Theorizing gender desegregation as political work: The case of the Welsh Labour Party

Organization studies offers a detailed understanding of the roots of gender segregation and the obstacles to its disman-tling in practice but has not proposed a conceptual framework that can help us understand how radical forms of desegregation may be made sense of and approached, particularly within a hotly contested organizational context. We provide an empirical analysis of the UK's only positive discrimination intervention, in the British Labour Party, and offer a conceptual framework of desegregation as political work, contributing by expanding knowledge of the contes-tations and possibilities inherent in desegregating organizations. We argue that successful radical desegregation is based on disrupting and contesting the foundational ontological values and identifications of a profession or organization, as gender is intimately enmeshed in these. From this basis we propose two political practices of desegregation: ‘ standing up ’ and ‘ walking with ’ .

Our contribution lies within the feminist organization studies literature, expanding knowledge of the contestations and possibilities inherent in desegregating organizations (Brumley, 2014;Cohen & Wolkowitz, 2018;Johansson & Ringblom, 2017).Conceptually, we enrich a view of gender as a political practice by drawing from agonistic and post-foundational theory (Mouffe, 2013), articulating these ideas with feminist organization studies to develop understanding of what it means to practice radical forms of gender desegregation.We highlight the ontological terrain upon which desegregating unfolds and posit two practices rooted in an acknowledgment of the conflictual work of desegregating -'standing up' and 'walking with'.Both carry implications for rupturing organizational ontology, foregrounding conflict and the 'agony' of change.Our analysis provides a conceptual language and practice framework that is relevant to raising awareness of the inseparability of radical desegregation from the broader power dynamics of organization and, therefore, also the contestation that may ensue.We conclude by exploring the wider relevance of this kind of action through arguing for more recognition of how the practice of gender is interwoven with the contingent and contested political terrain of organization.

| ORGANIZATIONAL SEGREGATION AND DESEGREGATION
Organization studies research provides clear insight into the causes of segregation and obstacles to desegregation, as well as some examples of proposals for desegregation.The issues are broad; a stereotypical division of gender roles begins with the segregation of reproductive labour within the private sphere (Dhar-Bhattacharjee & Richardson, 2018;Rafnsdóttir & Heijstra, 2013), with knock-on effects for women's progression within organizations (Casinowsky, 2013;Crompton & Lyonette, 2011;Larsson, 2007;O'Hagan, 2018).Likewise, pervasive social norms are often held up as contributing to gender segregation at work (see Ali & Syed, 2017;Brumley, 2014;Masika, 2017;McDonald, 2016).Dominant cultural logics of gender dictate certain normative frameworks that bear performatively on both sexes (Butler, 1999).Men continue to be stereotypically perceived as 'breadwinners', while women are perceived as 'caregivers' (Brumley, 2014;Cohen & Wolkowitz, 2018, p. 57), which contributes to the further entrenching of gender inequality in the home, and in turn, the workplace.
Men are clearly identified as a cause of segregation, in that they often, deliberately or not, defend their privileged and segregated organizational positions (Tiessen, 2004).The increase of women in the labour force is posited as a threat to men in terms of an overall decrease in wages or as a threat to an occupational identity (Acemoglu, Autor, & Lyle, 2004;Cohen & Wolkowitz, 2018).While men can contribute towards maintaining a segregated status quo, women can also stand in the way of desegregation efforts.Some women 'subscribe to the very ideologies that subordinate women' (McDonald, 2016, p. 30; see also McDowell, 2015) and can take blame on themselves for not resisting or not taking initiative (Coppock, Haydon, & Richter, 2014).
Various patterns to promoting workforce desegregation have been highlighted (e.g., Johansson & Ringblom, 2017;Ness, 2012;Noon, 2010).First, there is an instrumental or 'business' case.This is depoliticized utilitarianism, promoting a female workforce as profitable or as something that increases productivity (Ness, 2012).
