Who learns what from whom? Implementing gender mainstreaming in multi-level settings

Introduction

Gender equality policies are increasingly engaged with, developed and operationalized in a multi-level environment (Haussman et al., 2010; Outshoorn and Kantola, 2007). This is particularly true in the case of gender mainstreaming, which is expected to bring the inclusion of a gender perspective to all policies at all levels. In the EU, this mandate embraces a plurality of realms from the local to the supranational. This adds a greater level of complexity to a strategy that is often considered to be either ‘lost in translation’, subject to strong resistance (Hafner-Burton and Pollack, 2009; Mergaert and Lombardo, 2014), or simply a failure of implementation (Fraisse, 2008; Moser, 2005; Perrons, 2005; Schmidt, 2005), despite multi-level settings offering opportunities for development.

Knowledge on how to tackle gender inequalities represents an important asset for an effective implementation of gender mainstreaming (Hankivski, 2008; Squires, 2007). The EU polity appears to hold great potential to facilitate policy transfer; meaning that know-how travels through time and space (Dolowitz and Marsh, 1996). In the case of gender mainstreaming – and gender equality policies more broadly – this has been captured mainly by two key research agendas within feminist scholarship. First, studies have revealed the importance of Europeanization processes by looking at how domestic actors learn from EU policies and adopt EU ‘ways of doing things’ (Krizsán and Popa, 2010; Lombardo and Forest, 2012). Second, it has been noted that in compound states, this same replication of policies occurs between regions, leading to a ‘race-to-the-top effect’ (Celis and Meier, 2007, 2011), with policy-makers increasingly ‘trapped’ in these complex institutional settings. Thus, feminist analyses either focus on the Europeanization of state-level policies, or are restricted to the internal dynamics of specific federations. Consequently, such studies frequently miss the intricate ways in which policies travel within multi-layered settings.

This article seeks to address this gap by revisiting the classic question of who learns what from whom, developed in the literature on policy transfer (Dolowitz and Marsh, 1996), and applied to the implementation of gender mainstreaming. The case of the Spanish regions offers a best-case scenario to do so, as they have long been embedded in a highly complex institutional framework subject to EU influence and strong territorial dynamics. Moreover, gender mainstreaming features a level of implementation that allows for the exploration of whether or not actions developed in the so-called Autonomous Communities in Spain were ‘imported’ from elsewhere. In scrutinizing such processes, the key questions are: do regional policy-makers learn from different layers of government on how to mainstream gender equality? Does the EU multi-level setting provide incentives and/or knowledge to do so? Does this framework ease the implementation of gender mainstreaming at the Spanish regional level?

Research policies offer a rich source of persuasive empirical material from the EU, ¹ the Spanish government, as well as some Spanish Autonomous Communities that have been involved in promoting gender equality in science. Since 1998, the EU has sponsored several specific studies, launched institutional mechanisms and revised major policy documents to include gender-sensitive measures (Hafner-Burton and Pollack, 2009; Mergaert, 2012). Spain replicated the EU approach by setting up a Women and Science Unit in 2005. Similarly, key policy documents were modified to show concern over gender inequality in science. Starting with Catalonia, other Spanish Autonomous Communities have followed suit by transversally including a gender perspective in a number of research policies.

The multi-level commitment to gender equality in science gives three unique case studies with which to illustrate policy transfer: the regions of Galicia, the Basque Country and the Balearic Islands, which have a variety of factors in common. ² Galician and Basque research policies have incorporated a gender perspective since 2007, whereas Balearic policies remain gender blind; providing variation in our dependent variable. The time window from 2007 to 2010 for Galicia and the Basque Country, and from 2009 to 2012 for the Balearic Islands, allowed us to explore the role of the EU, along with national and regional precedents, in facilitating conditions for policy change. In order to account for policy transfer processes, the views of those involved in the making of regional research policies were recorded. First, those in charge of the Regional Ministries of Research and Development were considered critical actors and paramount to explaining why policies are imported from elsewhere. Additionally, the perceptions of the officials involved in the regional equality machinery and of women from the regional organizations oriented towards gender equality in science were also important. In 2011 and 2012, 31 semi-structured interviews were conducted with these three sets of actors in the three regions analysed. ³ Similarly, over 60 policy documents were assessed to determine if and how a gender perspective was effectively included in research policies. ⁴

The following section provides a consideration of the policy transfer literature, emphasizing the impact of Europeanization processes and domestic territorial dynamics. Then the third section in the article describes the development of equality policies and gender mainstreaming in Spain, reflecting how the transference of policies has impacted that process. The fourth section goes on to consider the implementation of gender mainstreaming by focusing on research policies in Galicia, the Basque Country and the Balearic Islands. It reveals how policies travel through the EU polity and how the multi-level setting in which they are inserted influences actors. Finally, the article concludes with a discussion on the need for a multi-level explanatory framework to analyse the practice of gender mainstreaming.

