Sticky Networks in Times of Change: The Case of the European Women ’ s Lobby and Brexit *

Addressing the case of the UK ’ s membership of the European Women ’ s Lobby (EWL), this article explores the signi ﬁ cance of European Union (EU) civil society organization (CSO) membership for domestic CSOs and, in turn, the impact of Brexit on (the nature of) that membership. Analysis adopts a territorial perspective to re ﬂ ect the UK ’ s ‘ four nations ’ approach to representation, which affords the constituent sub ‐ state CSOs notable access to this European network. Analysis of interview and text ‐ based data identi ﬁ ed shared core beliefs and resource dependencies as binding the UK ’ s sub ‐ state CSOs within the EWL, resulting in capacity building at the sub ‐ state level. These shared core beliefs can explain the EWL ’ s decision to formally accommodate the UK coordination post ‐ Brexit, with resource dependencies acting as a further incentive for sub ‐ state actors. However, beyond the policy sub ‐ system, there are practical constraints placed on the UK ’ s ability to collaborate and a diminished value of some resource.


Introduction
Feminist civil society organizations (CSOs) play an important role in the 'velvet triangle' (Woodward, 2004) of feminist activists, femocrats and experts that promotes gender equality in the European Union (EU). The European Women's Lobby (EWL) is the largest such CSO. As an EU-wide membership-based organization with formalized governance arrangements, for national members such as the UK Joint Committee on Women (UKJCW) the EWL provides a ready-made opportunity structure for horizonal and vertical collaboration between its 31 national coordinations and 17 European associations (European Women's Lobby, 2020). The nature of the UK's collaboration in these EU-wide CSO networks is, however, challenged by Brexit (Danisi et al., 2017;Guerrina and Masselot, 2018;Minto, 2018;Haastrup et al., 2019;Ritch, 2019), as the UK de-Europeanizes or disengages from the EU (Copeland, 2016;Burns et al., 2019). Such secondary impacts of Brexit compound the direct impacts on women's rights and gender equality of the UK moving outside EU legislative and governance frameworks (for example, Guerrina and Murphy, 2016;Fagan and Rubery, 2018;Guerrina and Masselot, 2018). An array of analyses from feminist activists and think tanks (for example, Fawcett Society, 2020;Women's Budget Group, 2018) also expose the gendered economic, social *Many thanks to the three anonymous reviewers for their valuable comments on an earlier version of this article and to Professor Daniel Wincott (guest editor of the Symposium) for his constructive feedback throughout the development of this article. Also, many thanks to all the interviewees for sharing their time, experiences and insights, both in the interviews and subsequent communications. 1 Witness seminars have been used to capture perspectives and events surrounding inter alia the Abortion Act 1967 (ICBH Witness Seminar, 2001a) and Section 28 and the rise of gay, lesbian and queer politics (ICBH Witness Seminar, 2001b), resulting in publicly available transcripts. resource dependencies that bind CSOs within the EWL (the drivers of their collaboration or collaborative work). Section III examines the CSOs' responses to Brexit, and assesses the significance and effects of the drivers for collaboration and the new post-Brexit context on the UK's continued participation in the EWL. The conclusion reflects on the gendered governance effects of Brexit. The article shows that the UK's participation within the EWL is driven by shared core beliefs, expressed within a common policy sub-system. Horizontal and vertical resource dependencies have formed within this community of shared beliefs. Resource dependencies have had effects at the sub-state level, resulting in sub-state CSO capacity-building within the UK. Post-Brexit, outside the policy subsystem, there will be challenges to the UK's continued participation in the EWL. The UK may drift from European frameworks and face transaction costs which may prohibit collaboration in some EWL activities and projects. However, drivers for continued cooperation remain within this 'sticky' network, evidenced in no small part by the EWL's accommodation of the UK's coordination post-Brexit.

I. Exploring Network Participation: The Case of the EWL and the UKJCW
The EWL was established in 1990 with support from the European Commission (see Hoskyns, 1991) which sought a single EU-level CSO interlocutor on women's interests (Strid, 2009, pp. 138-9). Its stated purpose is, 'to promote the integration of gender equality in all areas of policy within the institutions of the European Union' (European Women's Lobby, 2013: Article 1), and it is influential across a range of key policy areas (for example, Strid, 2009). These areas include women in decision-making and employment, and a dedicated Observatory on Violence Against Women comprised of national experts (European Women's Lobby, 2019a). Despite the contracting opportunity structure provided by the EU (Cullen, 2015), as an established membership-based CSO located in Brussels, the EWL is a key feminist network and important access point for its members to supranational policy-making.