Women are objectifiedwe/they are positioned as a means to achieve ends of more profit and higher productivity.This is accentuated in organizations that adopt this approach while maintaining organizational practices and processes (Acker, 2006) designed to reward traditionally masculine standards of performancewhere an employee Ringblom, 2017).This approach is based on forging a gender-blind culturea culture within which each individual will be valued according to their contribution and rewarded fairly regardless of sex (Sattari & Sandefur, 2018).However, the approach masks structural inequalities (Sattari & Sandefur, 2018), overlooking already existing disadvantages women enter organizations with, such as unfairly distributed reproductive labour across household members (Dhar-Bhattacharjee & Richardson, 2018), or 'discrimination in relation to sexual harassment' (French & Strachan, 2015, p. 240) to name only two.
Third is the more radical move of positive discrimination in favour of women, explored empirically here.This stands in contrast to equality of opportunity, which does nothing to acknowledge the transformative change needed to achieve equality of outcome (Noon, 2010).The most prominent manifestations of this approach are quota systems, which mandate a minimum proportion or percentage of a specific identity group in a cohort.Although such approaches are based on crude 'body counting' (Billing & Alvesson, 2000), which can be interpreted as a conservative perspective on identity (Forstenlechner, Lettice, & Özbilgin, 2012), we classify them here as radical interventions because of their inherently revolutionary challenge to established conventions of organization and organizing (Dahlerup & Freidenvall, 2005).Equality of opportunity or competitive equality systems are replaced with predetermined equality of results and outcomes.
From a feminist perspective, quotas ought not to be interpreted as discrimination against men or preferential treatment for women; rather, their proponents argue, quotas are a temporary compensation for structural barriers faced under the current inequitable social regime (Dahlerup & Freidenvall, 2005;cf. Noon, 2010).Crucially, Dahlerup and Freidenvall (2005) also suggest that when quotas are implemented and have changed the demographic, women have to consider critical acts to change organizational rules and norms, with the stated purpose of improving organizational life for themselves and others (cf.Cockburn, 1989;Krook & Norris, 2014).This final point signals that quotas go beyond their immediate utility in balancing numbers to the more political purpose of seeking to influence and change the underlying social norms preventing women from attaining senior positions within organizations.
To conclude, previous research provides a comprehensive overview of the causes of gender segregation at work, along with detailed accounts of the outcome effects of quota implementation in political organizations (although we note Dahlerup's (2008) call for more qualitative evidence, especially related to women's dilemmas in whether to engage with positive discrimination interventions).However, our reading of the extant research suggests a lack of conceptual analysis to help understand the lived political dynamics of radical desegregation.The next section develops a conceptual frame to that end.

| SEGREGATION AND DESEGREGATION: A POLITICAL VIEW
The primary purpose of this article is to propose a conceptual framework that assists in interpreting the contested and political dynamics of desegregating organizations.To do this we work with Chantal Mouffe's (2005Mouffe's ( , 2009Mouffe's ( , 2013) ) post-foundational account of the political.This perspective helps us view gender as an ongoing practice (Butler, 1999), maintained and contested through and against deeply embedded norms and everyday articulations.
Post-foundationalism foregrounds conflict and exclusion as ontologically constitutive of organization, and views attempts to subdue such politicization as themselves exercises of hegemonic power that seek to naturalize a status quo (Mouffe, 2005).
Post-foundationalists insist upon the contingency of meaning and language, as well as the inseparability of language, action and the material (Mouffe, 2013;O. Smolovi c Jones, Smolovi c Jones, & Grint, 2019;S. Smolovi c Jones et al., 2016).The grounding of language within meaningful structures is necessarily always contingent, dependent upon association and connection.As the world is ultimately contingent, attempts to finally ground meaning will be unsuccessful, as processes, identifications and power constellations are always projects in the making, open to contestation and further revision through the 'ever present possibility of antagonism' (Mouffe, 2005, p. 15).Hence the preference in Mouffe's work for the word 'identification' over 'identity' when she addresses the democratic.Identity, in Mouffe's (2009) terms, is a more static notion to be disassembled, a sedimentation of language that overlooks the possibility that identity 'can never be fully constituted, and it can exist only through multiple and competing forms of identifications ' (p. 56).As identifications are multiple and constituted through the contingency of language, she envisages the democratic project as one of maintaining the irreducibility of identity through conflictual discourse.
