 3) was actively mobilised by feminists who demonstrated under the flag of the European Women's Lobby in the streets of Amsterdam. Since then, GM has become one of the most widely discussed, praised and criticised con cepts in feminist academic literature on EU gender equality policy (see among other Liebert 2002;Lombardo/Meier 2006;Rees 1998;Squires 2005;Verloo 2001Verloo , 2005aVerloo , 2005bWalby 2005;Woodward 2003b, Woodward 2008. This burgeoning academic interest increased not only the attention paid to gender equality but also working knowledge of the EU policy machinery, leading to a criti cal re-evaluation of many Community policies from a gender perspective. Exam ples include macroeconomic policies (Hoskyns 2004), regional policy (Braithwaite 2000), and assessing the European constitution processes (Lombardo 2005 (EIGE 2014). With few exceptions (Jacquot 2015), the common theme linking these works has been recurrent praise for the transformative potential of the gender mainstreaming concept, coupled with criticism regarding the lack of effective implementation. As EU insiders, we will attempt in this article to pinpoint the opportunities inherent in GM, sources of resistance to it, the major implementation hurdles it has encountered, and reasons why it has yet to realise its transformative potential in European policy making. We use the results of this investigation to discuss the ways in which we think gender mainstreaming can be used as a major strategy in the future. We argue that even though the evolution of EU gender equality policies and the use of gender mainstreaming brought major progress after 1995, EU actors failed to recognize its value as a transversal, transformative strategy for use beyond the traditional social, employment and human rights spheres. In this context GM became vulnerable, and thus easily undermined (one could say: attacked) in conjunction with two major political-economic events that caused gender equality to lose its status as the "leading thread"5, reducing it to a minor part of the anti-discrimination agenda, namely, the EU enlargement processes since 2004 and the economic crisis since 2008. We build on Stratigaki's earlier argument (2005) that GM deviated from its trans formative course when attempts were made to use it as a policy strategy "countering positive (affirmative) action" during the first decade of its existence. Despite this deviation, however, gender equality policies remained visible on the policy agenda. We argue further that GM was eventually instrumentalised to demote gender equality as an EU policy priority, justifying its dilution by allowing member states to empha size other policy priorities. There are nonetheless good reasons to believe that abuse of the concept has not rendered it obsolete. Given its transformative potential, our goal is to highlight the opportunities for full implementation so that this promising instrument remains a source of feminist power. Our arguments rely largely upon our experiences as feminists and with feminists working in European institutions (Hubert) and/or in policy-making bodies at the national level (Stratigaki) during the crucial period, 2005-2015. We analyze EU doc uments and other inside sources of information through a gendered lens to com plement existing academic analyses. We add our understanding of the internal in stitutional conflicts and power struggles seen among diverse policy actors, political interests and pressure groups. Our argument is supported by key EU strategic doc uments and developments of the last decade, such as the Lisbon Strategy, Europe 2020, the Annual Growth Survey, the European Semester, Country-Specific Recom mendations, Multiannual Financial Framework Programmes and political guidelines issued by European Commission Presidents Jose Manuel Barroso (2004Barroso ( -2014 and Jean Claude Juncker (since 2015). We start with a brief review of GM in the early 1990s and discuss problems arising from related concepts. We then assess the backlash directed against gender equal ity, and the demotion of gender mainstreaming, precipitated by the 2004 Eastern enlargement, and the sovereign-debt turned Euro-crisis after 2008. We conclude by highlighting issues that suggest gender mainstreaming is actually the only effective strategy for tackling gender inequalities in the EU, as long as it is conceived and implemented in the "policy transformative way".
C o n flicts in la u n ch in g w ith u n p re d ic ta b le re s u lts Two aspects that did not attract much attention in the scholarly literature during the first phase of GM development in the EU (1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005) were, first, the context in which GM entered EU discourse, and the unpredictable results that emerged due to conflicting power relations accompanying its implementation. The second pertains to the ways in which GM terminology has featured in both generating opportunity and resistance to the concept.