Policy transfer in multi-level settings: An opportunity for gender mainstreaming implementation?

The current ‘institutionalist turn’ in the feminist research agendas has paved the way for a renewed emphasis on the relevance of downloaded and uploaded powers (Krook and Mackay, 2011). In complex polities like the EU, this has resulted in increased concern in the literature on Europeanization and federalism with processes of policy transfer. We note the extent to which policy knowledge travels, either through coercive transfer or lesson drawing (Dolowitz and Marsh, 1996), and consider how far this process has contributed to the development of gender equality policies.

The Europeanization agenda has been addressing the impact of the EU legal framework for a long time. There is general consensus regarding how EU equality policies have helped promote gender equality in member states (Kantola, 2010). Initially, analyses were devoted to demonstrating the extent to which EU binding norms affected national policies. In line with mainstream literature, a top-down approach was adopted to track the transposition of EU equality directives and account for country-specific responses to acquis (Caporaso and Jupille, 2001; Liebert, 2003). More recent contributions, however, align with emerging debates calling for a bottom-up and actor-sensitive approach (Exadaktylos and Radaelli, 2012). Thus, in reaction to the growing relevance of soft law within the EU (Open Method of Coordination), attention is shifting to ‘what actors make’ of EU policies (Radaelli and Pasquier, 2008), even when there is no pressure for adaptation. Learning from above or ‘using’ EU discourse to legitimize national and regional policies have thus become new ways of capturing the EU impact (Ladrech, 2010; Woll and Jacquot, 2010). The EU itself is seen more as a ‘policy transfer platform’ (Radaelli, 2003). Feminist scholars have reacted to this new perspective by exploring the impact of EU soft policies as well (Bruno et al., 2006; Idil, 2008; Krizsán and Popa, 2010; Lombardo and Forest, 2012). They looked at the growing role of non-binding measures, benchmarking and naming and shaming strategies (Beveridge and Velluti, 2008). To do so, they made use of sociological and discursive approaches to identify that both national and regional ⁵ actors learn from EU ‘ways of doing things’ in the absence of hard law (Lombardo and Forest, 2012). Gender mainstreaming offers a solid example of such tendencies as it is a soft strategy that frequently lacks binding capacities. The EU has influenced its development at the domestic level. EU moral leverage, the provision of EU funding, or the dissemination of good practices all represent significant incentives to implement this strategy in member states (Bruno et al., 2006; Kantola, 2010; Lombardo, 2004; Squires, 2007).

The process of downloading powers constitutes another area of interest in identifying policy transfer (Haussman et al., 2010; Outshoorn and Kantola, 2007). Studies on federalism have long been preoccupied with the positive/negative impact of this territorial model. In terms of policy outcomes, federations seemed to yield quite remarkable results at the sub-national level (Vickers, 2011, 2013). Many countries have contributed to the expansion of their equality architecture by setting up regional women’s policy agencies (Outshoorn and Kantola, 2007). Emulation processes between territories – also called the demonstration effect – that typically affect compound states facilitated this enlargement. However, the transfer of policies is especially remarkable in federations featuring distinctive territorial dynamics such as competitive patterns. Here, policies developed in a specific region/level of government are rapidly transferred to other territories that wish to avoid being left behind (Alonso, 2015; Celis and Meier, 2011). This also seems to facilitate innovation, as sub-national entities often strive to ‘lead the way’. These tendencies apply also to the implementation of gender mainstreaming. As the Belgian case vividly illustrates, competence among territories has been a powerful engine to boost its development. Measures have been transmitted between regions and levels of government, spurred by the leading role of the region of Flanders (Celis and Meier, 2007, 2011).