Mouffe thus translates her post-foundational positioning to an understanding of political and democratic practice that is salient for this study.Her differentiation between 'the political' and 'politics' is particularly useful.For Mouffe (2005), any society is the product of 'sedimented practices, that is, practices that conceal the originary acts of their contingent political institution and which are taken for granted, as if they were self-grounded ' (p. 15).This means that any organization will be founded on conflict and exclusionfor example, not viewing women as appropriate participants in civic and democratic processesand will proceed on the basis of 'concealing' such exclusions within an identity, as if the status quo were normal, natural and settled.
The name Mouffe (2005) gives to these sedimented practices that re-enforce a particular hegemony is 'politics' (p.8).This politics operates at the ontic level; when we engage with the procedures and language of politics at face value, we do so in a bounded and surface manner that does not disturb the status quo.The political, on the other hand, is what disrupts foundational assumptions about meaning and status and is therefore concerned with the ontological (Mouffe, 2005).'Political questions' (Mouffe, 2005, p. 9) cannot be solved via technical means from within the system but always involve a more radical challenge to the underlying meaning of people's multiple identifications.
Within the political and democratic frame, Mouffe (2013) differentiates between adversaries who engage agonistically and enemies who engage antagonistically.Agonistic relations are positioned as contests between adversaries who 'share a common allegiance to … liberty and equality' (Mouffe, 2013, p. 7).Through conflict, various adversaries will attempt to establish their own 'hegemonic order' (Mouffe, 2009, p. 13) without attempting to eradicate others they disagree with, but rather aiming to 'convert' them (Mouffe, 2013).Agonistic contestation ought to be disruptive (preventing full closure of identity/meaning), yet also productive (creating vibrant democratic engagement) (Mouffe, 2009(Mouffe, , 2013)).Outright antagonism is an ever-present danger, however, and is positioned by Mouffe (2013) in destructive, even violent terms, as enemies seek to eradicate positions or even presences.Moreover, Mouffe (2005) positions antagonism as arising in particular in situations where more generative forms of conflict have been suppressed.
From this basis we can begin to make sense of gender quotas as political solutions for desegregating, not simply in and of themselves but also in terms of their capacity to raise and contest assumptions of the role and status of gender within organizations, generating positive affirmations and antagonisms in response.The next section sets out our methodology for exploring this possibility.

| METHODOLOGY
The British Labour Party was founded in a small gathering of 129 trade union and left-wing group delegates in February 1900.Mirroring social tendencies, as well as the trade unions' mixed record on gender equality at work, the party has historically had a problem of gender segregation.While officially committed to universal suffrage from its inception, gender equality struggled to be considered a top priority by the party's leadership from its inception through to the mid-1990s.Uniquely among British political parties, however, gender representation is an issue that has been worked on, albeit against considerable contest and resistance, making it a particularly suitable setting for exploring the political work of desegregating.In the current parliament, elected in 2019, Labour counts more women (104) than men MPs.Between 1918 and 2018, Labour alone accounted for 57.8 per cent of all women MPs elected (Keen, 2015).In the devolved nations of the UK the picture is even more striking.Since the Welsh Assembly was established in 1999, Labour has had near 50-50 female-male representation and consistently has a higher proportion of women than other parties.This is remarkable progress; after the 1992 general election (the last without any form of desegregation intervention), only 37 Labour MPs were women from 271.At the 1997 general election, when Labour introduced gender quotas, 101 women were elected as MPs (24 per cent of cohort).Labour is far ahead of all others in this (compare, for example, the Conservative Party's 2019 result of 87 women elected, alongside 277 men).