Context of GM appearance
When gender mainstreaming appeared first in an EU policy document in the early 1990s, gender equality seemed to have reached a stand-still. Only non-binding EU recommendations on positive action were adopted. Their take-up in the Second Eu ropean Action Programme for Equality between Women andMen (1986-1990) (CEC 1986), extending incentives to companies, has been marginal, limited to training and childcare provisions. Meanwhile, pressure for progress was growing. Women's la bour market participation increased, the business world called governments to make it easier for women to participate fully in the labour market. Also, the 1989 creation of the European Women's Lobby offered the women's movement representation at the EU level. The fledgling Committee on Women's Rights and Gender Equality (FEMM) of the European Parliament (EP) (consulted on the Second Action Pro gramme) stated that equality would not be achieved as long as broader issues such as the status of women in society, gender stereotyping and the lack of women in decision-making positions were not addressed. Gender mainstreaming seemed to offer the ideal tool to support this type of policy expansion. In this context, conceptualizing and implementing GM became part of the general ongoing conflict over power relations in EU policy-making, involving institutions, national priorities, interest groups, etc. In 1991 the European Parliament (EP) Women's Rights Committee pressured for a sentence mentioning GM in the Third Action Programme. In 1995 similar pressure on the European Commission occurred when the new member state Sweden demanded to the EU in Beijing to include a mainstreaming provision in every chapter of the Platform for Action.6 Later in 1995, Anita Gradin, the newly appointed Swedish Commissioner, claimed the gender equality portfolio during the first informal meeting of the College of Commissioners. The conflictual debate this unleashed in the College triggered a series of adminis trative developments, such as the creation of a group of Commissioners for gender equality, which existed for twenty years. Others entailed increasing the gender equality budget, initiatives against the trafficking of women and girls, the DAPHNE Programme against gender violence, the formation of a High-level Group on Gender Equality for the Structural Funds, the introduction of a Women and Science policy, and new or enhanced directives for gender equality in the labour market and beyond. The new gender mainstreaming activities, combined with women's presence in pow erful positions (Commissioners Monica Wulf-Mathies, Edith Cresson, Anna Diamantopoulou; Erin Mac Nally in the EP Budget Committee) proved very decisive in producing a wave of positive gender equality developments in the late 1990s and early 2000s.

Contesting GM terminology
The ambiguous nature of the term "gender mainstreaming" played a crucial role in these developments, as experienced inside the institutions. Both "gender" and "mainstreaming" have evoked more questions, confusion and misunderstanding than any other term in EU equality policies (Stratigaki 2005). Vocabulary disputes and translation issues throughout various member states have intertwined with power play and political instrumentalisation (Jacquot 2015). At first, the concept was called "mainstreaming". Adding the term "gender" raised other controversies during the Beijing Conference, eventually resulting in "gender mainstreaming". The "mainstreaming of equal opportunities", first introduced in the Third Action Programme, initially meant widening the scope of equality policies "for women". Equal pay for work of equal value was seen to require action in other domains including women's role in decision-making, and their status in society. Ex isting equality policies often produced inadequate results at a time when women were sorely needed in the labour market. Both the feminist objective of women's economic independence and the utilitarian goals of the economy were forces leading to an extension of the scope of gender equality. The European Commission eventually adopted a lighter version, understanding GM as "a new partnership between women and men" (Stratigaki 2005, 175). Its ability to realize its transformational potential in terms of EU policy became further com promised given the confusion and conflict around the word "gender" which emerged in the wake of the Beijing conference. Feminists welcomed the term, insofar as it recognized the socially constructed nature of sex inequalities. Most