Research agendas on Europeanization and federalism reveal thus that multi-level settings offer a fertile ground for policy transfer – in the case of gender mainstreaming implementation also. Nevertheless, as these agendas have remained mainly disconnected, studies often look at single sources of influence and do not capture the complexity of the EU polity. The multi-level explanatory framework developed here will contribute to filling this gap.

Gender equality policies in multi-governed Spain

Of all the member states, Spain is most affected by regionalization and associated processes. In the aftermath of the Franco regime, the territorial model experienced exceptional and somewhat unexpected development. Efforts to resolve the Basque and Catalonian ‘questions’ – the historical push for self-government in these two regions – led to the inclusion of ‘regional issues’ as essential elements of the constitutional agenda. Instead of recognizing a federal state with specific regions, the Constitution sketched out a rather vague territorial model (Aja, 1993) and established two itineraries for becoming an Autonomous Community. The historic regions of Galicia, Catalonia and the Basque Country were granted the fastest access and the broadest set of powers. Other ‘differential factors’ such as regional languages or special fiscal regimes yielded an even more heterogeneous landscape. The tensions that arose between symmetry and asymmetry, homogeneity and heterogeneity drove the construction of the State of Autonomies (Estado de las Autonomías) (Agranoff, 1999; Máiz et al., 2010). These inherent tensions created a competitive cycle in which the slow-track communities continuously strove to acquire similar competences (playing catch-up) and the fast-track communities in turn sought to differentiate themselves again by acquiring new powers. The result has been described as ‘competitive federalism’ (Moreno, 2008), ⁶ in which all regions enjoy a broad set of powers that allow for self-government in a variety of policy areas.

Gender equality policies in Spain are fairly multi-level in nature and reflect the dynamics of the State of Autonomies (Bustelo, 2014; Ortbals et al., 2011). The first equality machinery, the Women’s Institute, appeared at the national level in 1983. The Spanish regions also included the promotion of gender equality in their regional statutes or constitutions, which were enacted between 1979 and 1983. Sub-state equality institutions soon emerged in all Autonomous Communities, replicating the state example and structures. The Basque Country and Andalusia were the first to establish women’s institutes in 1988 (Bustelo and Ortbals, 2007). Similarly, multi-annual policy plans and programmes targeting women were rapidly imported. Currently, the Autonomous Communities play a leading role in gender equality policies, whereas the national government increasingly plays catch-up (Alonso, 2015). The most recent and remarkable developments in this area, the Equality Acts, were first approved at the regional level. These norms regulate women’s rights, policy instruments and equality institutions in a detailed manner, representing a notable improvement with regard to former equality policies based on non-binding measures (Bustelo, 2014).

Interestingly, some groundbreaking contributions implicitly pointed to policy transfer as a meaningful factor for understanding this remarkable expansion. On the one hand, it is generally assumed that the EU policy framework has had a positive effect (Bustelo, 2004; Lombardo, 2004). Multi-annual equality plans were adopted during the 1980s in response to the European Commission Framework Programmes, and became the main instrument for promoting gender equality. Key EU directives were also transposed to domestic legislation, in some cases even before a country entered the EU (Lombardo, 2004; Valiente, 2003). The adoption of gender mainstreaming strategies in the mid-1990s was strongly related to Europeanization as well. While the impact of the Beijing Conference must not be disregarded, the EU is ubiquitous in gender mainstreaming policies, documentation and literature (Alonso and Forest, 2012) at both the state and the regional level. In fact, the Autonomous Communities pioneered the adoption of this strategy. Equality plans implemented in the Basque Country (1995) and Asturias (1996) were the first to state that all public policies must be revised from a gender perspective. The central government and other regions followed this example in subsequent years and gender mainstreaming became a key strategy throughout Spain.

Domestic dynamics also account for the rapid development of this policy domain. In the initial stages of decentralization, state-level institutions and instruments – including equality plans – were faithfully emulated at the sub-state level (Bustelo and Ortbals, 2007). However, as regional leadership and the strong influence of competitive patterns emerged in the 1990s (Alonso, 2015; Linos, 2003), Autonomous Communities strove to attain prominence between themselves, and with respect to the central government. Some territories played a leading role in this and other social policies, specifically Catalonia, Andalusia and the Basque Country, while others (including the national government) played catch-up (Alonso, 2015; Alonso and Verge, 2014). Again, a look at the implementation of gender mainstreaming helps identify these learning patterns. Key instruments for putting this approach into practice travelled quickly from one territory to another. For instance, gender impact assessment (GIA) was first regulated in Catalonia, in 2001. In response, Extremadura adopted GIA in 2002 and the national government followed suit in 2003. In fact, the national bill was presented before the Spanish parliament by the Catalonian governing party, Convergència i Unió (CiU). Subsequently, GIA was implemented in other territories and currently represents one of the most common instruments for incorporating a gender perspective into regular policies. Interdepartmental commissions, equality units, gender training and participatory councils were generalized in a similar manner (Alonso, 2015). The gender mainstreaming toolkit developed in Spain is thus closely related to the vibrant learning dynamics brought about by the State of Autonomies.