Labour's main approach to gender desegregation is the all-women shortlist (AWS).AWS implementation means that when there is a vacancy in a seat, party officials choose to restrict the choice available to local party members in their selection process to women only.Wales was of particular interest to us as it was the site of considerable contestation regarding the implementation of quotas.Firstly, arising upon the establishment of the devolved Welsh Assembly in 1999, an intense public debate emerged concerning the party's decision to implement a 'twinning' policy to select its candidates for the new institution.Under this plan, two winnable constituencies were twinned with one another and party members were mandated to select a woman as a candidate in one or both seats.The policy was narrowly approved in a special conference following months of heated disagreement, with opposition largely coming from poor and former industrial constituency parties.Then in 2005, the parliamentary constituency of Blaenau Gwent, a former centre of coal mining and steel production with high levels of unemployment and long-term illness, became the focus of the next controversy.At the 2001 general election, it was the safest seat in the UK.The incumbent Assembly Member (AM) for the area, a popular local Labour politician called Peter Law, led a local grassroots resistance to the imposition of an AWS.Buttressed by his supporters and a wave of popular support, Law left the party to stand as an independent candidate and he won the election, becoming both the constituency's MP and AM.He died less than a year later and antagonisms were reopened through two by-elections to fill the now vacant AM and MP roles.These were won by two left independent candidates -Law's wife, Trish Law, and his former agent, Dai Davieswho positioned Labour as elitist and out of touch with local needs.Both seats have since reverted to Labour.The twinning and Blaenau Gwent episodes were frequently cited by our participants as instances where issues of gender justice collided and were interspersed with broader issues of class and material justice.The gender quota system within Labour has therefore been successful in terms of generating change but has also provoked contestation and resistance.To better understand these contested dynamics, we selected a purposive sample of elected politicians, party members and officials within the Labour Party, which, at the time, was the only UK organization using a quota system to change its demographic professional profile.We conducted 21 semi-structured, in-depth interviews with party members, elected representatives and staff involved with the implementation of and campaigns for/against gender quotas.Of these, 11 were parliamentarians, two of whom were also former members of party staff; six were members of party staff; and four were party members.We interviewed 13 women and eight men.The interviews explored the reasons our participants had become members of the Labour Party; where they stood in relation to gender quotas; and their role in the initiatives to introduce and implement quotas.
The rationale for this design was to surface some of the foundational associations participants had with the organization and how they positioned notions of gender balance.Conducting interviews was a pragmatic choice of method, acknowledging the well-established difficulties of gaining sustained research access within political parties (Lees-Marshment & Smolovi c Jones, 2018).Interviews provided us with an extended and focused series of accounts of desegregation initiatives, with the historical nature of the events encouraging participants to glean some lessons from what they had experienced, and we acknowledge that this focus necessarily misses the possibility for a 'rawer' and more immediate set of impressions that would have arisen from a participant-observation approach.
We transcribed the interviews and analysed them informed by the principles of post-foundational discourse analysis (Jørgensen & Phillips, 2002;O. Smolovic Jones et al., 2019), with a particular focus on the contingency of language.We assumed that the articulations of our research participants were partial and contingent attempts to ground meaning related to gender and, indeed, the nature of the organization itself.This process followed three stages.
First, we focused on searching transcripts for, and analysing the operation of, empty signifiers (Laclau, 2015), which enjoy a privileged status in post-foundationalism.They are taken to indicate the underlying appeal and logic with which people hold together their sensemaking and sense of self.The emptiness of empty signifiers is achieved through their function as nodal words within a contingent system of language.They are marked by both 'overdetermination' and 'underdetermination' (Laclau, 2015, p. 66), meaning that they hold a curious status of being packed full of significance but also being light and loose of concrete meaning.As such, an empty signifier 'can be (almost) whatever one desires it to be' (Kenny & Scriver, 2012, p. 617; see also Islam, Holm, & Karjalainen, 2017), but can also indicate the 'compelling utopia' (Kenny & Scriver, 2012, p. 616) a particular group works towards.
Empty signifiers were identified through repetition and by the way they connect and hold together other associations within a person's articulation.Notions of 'equality' and 'fairness' in particular drew together a host of memories and associations.We bracketed each instance of equality and fairness and analysed the sense made of these in relation to the extended articulations surrounding them.
Second, we analysed the meaning articulated in relation to empty signifiers.We worked with the postfoundational concept of the chain of equivalence, seeking to understand how an articulation is assembled and who is evoked by a speaker to accomplish a particular grounding of meaning (Islam et al., 2017;Kenny & Scriver, 2012;O. Smolovi c Jones et al., 2019).This means paying attention to who and what are regarded as bolstering a certain claim, alongside who is regarded as standing in antagonistic opposition to the articulated chain.Antagonism is of central importance, as chains of meaning are constituted as much by the antagonisms they exclude as by the more positive associations included (Mouffe, 2013, p. 18).For Mouffe (2013, p. 18), the external antagonist, a '"they" … will serve as a "constitutive outside" for the "we"', and this antagonistic relationship is the hallmark of the 'moment of the political'.