Most research on the EWL has explored its nature and role within supranational politics and policy-making (for example, Helfferich and Kolb, 2001;Strid, 2009;Bygnes, 2013;Cullen, 2015). Research focused on its national members is limited. Karlberg and Jacobsson (2015) have examined the Europeanizing effects of EWL membership on organizational structures and CSO cooperation in Sweden. Lafon (2018) has noted cultural, organizational and individual factors as shaping different Europeanization outcomes in Belgian and French CSOs. These focused studies speak to a broader literature on national CSOs' Europeanization of their activity, including through participation in EU CSO networks. Many scholars recognize that CSOs at once seek to advance particular policy goals and, as organizations, seek to sustain and expand themselves. Normative, capacity and organizational factors shape and are shaped by domestic CSOs' participation within Europe (for example, Beyers and Kerremans, 2007;Karlberg and Jacobsson, 2015;Kröger 2018;Lafon, 2018). Sensitive to multiple levels of governance, Santiago López and Tatham (2018) have explored the Europeanization of interest groups at the sub-state level, observing a relationship between the level of a state's decentralization and regional interest groups' Europeanization. The phenomenon of Europeanization at multiple levels of governance is now well-established (for recent examples, see Tatham, 2018;Huggins 2018), emphasizing the salience of a territorial perspective. In seeking to understand how and why the UKJCW (itself comprised of four sub-state CSOs) participates in the EWL, this analysis focuses on the notions of shared beliefs and resource dependency as driving CSO engagement.
Combining two dominant theories of network formation provides a useful framework to distinguish between these different drivers for collaboration. Both provide explanations for the 'stickiness' of networks. In turn, identification of these drivers provides insight into network cohesion and CSO participation post-Brexit. One theory highlights the role of shared beliefs (Weible et al., 2009) and another focuses on resource dependencies (Pfeffer and Salancik, 2003;Casciaro and Piskorski, 2005). The Advocacy Coalition Framework [ACF] asserts that, within a policy subsystem, shared beliefs bind actors together. It is through these beliefs that actors simplify and interpret the world (Sabatier, 1987(Sabatier, , 1988Jenkins-Smith et al., 2017). Beliefs are expressed through policies. Belief systems are themselves hierarchically structured. Deep core beliefs are normative, entrenched and impervious to change. Policy beliefs are policy translations of more fundamentally held beliefs. At their outer layer are secondary beliefs, which are more easily contested policy interpretations (Jenkins-Smith et al., 2017). Here, ACF helps investigate how shared beliefs bind sub-state CSOs within the EWL network.
Resource Dependency Theory (RDT) offers a more rationalist account of the relations of interdependence that drive actors to collaborate. Ties enable exchanges of 'monetary or physical resources, information or social legitimacy' (Pfeffer and Salancik, 2003, p. 43), influenced by power imbalances and mutual dependence between actors (Casciaro and Piskorski, 2005). RDT draws attention to the capacity building potential for CSOs of voluntary collaboration in the EWL. The idea of 'critical resource dependencies'resources essential for organizational functioningis particularly important here given the nature of the CSOs under study. Beyers and Kerremans (2007) have highlighted the risk of overlooking critical resource dependencies acting at the domestic level, which may restrict the Europeanization of domestic CSOs. As such, the critical resources of the CSOs are considered here.
Various scholars have not limited their analyses of networks to one or other theory (for example, Henry, 2011;Leifeld and Schneider, 2012;James and Christopoulos, 2018). Findings from this research reveal the significance of both shared beliefs and resource dependencies in network formation. Interestingly, Henry's (2011) research reveals that drivers are nested, with resource dependencies between coalitions of actors established within a wider network shaped by shared beliefs; and James and Christopoulos (2018) find that drivers are 'distinct and complementary' (p. 534). Leifeld and Schneider (2012) stress that opportunity structures and minimising transaction costs are key to tie formation between actors. This current analysis treats the relevance of and relationship between belief-based and resource-based drivers as an open question. It explores empirically the (voluntary) tie formation within an existing, formalized network. When compared with diffuse coalitions of actors, these networks reduce transaction costs significantly for national and European CSO members to establish horizontal and vertical ties. Changes to the 'fit' between sub-systems and the subsequent 'value' of resource may shape future collaboration in EU CSOs post-Brexit. Beyond even these alterations, the changing nature of this opportunity structure for the UKJCW remains highly relevant.