The three regions analysed here enjoy a significant level of development in gender mainstreaming, which goes beyond its adoption as a mere strategy. Even though the Basque Country is the champion in this respect, Galicia and the Balearic Islands also have Equality Acts that embrace specific instruments to further its implementation. It is thus worth exploring how gender equality is mainstreamed into their regular policies.

Learning how to mainstream gender equality: A focus on the Spanish regional level

Conclusions

Gender mainstreaming has been the subject of numerous scholarly works, eager to find the formula for its effective implementation. In the EU, Europeanization and federalism research agendas have shed some light on how policy transfer can facilitate its practice and benefit gender equality more broadly. However, they have not provided convincing explanatory frameworks that capture how policies travel within multi-level settings.

This article has integrated elements of both literatures to engage with the classic question of who learns what from whom, and to disentangle how policy transfer processes contribute to gender mainstreaming implementation. The Spanish Autonomous Communities, conditioned by the strong political and territorial dynamics of the Spanish multi-governed state, provided a suitable case study to do so. We looked specifically at the inclusion of a gender perspective in research policies. The presence of solid precedents in this area at the EU, state and sub-state levels made it possible to track the extent to which lesson drawing occurred. The examination of these processes through interviews with key policy actors in three Autonomous Communities constituted the core of the analysis. The cases showed that the key actors involved in both research and equality policies recognize their multi-level nature, and often map and borrow good practices. The article further shows that Europeanization appears to influence beyond the transposition of binding norms at the state-level. The inclusion of gender equality goals and measures in the regions analysed corresponds strongly with the existing EU commitments. In Galicia, this compromise was encouraged by the presence of EU funding, whereas in the Basque Country interest in surpassing state-level policies played a significant role. In line with what recent literature has suggested, the actor-centred and bottom-up approach adopted here was helpful for capturing the impact of the EU in the absence of pressure for adaptation from the EU itself. It provided a clearer grasp of the influence of EU research policies, which draw on the Open Method of Coordination and its related soft tools (e.g. benchmarking, naming and shaming). This approach made it possible to explore the regional level, which has been largely disregarded in scholarly works on Europeanization.

Second, domestic dynamics emerged as a significant catalyst for change in our three cases. Spanish competitive federalism facilitated the inclusion of gender equality in two different ways. In the Galician case, policies undertaken at the state level, and by other regions, provided both an incentive for introducing a gender perspective, and a knowledge resource for implementation. As highlighted before, Basque policy actors emphasized their intention to follow the EU policy framework as a way of ‘doing it better’ than the national government and preserve its traditional leading role. Both cases show how the competitive cycle powerfully informs the Spanish policy-making process. Similarly, they accurately capture the intricate way in which the EU multi-level setting facilitates the implementation of gender mainstreaming by providing multiple sources of influence. Nevertheless, it became clear that the existence of significant precedents for promoting equality in science is not a sufficient condition for policy change. Balearic policies reflect how the actors’ willingness to pursue change remains a crucial factor. Multi-level settings provide ‘platforms for policy transfer’, though actor resistance to implementing gender mainstreaming, or equality policies in general, has been widely acknowledged and should be taken into consideration.

If the relevance and comprehensiveness of the patterns revealed in these cases are to be adequately assessed, then further analyses of other countries, policy areas and levels of government are needed. The current implementation of austerity measures, with potential consequences for equality policies across nations and regions, calls for new studies as well. In such an environment, the presence of policy transfer may lead to the shrinkage, rather than the expansion, of gender mainstreaming practice. Future studies should explore whether multi-level settings remain significant for the promotion of gender equality, or become a threat if actors ‘learn’ to dismantle existing policies.