Third, we sought to make sense of how the process of managing or resolving conflict was articulated.Mouffe's (2013) distinction between antagonism and agonism proved a useful initial heuristic to understand underlying logics of how actors and groups related to one another in various disputes.Yet after careful reading and rereading we were unsatisfied with this somewhat binary distinction, noting significant overlap and nuance.We now elaborate on these nuances through our political analysis.

| POLITICIZING THE ONTOLOGY OF ORGANIZATION: EQUALITY
It was striking that research participants drew on the same empty signifiers in their articulations in favour of and against the use of quotas.This signals a sedimentation of the core values of the organization; yet quotas had the effect of raising and magnifying differences, surfacing previously subdued feelings of hostility towards perceived power imbalances and institutionalized inequality.Gender as a practice was therefore politicized and politicizing, its taken-for-granted status contested, as notions of gender justice collided with notions of economic and social justice, raising doubts about the commitments to equality that underpinned the organization's sense of being.
A strong socialist ethos accompanied the push for equality through quotas.For many, it was about the 'equality principle' (woman, parliamentarian) that laws should be made by legislators who are representative of their communities, so as not to 'exclude' anyone.For one male former party official, gender equality and socialism were inseparable and the 'battles' for both of them deeply 'political', as he illustrated in this story: My mother was a miner's widow, buying our first ever fridge.We were becoming middle-class.My father had died and she was working in a factory, well paid, but, you know, and she was getting this fridge on hire purchase and they said 'Right, we'll need the guarantor.'They said 'It's a man who will say that we can do this', and she said, 'Well, I haven't got a man, I'm a widow and it's my money.'You've got a patent connection between outdated laws not being changed and an absence of the people affected by those laws in power and it's not just about gender, it's about class as well, obviously.But, you know, women get it worse, working-class women have a harder time than working-class men.And here was an opportunity.
Equality in relation to gender is articulated in a chain of equivalence that draws in an honest and hard-working working-class widow trying to build a home, who is thwarted by an external antagonist and an unnamed presence responsible for creating discriminatory law.Gender quotas are framed in the story as presenting an 'opportunity' to defeat this sort of taken-for-granted politics-as-usual where women are assumed to be answerable to men.The participant here is making the claim for desegregation indirectly, stating that unrepresentative institutions are more likely to produce unequal everyday social experiences.
Women research participants had assumed the existence of a generally shared commitment to equality within the partya chain of equivalence of people and language that gave primacy to notions of equality.This assumption was shaken by their experiences of quotas.In the words of a female former party official: 'I couldn't understand because I start on the premise that I will be treated equally.'She realized she was working for an organization where others saw its foundational premises as being very different.When she worked later as part of the implementation team for quotas in the Welsh Assembly, she recalled discussing the matter with an influential councillor.She warned him that not taking action would result in the Assembly looking like a prominent county council, which at the time was almost exclusively occupied by older white men.The man replied: 'What's wrong with that?' In local constituency parties research participants also grappled with a sense of unravelling as chains of unity amongst comrades committed to equality, established over years of campaigning together, came apart.In the words of a constituency officer, that was … hurtful.These people who claimed to be … you would expect people on the left to be, to be fighting for equality and fighting for poorer people.We're a party of equals, aren't we?
A sense of disorientation was articulated as what formerly constituted a strong chain of equivalence around the notion of equality was disrupted, prompting a political response.That political response manifests as a repositioning of the chain, with previous allies now recast in adversarial terms outside the chain of equivalence, actors whom the newly forming chain must now overcome to achieve equality.