II. Shared Beliefs and Resource Dependencies in the EWL Network
Structure and Organization of the UK's National Coordination '… it is hard work, working across the four countries and working from a four-nation perspective. It's always been a source of tension and contention, but the sisters have managed to pull it off. ' (quote#1) Interviewees provided rich narratives of the establishment and development of the UKJCW, both first and second-hand accounts. As a body coordinating the UK's representation at the EU-level, a striking feature of the UKJCW is that it comprises four 'constituent groups': from England, Northern Ireland, Scotland and Wales (UK Joint Committee on Women, 2013). Unlike most intra-UK structures, it operates as a partnership of equals, treating each part of the state as equally authentic and legitimate. Furthermore, it did so prior to devolution. As with the Swedish coordination (Karlberg and Jacobsson, 2015), the requirements of the EWL shaped the organizational structure of the UK's coordination. EWL rules stipulated a single coordination per national member, each holding four seats. EWL rules have since changed but the quasi-federal UKJCW model remains. Today, these four sub-state groups are the National Alliance of Women's Organisations (NAWO) in England 2 ; Engender in Scotland; Women's Equality Network Wales (WEN Wales); and the Northern Ireland Women's European Platform (NIWEP). NAWO and NIWEP were amongst the four founding CSOs of the UKJCW (and EWL) in 1990; the Scottish and Welsh representations have shifted over time. 3 The EWL statues (European Women's Lobby, 2013) require that each national CSO is a network of CSOs (Title II, Article 5, paragraph 1). UKJCW roles are distributed across the four organizations: one UKJCW member is the EWL Board member, another is the alternate; with the two other organizations acting as UKJCW chair and secretariat respectively. These roles rotate on a two-yearly basis. An annual face-to-face meeting is complemented by an informal catch-up in the margins of the annual General Assembly, and supplemented by regular Skype meetings and email exchanges between UKJCW CSOs.
Each sub-state CSO is a distinct organizational entity, firmly embedded within its own territory, from which its operational funding ('critical resource') may be sourced. The resourcing of sub-state CSOs has varied over time, posing a challenge to the functioning of the UKJCW. As one interviewee stated, '[cross-national working] has always been quite tricky and I think that the particular reason for the trickiness is the difference in the resourcing of the four member organizations. Historically this has waxed and waned as different organizations have attracted different levels of funding' (quote#2). Differentiation in resource is apparent today. Currently, NAWO is the least well-resourced (with a recorded income of £3.6k for the year 2018, which was down on previous years) and is run entirely by volunteers (NAWO, 2018); whilst Engender is the largest and most well-resourced, with a staff of seven and an income of c.£350k for the year 2018/19 (Engender, 2019a). WEN Wales comes in second with a staff of two and an income of c.£160k for the year 2018/19 (WEN Wales, 2019) (although staff levels have since increased). Both the Scottish and Welsh CSOs receive funding from their sub-state governments (nearly 80 per cent in Scotland, and over 70 per cent in Wales for the year 2018/19). NIWEP's income was c.£35k for the year 2018/19 (NIWEP, 2019), boosted by a grant which covers some staff costs; otherwise its activity is supported through volunteers from the Board, who act as delegates to the UKJCW. These differing resource levels could compromise sub-state CSOs' ability to engage at a European level, with knock-on implications for domestic capacity development. It is particularly notable that NAWO (the CSO representing the significantly larger and more dominant Anglo-British part of the state) is the least well-resourced of the four.