policy makers, however, construed it as a way of adding "men" to equal opportunities and equal treatment policies. Adding to the confusion, "gender" was often impossible to trans late into other EU languages giving the impression that, when used in English, it was a concept that did not belong there. Its use in the Latin languages became associated with an anti-traditional family message and implied (sexual and homosexual) immo rality thanks to the co-optation of the term by religious circles. The profound nature of the conflict surrounding the word "gender" has resurfaced in recent public debates on same sex marriage, in which opponents have rejected the term in order to denounce the blurring of sexual identities. Some positions espoused by the Holy See are reminiscent of the vehement exchanges over the definition of the term that took place in Beijing in 1995 between the Vatican spokesman, the Iranian representative of the Khomeini Government and Cristina Alberdi, the Span ish minister for social affairs who held the EU Council presidency at the time. The conflict abated with the arrival of Irish EU commissioner Padraig Flynn and French director Odile Quintin, both faithful Catholic believers. What they said in Beijing remains confidential, but they subsequently made sure that the term "gender" did not challenge the traditional division of gender roles in EU policy documents. The 1996 Communication of the Commission on gender mainstreaming, written under the tight control of Quintin is a telling example (Stratigaki 2005). The Catholic Church's resistance to the notion of socially constructed gender persists and it avidly follows EU legislation. This demonstrates the revolutionary potential of the term "gender" which points to socially constructed male domination. The problematic connotations of the term "gender mainstreaming" make it hard to offer a definition, simple enough to be understood by policy makers inclined to be dismissive of gender equality. Indeed, no definition has ever been formally accepted at the EU level. The Commission uses the definition of its 1996 Communication that is limited to a vague "integration" objective referring to "women and men" instead of "gender".7 The EP prefers the definition advanced by the Council of Europe, pri oritizing GM's concrete transformative policy goals: "the (re)organization, impro vement, development and evaluation of policy processes, so that a gender equality perspective is incorporated in all policies, at all levels and at all stages by the actors normally involved in policy making" (Council of Europe 1998). Drawing on multiple interviews conducted with policy makers in Brussels, Sophie Jacquot (2015) identified four types of (mis)understandings associated with gender mainstreaming. The first draws on the extensive conceptualization utilized by mem bers of the "velvet triangle" i.e. policy makers, the research community and civil society (Woodward 2003 a) for whom the challenge is to deploy it as an effective instrument for equality. Next there is the minimalist conceptualization, upheld by a majority of policy makers who see GM as a "politically correct" idea that does not require much effort, as if to imply that "gender equality (already) exists". Third, those who want to maintain control over their own areas of competence, by denying the relevance of a gender perspective in their (supposedly neutral) sector, adhere to a defensive conceptualization of the term. Last but not least, there are those who em ploy a conservative conceptualization, drawing on value based arguments to dem onstrate the irrelevance of gender mainstreaming for their sector. The only "rele vant" differences between women and men, in their eyes, concern traditional gender roles (e.g., the safety of cosmetics, different toys for girls and boys; product labels for women consumers).8

EU E n la rg e m e n t v e rs u s G ender E q ua lity
The EU enlargement in 2004 to Central East European states implied serious budget ary restrictions and administrative reforms for gender equality policies. The Com mission construed the goal of gender equality, first, as part of a larger anti-discrimi nation objective and, secondly, as a matter of fundamental rights. This tendency to increasingly merge gender equality with other inequalities resulted in merging insti tutional and administrative structures in the EU with the exception of the European Institute for Gender Equality (EIGE) which has retained its gender specificity since its foundation in 2006 (Hubert/Stratigaki 2011).