Opponents of quotas also cited equality as fundamental to their struggle.These people were predominantly from former industrial towns, amongst the poorest areas in Europe.Investing time and effort into the Labour Party was one way in which they could proactively work for a more equal society, but their opinions of the Labour governments of the late 1990s and 2000s were largely negative; a sense of an inequality even within Labour prevailed and they shared a view of the party hierarchy as privileged outsiders.Proponents of quotas were articulated as outsiders to their chain of meaning, people removed from daily class struggles.In the words of a male former MP: If we're talking about equality and equality of representation, then how on earth can someone who's never walked outside an office environment really understand your factory worker, your fireman, your policeman?
This is articulation of a chain that positions 'ordinary' working-class people (although we note use of the male 'fireman', 'policeman') in opposition to the 'office worker'.It is given force through use of the verb 'walking', connoting a sense of the speaker as an active politician-worker embedded within the struggles of a working-class community and outside the sedentary 'office' bubble of establishment politics.
Another male former MPwho agreed with quotas but also understood the oppositionpositioned inequality in terms of the erosion of previous chains of solidarity and sociality within working-class communities, described as 'a sense of a community that feels increasingly left behind, losing out, and expressing itself in the structures of the Labour Party as well': If you were a miner, it might have been a shitty job and a dangerous one but you had pride, you had comradeship, the Royal Legion, or rugby to go to and to be a part of that culture.There was a sense of identity and a sense of future and their kids got houses and you got a pension.And your kids, if they made it to university, the handful that did, they got grants and there weren't bloody fees at the end of it.[Now] you can't get houses, there are no pensions, the few bright kids who make it through the system get clobbered.Jobs are insecure and there's no pride in them in that sort of working-class-solidarity sense.I think that's what's behind it all and that infects, and I think infects is the right term, it infects the milieu in which these AWSs operate.
A sense that Labour had been passive or inactive in the face of these profound social and economic inequalities, in the view of the former MP, had coloured responses to quotas, magnifying a sense of alienation from the party.In this extract we see the string of chain associations -'shitty job', 'comradeship', 'Royal Legion', 'rugby'unravelling and the 'infection' of antipathy setting in as a result.Likewise, a female party activist active in opposition to quotas articulated her position in terms of addressing economic inequality in communities as a first priority: 'Welfare benefits, we'll start there.Then we'll move towards a more equal society as we go up the scale.'For her, the use of quotas for elected representatives, far from addressing an issue of inequality, was itself a symptom and articulation of inequality, of an organization seeking equal gender relations but acting contrary to principles of equality in relation to the poor.
Introducing quotas, therefore, had the effect of bringing to the surface a range of previously subdued disaffections related to the key notion of equality, which in turn unravelled previous alliances and understandings.

| POLITICIZING THE ONTOLOGY OF ORGANIZATION: FAIRNESS
Notions of fairness were important empty signifiers, helping participants articulate the need for, and opposition to, gender quotas.Fairness was connected to procedural fairness, justice and the role of organizations (including governments) in intervening to redress opportunities and outcomes.
Opponents of positive action on gender shared stories about unfairness in the procedures and practices of the party that gained fresh significance in light of quotas.These were chiefly concerned with examples of 'safe' Labour seats where men in favour with the party leadership had been allowed to stand for selection, when an AWS should have been imposed.Opponents could therefore say they did not object to the principle of quotas but were concerned about implementation, making an association with fairness.In the words of one male former MP, 'If we are talking quotas, I've got no problem with that, but it must be done fairly.'A woman party member said: 'You're right it [candidate selection] is changing too slowly.It is unfair towards women.There's no question about that.'However, she then continued, describing how an 'unfair' system of quotas would undermine trust in party and the women selected.Such sentiments were widely shared and expressed, with one female quota supporter expressing concern that they were being used as a 'blunt instrument' that was damaging the sense of 'fairness' party members expected: 'On the surface you've got a very fair system which will work, but then I think it's being tainted by the way it's being adopted.I mean, where's the criteria?'For quota supporters, their introduction was related to reasserting 'fairness' in counter-distinction to alterative moral signifiers viewed as antithetical to the values of the organizationchiefly a failed form of meritocracy.As one male party official said, You know, it's not a meritocracy.It's flipped on its head.Our processes are not fair right now.If they were fair, we would have had a damn sight more women selected.