Belief-Based Drivers of Collaboration: Feminist Solidarity in a Policy Subsystem
The UKJCW Protocol (UK Joint Committee on Women, 2013) sets out the constituent CSOs' common objectives to promote women's rights and gender equality (paragraph 3), commitments which are shared with the EWL (European Women's Lobby, 2013). These commitments are presented as at the core of each of these individual, values-based organizations, both by the interviewees and in policy texts/communications. For example, NAWO, 'seeks to promote gender equality and social justice for women' with its mission 'to promote the human rights of all women and girls, with a special focus on gender equality and Europe' (NAWO, 2015, p. 3). Engender presents itself as 'Scotland's feminist membership organisation' holding a vision, 'in which women and men have equal opportunities in life, equal access to resources and power, and are equally safe and secure from harm.' (Engender, 2020a). WEN Wales seeks, 'to ensure that Wales is the safest country in Europe to be a woman and that Wales is a truly feminist nation, where women and girls can flourish and actively participate in their communities' (WEN Wales, 2018, p.2). Finally, NIWEP's core objective, 'is to give women and girls in Northern Ireland a voice at the national and international level' (NIWEP, 2017). Notably, the work of each constituent organization is territorially located but the beliefs of women's rights and gender equality transcend these territorial boundaries. Participation within a European community of feminists (through the EWL) was understood to provide an opportunity for the development of feminist thinking. As one interviewee stated, '[The EU provides] a space in which some of those international norms about women's rights and gender mainstreaming, however imperfectly realised and implemented, are shaped, and created, and transmitted, and restated …' (quote#3). This conceptual development displays elements of resource dependency; however, it is shared feminist beliefs that appear to cohere actors together.
Today, EU-focused activity represents one element of the work of these CSOs (for example, Engender, 2019a;NAWO, 2015;NIWEP, 2017;WEN Wales, 2020), as they undertake a wide range of activity in their distinct territorial locations, with some UK-wide work as well as engagement with international women's rights and gender equality frameworks, specifically the UN's Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW), the Beijing Platform for Action and the Council of Europe's Istanbul Convention on preventing and combating violence against women and domestic violence. This multi-level focus is reflected in the UKJCW Protocol (2013) which sets out the specific aims of the UKJCW (paragraph 5) as primarily to coordinate the UK's representation to the EWL but also to act with respect to national (namely, sub-state), UK, European and international levels of governance 'as may be helpful'. Notably, these European and international frameworks for women's rights and gender equality mirror those followed by the EWL.
Beyond legislative frameworks, since the adoption of gender mainstreaming in the late 1990s, the EU has developed an array of 'soft' governance tools in the area of gender equality, including the creation of an agency for gender equality, the European Institute for Gender Equality (EIGE) from 2005 (Hubert and Stratigaki, 2011). In 2013, EIGE launched the Gender Equality Index that evaluates all EU Member States against a range of gender equality indicators (Valkovičová, 2017, pp. 25-6). The significance of this tool was noted. One interviewee stressed that, as a policy tool, 'the Gender Equality Index … is really helpful when I'm talking to Ministers here … It's really quick and useful; you've got 28 countries there just like that' (quote#4). She emphasized the normative significance of this tool for her advocacy activity: 'I am more likely to look at the European context because it's kind of closer and seems more comparable to us' (quote#5) … '[looking across the EU], we're more comparable … comparing ourselves … to countries that are really a long way behind us economically isn't perhaps as useful as comparing ourselves to our European neighbours …' (quote#6). Interviewees also referred to the Fundamental Rights Agency's (FRA) activity and reporting around violence against women (for example, Fundamental Rights Agency, 2014). As stated by one interviewee, 'It's a huge tool for us to be able to come back here and say the Fundamental Rights Agency says that one out of three women in Europe are experiencing A or B or C, and that [this nation] is part of the UK which experiences D …' (quote#7).
However, multiple interviewees stated that the value of these tools was limited. For example, one interviewee said, 'the only frustrating thing about it is that [the devolved-level] isn't disaggregated. It is still useful, obviously … but [in a recent meeting with a Minister] I used a couple of comparisons but it was really annoying because I couldn't say, "And [we are] here" because the Gender Equality Index obviously doesn't go down to [that level]' (quote#8). Another reservation was raised from a territorial perspective, with an interviewee saying that, 'I think I've got mixed feelings about [this data and reporting] in that we often see … a little bit of obfuscation of the national differences inside those pieces of UK data … I think that there are things that could have been improved within those UK structures' (quote#9). Therefore, although these EU tools situate the sub-state CSOs within the EU's policy subsystem, there are limitations that reflect the mismatch between the EU's state-centric reporting and the territorially differentiated reality of gender equality governance in the UK.
ACF treats shared beliefs as fundamentally significant to formation of ties between actors. In this case, the shared beliefs of women's rights and gender equality were common amongst all CSOs (the constituent CSOs of the UKJCW and the EWL). Importantly, these shared beliefs existed within a particular policy subsystem with data revealing certain elements of the policy subsystem as significant to the constituent CSOs of the UKJCW, including legislative frameworks (European and international) and governance tools. In turn, the policy sub-system conditions the meaning and 'value' of resources exchanged within the EWL network.