Budget restrictions
The resource scarcity currently afflicting most public administrations began before enlargement and the economic crisis. The onset of the neo-liberal era in the early 1980s was marked by a constant vilification of taxes and high praise for privatisation of public services. It has seen many variations leading to the reconfiguration of pub lic administrations, reduced spending and reassessed national priorities. The neo-liberal impact on the EU budget amounted to a major "recast" exercise, which merged some budget lines. The specific budget line reserved for gender equal ity policy in the 1990s, for example, was merged into the PROGRESS programme (2000), managed by the Directorate General for Employment (DG EMPL) based on the argument that gender should be mainstreamed into social and employment policy.9 The EQUAL Community initiative (2000)(2001)(2002)(2003)(2004)(2005)(2006) replaced NOW (New Op portunities for Women) (1994)(1995)(1996)(1997)(1998)(1999) in the European Social Fund (ESF). Eventually even EQUAL was replaced by a small amount of funding to implement GM as of 2007. In 2010 the specific budget line inside PROGRESS10 was moved into the Rights, Equality and Citizenship Programme managed by the Directorate General for Justice.11 Since then, and including the present ESF programming period 2014 2020, financial resources available for the specific pursuit of gender equality can no longer be identified. The "budgetary fence" protecting expenditures for gender equality in the EU structural funds has disappeared, as have funds for genderspecific pilot programmes. A tentative estimate regarding all EU funding for equality promotion shows that the financial resources for gender equality rose from €70 million in 1991 to €220 mil lion in 1999, falling below €50 million in 2012; this accounted for 0.20% of the EU budget in 1999, dropping to a mere 0.03% in 2012 (Jacquot 2015). In the "Strategic Engagement for Gender Equality 2014-2020", it is estimated that €5.85 billion of the European Structural and Investment Funds will be spent on measures promoting gender equality in the period 2014-2020. This amounts to 0.06% of the EU budget. On the positive side, enlargement has been accompanied by an extended commit ment to "promote gender equality and eliminate discrimination" in "all" EU policy areas, as codified in the Lisbon Treaty, in contrast to the community policies formerly mentioned in Article 3 of the Amsterdam Treaty. This major extension, along with the proclaimed "deepening" of EU policy, suggested a need for even greater human and financial resources in the area of gender equality policy. However, the common belief among EU policy makers was that, in contrast to positive action measures, gender mainstreaming was a simple matter of "reorganisation" that requires no addi tional investment, serving the argument for diminishing the gender equality budget. This position, however, contradicted the gender mainstreaming guidelines of the Commission, which recommended that policy actors "provide an adequate budget to cover the costs of gender mainstreaming". The guidelines provided examples of possible additional expenses: "You may need to buy in outside expertise, organise awareness raising and gender training, develop gender analysis tools, organise the consultation and participation of interested parties, ensure monitoring and evalua tion and measuring the results of the impact" (European Commission 2005).

Weakened gender institutions through administrative reform
Administrative reforms occurring in the wake of EU enlargement have been im plemented differently, depending on the policy area, the officials involved and the level of governance. With few exceptions12 gender equality institutions and pro grams were hit hard, resulting "in a reduction in the existing institutional capacity for gender equality policies and a tendency to view gender equality as a human right requiring only legal, judicial measures that address discrimination at an individual level"(Aseskog 2015, 67). The "streamlining" process began with threats to the existence of the Women's Rights Committee of the EP. The European Commission itself went through a covert process of administrative shifting, which included dismantling certain gender-spe cific structures. The most significant change in this regard was the "neutralisation" (from 2005 onwards) and eventual elimination (as of 2010) of the Commissioners' Group on Gender Equality, which had played a decisive role since 1995 in stimula ting political interest and creating momentum for gender policies in fields beyond employment (i.e. research, structural funds, development and cooperation, agricul ture). The Interservice Group that had served the Commissioner's Group is the only remaining internal coordination tool for GM. Originally directors-general attended its meetings, but the level and the mandate of its participants has been consistently reduced. The post-enlargement period also brought the transfer of the Gender Equality Unit from DG EMPL to DG JUST that implicitly shifted EU priorities from gender main streaming EU policies to the human rights agenda, including gender based violence. The second Barroso Commission (2010-2015) offered no functional or theoretical justification for shifting this Unit from its historical and legally stronger home in DG EMPL to the rather smaller scope portfolio of justice accorded to Commissioner Vivianne Reding. The Gender Equality Unit became part of a newly created anti discrimination directorate in DG JUST. The risk here was concentrating on tackling individual cases, thereby undermining the transversal and structural character of the EU gender equality policy agenda. These administrative developments all converged weakening the institutional setting for gender equality in the European Commission. Our (insider) assessment of