Politicizing gender for these officials and campaigners therefore meant challenging assumptions about the 'fairness' and associated 'merit' systems supposed to be at work in the party, a 'flipping' of the conventional logic held as synonymous with what it meant to belong to a party committed to fairness.
Fairness also articulated a sense of moral consistency amongst supporters of quotas, with argumentation offered that as Labour politicians stood for implementing fairness in society through legislation, it was inconsistent, even hypocritical, for these same legislators not to reflect the societies they represent.Gender therefore became a surrogate for more generalized fairness.In the words of one female MP: 'If you're making laws with just one section of society, you're bound to be making laws that are less welcome, less inclusive.' Here the actor of 'woman' performs a political function in disrupting a taken-for-granted notion of fairness, with fairness being vital to the process of making laws that work for a whole society.This notion of woman acting as surrogate for a more general politicizing of fairness was captured by one female AM, who, after introducing the notion of 'fairness', explained: It's not just with women but with other groups: how can you say you truly reflect the general population?I always think when people get really cross about [quotas] and say 'Oh it's terrible because, you know, Joe Bloggs, he's missing out on a seat', that actually it's not about the individual, it's about the collective isn't it?It's about the collective good, the common good.
She builds a chain of equivalence around the notion of fairness.The central actors are women but she soon expands her articulation to include 'the general population', because law affects everybody in a society.External to this chain are a generalized 'people' who 'get really cross' and a (male) figure of 'Joe Bloggs'.These people are positioned as missing the point: fairness, rooted in a moral system of the common good, needs to trump individual self-interest.
The notion of fairness for women is therefore placed in a chain that serves the purpose of generalizing a political project of fairness, politicizing gender, certainly, but also the party's sedimented assumption of fairness as core to its existence.
So far we have drawn out two empty signifiers politicized in intra-dependence with gender.This was ontological in the sense that it disrupted what it meant to belong to the organization.As equality and fairness were positioned as existential commitmentsthe very reason for the organization's beingdisrupting them necessarily entailed political contestation.Yet identifying and expanding on these ruptures does not bring us closer to conceptual clarity.
Next, therefore, we proceed to theorize two practices for pursuing desegregation: 'standing up' and 'walking with'.

| STANDING UP
Quota supporters explained that they saw only two options, either backing down or standing up to what they viewed as intransigence from constituency-level parties or, worse, outright sexist and misogynist attitudes.Standing up falls somewhere between an agonistic and antagonistic approach.It seeks conflict but agonistically, through the machinery and processes of liberal democratic institutions, such as internal party selection procedures with nominations, hustings and voting processes, and if opponents of quotas left the party to stand their own candidate, in a public election.Standing up entailed confrontationstanding up to be counted (literally) in an election where a woman would be the candidate.Yet this agonistic approach did sometimes spill over into antagonism.The decision for opponents of quotas to 'stand up' was not a straightforward choice, as they felt a sense of loyalty, knew that they would need to leave the organization to support an independent candidate and feared an antagonistic clash with former comrades.
Both opponents and proponents of gender quotas recounted provocations to stand up.Recalling a meeting convened to determine whether a certain constituency would have an AWS, a female former party official said: And [constituency officers] came in and went, 'We dare you.'And it was like, 'Come on then.'They just sat there, arms folded.So instantly they raised the mood.So me and [MP] snuck out of the meeting and went, 'We can't pick them.It would be awful.'And [another MP] … persuaded everybody in that meeting, said, 'No, you can't let them get away with it.We've got to make an example.'Which was exactly the right thing to do, although super painful.
To contextualize: it was normal practice at the time for the Welsh executive committee to call in officers of a constituency with a candidate vacancy for a meeting, where they would be asked to report on achievements in promoting the progression of women before a decision on AWS was made.This extract could describe the beginnings of a physical altercation rather than a democratic discussion, howeverwith provocations ('Come on then', 'We dare you'), a moment of indecision, and quota supporters deciding whether to back down or stand up.
The adjective 'brave' was used by a number of participants, suggesting that they knew these decisions could result in antagonistic confrontations, with attendant stresses and anxieties -'like a civil war, like splitting families' (male, former MP).The antagonistic language of battle was significant, with quotas described as a 'crusade' by one male former party official.This 'crusade' was cloaked in strong moral terms, as shaping 'this new Wales', with others talking of a 'responsibility' as the largest party in Wales -'if we didn't do it, no one would and nothing would change' (female party official).