Resource Dependency Drivers for Collaboration: Capacity and Influence
• Policy expertise Multiple interviewees stressed the significance of their participation within EWL forums and campaigns for information exchange. For example, one interviewee highlighted her participation in the Feminist Economics Working Group, which met for the first time in October 2016 (European Women's Lobby, 2017). Speaking of her engagement with the Working Group and its work on the European Investment Bank's Gender Equality Strategy, this interviewee reflected that 'there's been some capacity to do some quite detailed learning from the European experience which we're then feeding into our domestic policy process' (quote#10). Another interviewee pointed to her participation in the EWL's Politics and Leadership Working Group: 'because that is something we're focusing on very much here [in our home nation] … and so now I've got this great network of people who are also working in that area that I can share ideas with, see what's worked in their country to get more women into [political office]. So, it's really useful in terms of sharing ideas …' (quote#11). The significance of the multi-level opportunity structures was emphasized explicitly. As articulated by one interviewee, 'the learning from the rest of the world doesn't get to us [at the devolved level] unless we participate' (quote#12). In addition to the EWL's long-standing Observatory on Violence against Women, other Working Groups and Task Forces (including Sexual and Reproductive Health and Rights, and geographically-organized groupings) have been established with increased inclusion in mind. One interviewee from the EWL stated, 'so now it is not only the Board Member representing each national coordination or European association but also members of those organizations who can be involved in those different groups -Working Groups and Task Forces and so onso it is helping to have more women involved in [the EWL's] work' (quote#13).
The significance of the EWL for information exchange, learning and capacity building was highlighted by all four sub-state CSOs. The structure of both the UKJCW and the EWL (including Working Groups and transnational campaigning) allowed UKJCW constituent CSOs to participate, with minimal transaction costs. Therefore, and notably, resource dependencies were territorially differentiated across the UK, with information exchange supporting targeted capacity-building for NAWO, Engender, WEN Wales and NIWEP. The UKJCW was not highlighted as a key structure for extensive information exchange, although the potential of the structure was noted by interviewees. One stated '[I think] that the nations need to get together … we need to share [our] learning and we need to share between the nations rather than all being totally exclusive … there's great learning between each one … we can share information and we can say "Well, we do it this way and how are you doing it? Could it work another way?" …' (quote#14). More recently, there has been increased activity by UKJCW members to bolster the UKJCW's role and profile including, for example, the launch of a biannual UKJCW newsletter (Engender, 2020b). This activity builds on exchanges at both feminist events and also coordinated working in the international sphere (although in some of this work, England-focused activity is undertaken by the Women's Resource Centre (WRC) not NAWO).
• Resources for joint working The EWL's financial contribution supports their members' access to the EWL's opportunity structure. Specifically, funding covers the attendance of Board members at the EWL's General Assembly; an annual decision-making forum for EWL members, held over multiple days. Notably, Commission funding to support attendance at the General Assembly is specifically for members from EU, European Economic Area (EEA) and Accession states (EWL representative 2019, personal communication). The UKJCW Protocol (Article 8) stipulates that all four CSOs will be present at the General Assembly, with those not funded by the EWL funded jointly by the UKJCW's constituent groups. The EWL's funding is predominantly provided by the Commission (for example, data for 2019 shows it provides over 70 per cent of EWL's budget), 4 although the EWL has sought to diversify income streams. One interviewee from the EWL stressed that it 'also receive[s] money from time to time from the Council of Europe … and we also receive money from private foundations like Google or [The] Chanel Foundation …' (quote#15). At present, however, the EWL is largely dependent on the European Commission for its critical resource, reinforcing the EWL's ties to the EU political system.
More broadly, the EU provides funding for gender equality activity (beyond that provided through the EWL). Interviewees associated with the Observatory on Violence Against Women made particular reference to funds for projects related to gender-based violence. These interviewees highlighted the EWL's Observatory on Violence Against Women as a particularly rich site for the establishment of multi-lateral relationships which could draw on EU funding. As an illustration, one interviewee stressed that 'I still have relationships from that [time in the network] … I'm doing [this project with EU money] with women in Portugal, Cyprus, Bulgaria and Romania … and the Bulgarians, Romanians and the Cypriots, I all knew because of the Observatory' (quote#16). In turn, these projects provide opportunities for further resource exchange between participants.