Reding's five-year record is not positive. There was little action to further the original GM agenda of moving gender equality concerns into all EU policies. Instead there was backpedaling, for example, when the Annual Gender Equality Report (produced since 1996) was absorbed into the Fundamental Rights Report. Reding's main achievements consisted of prioritizing a campaign against Female Genital Mutilation (EC 2013) and proposing a directive on Women on Boards in private enterprises13. While these issues were controversial enough to draw media attention, they were too narrow to become part of a GM transformational agenda. Overall, the political and administrative legitimacy that gender mainstreaming ac quired during its first decade has vanished over the last ten years under the pressure to achieve economies of scale following enlargement. Administrative reform pro vided the space for political priorities and personal strategies which were hostile to gender equality but did not have the chance to become operational before then. In fact, the effort to anchor the dual approach of gender mainstreaming and positive ac tions into the EU gender equality doxa (Commission of the European Communities 1998) supported by the strong political will of the Commissioner in charge, Anna Diamantopoulou (1999Diamantopoulou ( -200414 had "protected" gender mainstreaming's strategic potential successfully in the EU policy agenda. In the post enlargement period, one more negative factor was that in contrast to the 1995 enlargement (Austria, Finland and Sweden), none of the new EU members pos sessed a progressive understanding of gender equality. It is no coincidence that since 2004, gender policy documents have been repeatedly rejected by the Council, cul minating with the Maternity Directive that was eventually withdrawn by the Com mission due to a lack of agreement. The new member states are not the only ones rejecting gender proposals. Enlargement has created a downward spiral in which a lot of "old" member states have found significant allies, bolstering their reluctance to advance gender equality policies.
G ender m a in s tre a m in g in th e e con om ic c ris is Many factors have converged, rendering the first ten years of the century the most "anti social policy" years in EU history: The triumph of neoliberalism favoured high levels of employment building on low-paid workers, mainly women, while enlarge ment brought a wave of ultraliberal enthusiasm as a reaction against communism. Damaging EU referenda in France and the Netherlands weakened the European Commission. Abandoning the "shared competences" on social policy foreseen in the Treaty, the few "social and employment" documents adopted (e.g. the Renewed Social Agenda, COM (2008) 0412) reaffirmed that the main responsibility for social policy belonged to the Member States. The financial crisis in September 2008 could have been a wake-up call in this regard, but fiscal consolidation rapidly gained absolute priority on the EU agenda. In the em ployment and social policy field, unemployed auto workers, manufacturing closures, and overextended banks directed attention and money towards men15. The lack of pri ority accorded gender equality was paradoxically enhanced by increasing inequality between women. Highly educated women had seen their possibilities improved, such that gender equality was no longer an issue for many. Meanwhile, underprivileged women and single mothers were left to struggle with the consequences of the crisis.
The Commission spent much of its energy on designing economic and fiscal gover nance processes to monitor member-state budget balances. Social proposals were de layed or sent back. Bettio et al. (2012) and Karamessini/Rubery (2013) extensively documented women's worsening situation under the crisis. During that period, the words women and gender disappeared from the major EU policy documents. Even Spain's insistence (as Council President in 2010) on inserting the word women into the Europe 2020 strategy employment targets, was only partially successful. The strategic objective was to increase the overall employment rate. The failure to dif ferentiate between male and female rates meant that increasing male employment figures could serve as a cover for a stagnating or declining female employment, increasing gender inequality in the labour market. The major documents used to monitor post-crisis trends such as the Annual Growth Survey, the European Semester and the National Reform Programmes paid little attention to the Treaty's gender mainstreaming provisions. The Gender Equality Unit efforts to integrate gender issues were rarely supported by the rest of the Com mission. Some member states were occasionally asked to improve their childcare facilities or to promote more "flexible employment" to allow for the reconciliation of work and family. Posing no challenge to gender roles, the use of these two policy issues was as far as the Commission seemed able to go in addressing women's unem ployment (EC 2015b). Reversing commitments made in previous policy documents to promote more sharing of family and work responsibilities, this period jettisoned the transformative potential of gender mainstreaming. The UN Millennium Goals Agenda, along with the EU Social Investment Package (approved in 2013), slightly revived policy makers' interest in gender equality. A few of the financing programmes included in the 2014-2020 Multiannual Finan cial Framework (HORIZON 2020, EASI, Structural funds) used the terms gender or women. The references were nonetheless cautious, producing a very limited impact as streamlining took its toll: for instance, EU technical support for gender mainstreaming in policies and projects funded by the European Social Fund (2007 2013) at national level merged with other horizontal ESF priorities in the following programming period 2014-202016.
G ender m a in s tre a m in g : too hard to die? Twenty years after the 1995 Beijing Conference, when GM first appeared as an alternative strategy, there are still many reasons to believe that a broader, deeper approach is necessary if the EU is to honor the equality commitments it has made across the Treaties. At present, gender inequalities are stable or on the rise within the EU member states (EC 2015c) amidst other mounting, crisis induced inequalities (OECD 2011; Karamessini/Rubery 2013). In 2016, it is still possible to have an all male government (in 3 EU Member States), and despite higher numbers of women in leadership positions we are far from equal political and economic representation.
Violence against women is still widespread and unduly tolerated, with drastic con sequences. The overall difference in all areas of life from money to health between women and men was at 54 on a scale of 100 in 2012 according the Gender Equality Index of EIGE17. This means that the level of gender equality is still only half way there after decades of policies to "promote gender equality and eliminate inequali ties". At the Commission level, it appears that gender equality has only superficially or temporarily influenced the administrative culture and decision-making processes. The recently issued "Strategic Engagement for Gender Equality" (2015-2019) (2015 a) of the Commission contains interesting benchmarks, such as calling for a report on gender mainstreaming in the Commission in 2017, as well as for an internal report on the Strategy's implementation. However, the text is not a Communication to the EP and the Council. It has the lowest status as an internal document issued by the Commission services without approval by the College of Commissioners. It contains no binding provisions or requests for member state commitments, and uses conditional tense. Even the Council criticised the Commission's very low-key approach. The negative assessment of the current situation and the failure to achieve progress has reached such a level that it is now censured by relevant European organisations (European Women's Lobby, European Trade Union Confederation) and by the Eu ropean Parliament.
A wake-up call leading to a rebirth of GM?
Along with countless other researchers, we believe that the "limited impact of gender mainstreaming on EU policy outputs, therefore, reflects not an inherent flaw of the mainstreaming concept, but rather the Commission's choice to rely almost exclusively on soft incentives in implementation" (Hafner-Burton/Pollack 2008, 0). It is also argued that it reflects the choice of an implicitly integrationist approach rather than a fundamental rethinking of policy making from a gender perspective (Verloo 2001). One major evidence of feminist efforts to change practices and poli cies at the EU level is the collective knowledge acquired across Europe, often on a smaller scale of good practices and projects that can have a big impact on the design of major policy tools and methods.18 Twenty years of experience in implementing, theorising and analyzing gender main streaming have provided lessons and ideas as to how this "transformative tool" could become valuable for gender equality policy in the EU and the member states. In this last part of the article we highlight the most important assets that allow us to argue that GM is not yet dead and that it is worth investing in it as a major policy making strategy for gender equality.