Despite the significance of EWL funding, funding did not include critical resources for organizational survival, as the four constituent CSOs of the UKJCW rely on domestic funding to sustain their operation (with significant variation in resource between them). Instead, EWL funding reduces or eliminates transaction costs to access to the opportunity structure (identified as fundamental for tie formation between actors; Leifeld and Schneider, 2012). EU funding for gender equality projects also emerged as an important resource, supporting partnership-building and information exchange opportunities within the EWL network.
• Political legitimacy and influence in a multi-level political system Interviewees from Scotland, Wales and Northern Ireland stressed the benefit for sub-state CSOs of their own distinct access to and participation within EWL structures. One actor said, 'we're able to share very distinct [devolved] positions … that wouldn't be known by somebody from England. I think that enables us to have that presence and be seen' (quote#17). As stated by another interviewee: 'the work that's happening in Wales, or Scotland or in Northern Ireland is invisible to the rest of the world, which is why we need to be involved in those mechanisms' (quote#18). Interviewees noted the general lack of awareness of the UK's territorial nature. For example, an interviewee from Scotland stated that 'part of the difficulty of working with European institutions and the UN is that there's next to no interest in understanding the devolution settlements' (quote#19). However, despite this lack of understanding, both the EWL's and UKJCW's opportunity structure affords sub-state CSOs access and visibility within this European network. One interviewee provided a particularly focused account of participation within the EU's multi-level system, as a means to achieve domestic policy influence: 'because so much power is gate-kept in London … our experience has been that in order to both influence engagement … in Europe and also benefit from those links, we had to have our own links … and that's what almost all of our European work has been about. It's about levers of power in London, [here in our home nation] and in Brussels … It really does work … and it is a bit about moving through the different levels of government …' (quote#20).
The significance of this distinct access and visibility provided to sub-state CSOs was understood differently across the UK. Specifically, the interviewee from NIWEP noted, 'being part of the Lobby, it has also meant that we have been included in, say, work on conflict resolution where other countries would not be conscious of linking the UK in … so they wouldn't, you know, instantly think, "Oh, yes, that's for the UK because they've got a conflict area"' (quote#21). This level of access and visibility is usually reserved to Member State coordinations. The EWL therefore reinforces the political legitimacy of these sub-state CSOs, as distinct actors embedded in distinctive sub-state polities and, in turn, creates opportunities for awareness-raising around the UK's territorial nature.
Related to the notion of political legitimacy, one interviewee said, 'we absolutely don't have the capacity as individual organisations to engage in detail with [the European level]. So, I suppose, the key benefits of participating in the Lobby is to strengthen a feminist organisation that will advocate on the detail of bits and pieces … The European Women's Lobby facilitates national participation of feminist coordinations … they do quite a lot of the administrative heavy lifting around quite abstruse policy processes.' (quote#22) This quote contains many important elements, including: the resource constraints that prevent CSOs from undertaking lobbying at the EU-level (despite the relevance of EU legislation) and the value of the EWL's specific expertise. This is an example of the amplification of political influence afforded to sub-state CSOs through collaboration within the EWL, as the impact of advocacy activity at the supranational level is significantly increased.
Two elements of this heightened sub-state legitimacy (and related political influence) are particularly interesting. Firstly, through the UKJCW, each part of the UK state is treated as equally authentic; in contrast to standard practice, which sees the Anglo-Britain/England dominate. Strengthening political legitimacy is beneficial for the sub-state CSOs within a territorially differentiated UK-state. Secondly, the UKJCW's territorial structure does not map on to the EU's own Regional politics. Whilst Scotland, Wales and Northern Ireland were indeed EU Regions, England comprised nine Regions.
RDT emphasizes the formation of ties for resource exchange. Analysis of interview and documentary data uncovered clear evidence of both horizontal (between national CSOs) and vertical (between the supra-national and sub-state levels) resource exchange relationships. Sub-state CSOs gained resources such as policy information/expertise; resources for joint working; and political legitimacy in a multi-level system. Although critical resource dependencies did not operate vertically (from the supranational to the sub-state level), tie formation within the network was enabled by low transaction costs for access to the EWL opportunity structure. In this case, the result was capacity-building at the sub-state (as opposed to Member State) level.