First, there is a strong legal and policy framework fo r gender equality, resting on the foundation of gender mainstreaming and positive actions embedded in the exist ing treaties. Despite the fluid meaning of gender mainstreaming found in policy documents and objectives, the legal basis remains a source of legitimacy. Given the right circumstances, it can function as a source of human and financial support for appropriate policies. One example derives from the ways in which Article 3 of the Amsterdam Treaty offered a basis for the Communications of the Commission and Conclusions of the Council which formalised a women and science policy in DG Research in the late 1990s (Hubert 2010; Mergaert 2012). Secondly, there is more than enough expertise, with relevant structures rooted in EU and member state institutions that can be harnessed for the purpose of further ing gender equality and gender mainstreaming (EIGE 2014). Knowledge, argu ments, tools and mechanisms have been refined over a span of twenty years, and many mainstreaming processes (often using gender budgeting as a central tool) have been implemented at the regional, local, national and EU levels in pursuit of equality. Thirdly, gender-analysis has proven that equality policies must cover a whole range of policy arenas (health, IT, space, environment, etc.) and not only economic and social domains. It is crucial that gender analysis in all European policy fields be sup ported by gender impact assessments and gender budgeting, so that policy makers are aware of available resources and their distribution by gender. Since the gender budgeting feasibility study commissioned in 200819, the Commission has taken very timid steps to integrate gender into its policy processes but sufficient examples and methodology are now available if the college were willing. Fourthly, EU policies influence and, in some areas, shape national policies through binding measures, incentives and the exchange of good practices.20 It is therefore crucial to continue evaluating the impact of EU GM strategies at the national level. When GM is linked to eligibility for EU funding, its impact has to be acknowledged, evaluated and credited to EU policies. In some cases, "pressure" from EU institu tions may have lapsed, especially when economic and fiscal priorities predominate, but this, too, could change.21 Finally, one additional strength of EU gender equality policy has been its ability to maintain close links with society. The social "use" of the EU legal basis continues to depend on the links between collective action and the institutional framework (control mechanisms, etc.). Collective action relies on vibrant social (feminist and women' s) movements, while institutional control mechanisms rely on political will and diligent administration. The role of EU-funded umbrella associations like the European Women's Lobby remains a key factor for mobilising women's organisa tions behind EU policymaking.

C onclusion
In conclusion, we argue that even though GM has been hindered by a political con text dominated by enlargement and the economic crisis, on a back drop of definitio nal problems, administrative reforms and budgetary restrictions, the dual approachan EU policy characteristic -remains a firmly embedded, well documented strategy for advancing gender equality in relevant policy domains. Introduced in the 1990s, gender mainstreaming offers a revolutionary approach to gender equality. Departing from older efforts to "compensate the disadvantages suffered by women on the la bour market", gender mainstreaming (along with the concept of parity democracy from the same period) opened policy makers to an approach challenging the social construction of gender roles in all policies. It shifted the focus from injecting women into an androcentric world to an understanding of gender inequality as a structural problem upheld by institutions reproducing the subordination of women (Alfama Guillen 2015). While it may sound bureaucratic, gender mainstreaming is a highly "political exer cise" that has impelled the Commission to move a long way through a mix of inter nal tensions and opportunities. The backlash against gender equality experienced in most fields over the last ten years has not undermined the legitimacy of EU equality policy; gender mainstreaming and positive action have been implemented at national and other levels with some degree of success. The mobilisation of administrative actors and civil society is an asset, but generating the momentum necessary for this transformative policy requires explicit political leadership. As the main EU actor, the European Commission still possesses the ex clusive right of initiative according to the Treaties, and thus remains the main driver. Fighting member state resistance to gender mainstreaming and creating a more in clusive Europe go hand in hand. a c c o rd in g to P re s id e n t o f th e E u ro p e a n C o m m is s io n J a c q u e s S a n te r (in a u g u ra l sp e e ch in the E u ro p e a n P a rlia m e n t in 1995).

6
The C o m m is s io n u n d e rs to o d th a t fo r S w e d e n g e n d e r e q u a lity w a s a q u e s tio n o f n a tio n a l id e n tity on th e in te r n a tio n a l scen e. E u ro p e a n C o m m is s io n , 2015a: S tra te g ic E n g a g e m e n t fo r G e n d e r E q u a lity 2016-2 019 . SWD (2015) 278. J a c q u o t, S oph ie , 2015: T r a n s fo rm a tio n s in EU G e n d e r E qu a lity. F ro m E m e rg e n c e to D is m a n tlin g . M e rg a e rt, L u t/L o m b a rd o , E m m a n u e la , 2014: R e s is ta n c e to Im p le m e n tin g G e n d e r M a in s tre a m in g in EU R e s e a rc h P olicy. In: W e in e r, E la in e /M a c R a e , H e a th e r (Eds.): The P e rs is te n t In v is ib ility o f G e n d e r in EU P olicy. E u ro p e a n In te g ra tio n O n lin e P a p e rs. S p e c ia l issu e 1, vol. 18, a r tic le 5, 1-20.