III. Sticky Networks in Times of Change
The EWL provides an accessible political opportunity structure for national CSOs operating within the EU's political system. Analysis identified shared beliefs and resource dependencies as underpinning voluntary collaboration vertically and horizontally within the network. The four coordinations of the UKJCW formed ties according to shared core beliefs within a policy sub-system; and established relationships of resource exchange to build capacity at the sub-state level. Identifying these drivers that bind actors together within the EWL network provides a useful starting point for exploring continuity and change in network participation post-Brexit.
Following the UK's EU Referendum, the EWL's Secretary General stated that the EWL was 'shocked and dismayed' at the outcome (Fawcett Society et al., 2016: 1); stressing that the UK ought to remain part of European and international feminist networks and, specifically, that '[the] EWL and the UKJCW will remain in close contact as they have throughout the past 25 years' (Fawcett Society et al., 2016, p. 2). Steps were taken to secure the UK's continued EWL membership. As one UKJCW representative recalled, 'actually our sister national coordinations were really positive about us staying, which I think was immensely heart-warming to see; so we [the UKJCW] determined we wanted to hang on in there' (quote#23). The strength of feeling was tangible amongst interviewees, with one from the EWL Secretariat emphasizing that, 'we absolutely want to do everything in our power to stop our members being abandoned from our network' (quote#24). In accordance with the EWL's governance processes, changes were made to the EWL's internal rules such that a national coordination can maintain full membership even if the state they represent changes its formal relationship with the EU (Article 2, European Women's Lobby, 2019c). This amendment was supported by the Board in February 2019 [with the Board 'massively in favour' (quote#25)] and subsequently approved unanimously by the General Assembly on 7 June 2019 (European Women's Lobby, 2019d).
The public announcement of this watershed amendment to the EWL's internal rulesentitled 'Solidarity with EWL's UK Coordination' (European Women's Lobby, 2019d)evidenced both shared beliefs and resource drivers for continued collaboration. The EWL President stated: 'We must show support and keep our sisters with us, making sure that solidarity and sisterhood of our Feminist Movement go beyond the political context'. Statements from the UKJCW coordinations were also included in the announcement (European Women's Lobby, 2019d). They referred to inter alia 'continued solidarity and sisterhood', 'our sister organisations' and 'a feminist Europe' (Engender); 'the feminist movement in Europe' (WEN Wales); and 'sisters across Europe' and 'a flourishing feminist Europe' (NIWEP). Whilst attention to a community of shared beliefs was particularly pronounced, statements also touched on resource dependencies. These statements included the 'vital space … to amplify our voices on advocacy, collaborate on campaigns at a national and European level, and push back against the rise of populism that is undermining women's equality and rights' (Engender) and the 'huge benefit to us in the UK' (WEN Wales); with NIWEP stressing the value of feminist collaboration 'in building sustainable peace, securing women's reproductive and health care rights and working to the goal of ensuring women's substantive equality'.
Further analysis of the drivers for collaboration provides a more nuanced picture of future UKJCW collaboration in the EWL network. Brexit impacts on both belief-based and rational drivers for collaboration, albeit to varying degrees. Core beliefs shared by CSOs transcend political boundaries so will continue to drive collaboration post-Brexit. However, these beliefs operate within a common policy sub-system which comprises EU-wide legislation and governance tools. UKJCW members have individually and collectively emphasized the risks of falling outside EU legislative frameworks (for example, UKJCW, 2017;WEN Wales, 2018, p.8;Engender, 2019b). In October 2017, the UKJCW sent a letter to the Secretary of State for Exiting the EU in which it stressed the risk of Brexit to women's rights in the UK and highlighted the lack of gender-sensitivity in the Brexit negotiations (UKJCW, 2017). Interviewees also expressed concern about the UK withdrawing from EU-wide reporting. Referring to the FRA, interviewees stated that, 'It's going to be devastating to lose that kind of information' (quote#26) and that it would be 'horrific to lose all of that' (quote#27). The impact of this drift from the EU policy sub-system was highlighted by one interviewee from the EWL: 'so what is the interest then of our UK members to engage in processes like that [EU lobbying activity]; they wouldn't be bothered because there's no actual impact in the UK anymore. They might be interested because then they can say, "This is what's happening in the EU, we need to align with that"' (quote#28). Notwithstanding this matter of continued relevance, the influence of shared core beliefs as binding actors together was evident through analysis of the data. As one sub-state CSO actor stated, '… even though some of the work would then have less immediate relevance to us. It seemed just so important to maintain that international solidarity and to continue to be part of the conversations and to find out what's happening in Europe' (quote#29).
Regarding resource dependencies post-Brexit, incentives will remain for the UKJCW's sub-state CSOs to form horizontal and vertical ties to secure additional domestic capacity, although outside the policy sub-system the 'value' of resource and attached transaction costs change. Information exchange (a key cause of collaboration within the EWL) will remain relevant when not exclusively attached to the policy sub-system. The EWL's recent introduction of more forums for thematic working (in addition to EU-focused legislative and policy-specific work) provides scope for continued participation. Transaction costs for some participation will likely increase as the UKJCW's eligibility for funding to attend the General Assembly is dependent on the UK's future relationship with the EU (given the aforementioned eligibility requirements for European Commission funding). Furthermore, it is highly unlikely that UK-based CSOs will have access to EU funds for multi-lateral projects. For example, the January 2019 call for funding for projects to combat violence against women was opened to organizations in EU Member States, Iceland and the Republic of Serbia (European Commission, 2019). As a non-EU Member State and without a comprehensive relationship with the EU, the UK would not have access to these funds. These practical barriers to participation go beyond the UK's formal relationship with the EWL but impact collaboration nonetheless. As one interviewee stressed, 'I think it is a nightmare for us, it's really a nightmare … none of that [project funding] will be available to us again and not only for us to apply for it but for us to participate along with other people.' (quote#30).
The UKJCW's and EWL's responses to Brexit and the maintenance of the UK's full EWL membership reflects both belief-based and resource drivers for collaboration. The common core beliefs shared by the EWL's community of gender equality CSOs appeared to underpin the continuity of the UK's collaboration within the network; with resource dependencies acting as a further incentive for sub-state actors to maintain relationships for capacity-building at the domestic level. This nesting of resource-based ties within a belief-based community echoes the findings of previous research (Henry, 2011). However, with the UK beyond the policy sub-system, there are practical constraints placed on collaboration and a diminished value of some resource.

Conclusion
Changes to the UK's participation in EU-wide CSO networks is part the wider transformation of the UK's gender equality landscape post-Brexit. Situated at the intersection between feminist CSOs and sub-state spaces and identities, this analysis makes a dual contribution. First, it begins to document the history and evolution of the four-nation coordinating body and its constituent sub-state CSOs, expressly established for the UK's representation to the EWL. Analysis recognizes, highlights and explores the UKJCW's unique structure that, although created pre-devolution, reflects the territorially differentiated context for and approaches to promoting women's rights and gender equality across the UK. Further research is required to fully capture and preserve the history of the UKJCW and its constituent organizations. Secondly, the analysis provides insights into the drivers for collaboration within the EWL network and, in turn, the significance of this opportunity structure for domestic CSOs. Analysis of data from semi-structured interviews with CSO actors and policy texts/communications revealed both strong belief-based drivers for collaboration (shared core beliefs, within a policy sub-system) and resource dependencies (including expertise; resources for joint working; and political legitimacy in a multi-level system), with the latter seemingly nested within the former.
Shared beliefs and resource dependencies were useful explanatory factors for the UKJCW's desire to continue participation in the EWL post-Brexit and the EWL's subsequent amendment of its internal rules to enable this continued membership. However, systematic consideration of these drivers, the EWL's opportunity structure and the wider policy sub-system highlights practical factors that will constrain future participation. Indeed, being outside the EU's legislative, policy and funding frameworks, and removed from EU-wide comparative datasets and reporting raise barriers for continued participation, with capacity implications for the sub-state CSOs of the UKJCW accompanying the UK's 'drift' from European norms. Countering this, the EWL has established thematic working groups (opportunity structures) that are not focused around the EU's legislative agenda and is continuing efforts to secure funding from non-EU sources. The inclusion of Iceland as the first EWL member from the EEA in February 2019 (European Women's Lobby, 2019b) is perhaps another indication of a more systematic broadening of the EWL's reach beyond the EU. Finally, research findings underscored the significance of EWL membership for sub-state CSOs, with resource dependencies established from the sub-state level in the UK. Therefore, capacity building took effect at the sub-state level, as CSOs from Scotland, Wales, Northern Ireland and England sought resource for distinctive domestic agendas. The significant variation in resource across the four UKJCW constituent CSOs will result in differential impacts of resource loss across the nations, with knock-on implications for gender equality governance.
Correspondence: Rachel Minto Wales Governance Centre Cardiff University School of Law and Politics email: mintor@cardiff.ac.uk