The toll taken on both proponents and opponents was antagonistic.While opponents tended to talk about their campaigns in more liberating language, one male who defeated the party AWS candidate as an independent in an election and became a parliamentarian, described his experiences in terms of isolation, as people he regarded as comrades shunned him and behaved antagonistically.He recalled his experience of walking into the parliament chamber, greeted by 'two individuals sat on the Labour benches.They actually shouted at me and said, "We spit on you, traitor.We'll never forgive you."'The antagonistic act of spitting, this time literal rather than verbal, was also described by one quota supporter, who recalled giving a broadcast interview outside a party event and 'they were all there … spitting [at me]'.
The experience for women candidates selected on an AWS could certainly be antagonistic, with one describing the bullying of local party members a decade after the original selection.Another described veiled comments received about her accent and not dressing appropriately.A candidate who lost an election having been selected through AWS described the election count as akin to experiences of dealing with National Front activists, but under the glare of national media presence and a packed hall: 'People were pushing, they were shouting, you know, shouting unpleasant things.'For party officials, usually invisible actors in election campaigns, the impact was both embodied and antagonistic.
In the words of a male former official, a particular constituency battle became his 'whole life', with all other work, private and social commitments put on hold.Likewise, a woman official described a period of intense over-work: 'I just did it all myself'.Another male party official described sleepless nights followed by a period of illness after one local campaign.
Experiences of standing up were visceral for proponents and opponents, both sides working towards vindication of their sense of what it meant to be a party that stood for equality and fairness, but both also working towards the defeat of the other.Final satisfaction and understanding was often not reached for either side.

| WALKING WITH
This was a practice described in terms that acknowledged the need for disruption of ontological assumptions of fairness and equality, but in ways that sought to reach common understanding.We think it is important to acknowledge it as an alternative and subtler, more agonistic practice than standing up, although perhaps also one with less potential for immediate, dramatic and radical change.
We see possibility for walking with in the common language and identifications drawn on by participants on both sides of the debate.Although differences existed in terms of examples of breaches of fairness and equality, both sides appeared to register that these were essential principles.Furthermore, opponents of quotas acknowledged the twin problems of misogyny and sexism.One male, a former constituency officer, described the goal of securing more women's representation as a 'laudable ambition' executed 'clumsily' through AWSs, while others disapprovingly shared stories of local male councillors enacting protectionist and exclusionary practices that held women back from equal participation in intra-party processes.The stories told on both sides of the debate were strikingly similar.Both also remembered past arguments in largely negative terms, as bitter, unhealthy and corrosive, whose legacy remained in damaged relationships: memories of standing up and losing were strong for all.
In terms of practice, walking with was described as time-consuming and subtle work, of stepping towards another to seek understanding, while being willing to reflect on harsh critique: 'You spend, you know, time on the phone, time meeting them, sitting down with them', in the words of one male party official.These meetings, he said, could be tense and fraught but allowed all sides to articulate feelings and identifications: what it means to be a practitioner committed to equality and fairness and what that means in terms of practical actions.This work had a temporal dimension.By beginning discussions early, officials believed they reduced the risk of being labelled as 'outsiders'.It was also important, we were told, to maintain dialogue after a decision had been taken centrally:

It's the fall-out from the row that kills you. It's not the decision you've taken, necessarily. It's whether or not you're able to hold people together after the decision has been taken. ( Male former party official)
There is an acknowledgment here that 'rows' will happen, that people will fall out and tempers will be lost.There was always a danger present, therefore, that walking with would spill into an antagonistic standing up.Yet rather than viewing such conflict as a reason to stand up, it was drawn on as justification for walking with, for staying present with interlocutors after a decision had been taken, showing that you were committed to maintaining a partnership rooted in common values.This proximity to antagonism sometimes necessitated, in the accounts of officials, a slowing down of desegregation activity, a 'stepping away' within 'walking with'.Stepping away does not equate to giving up or not trying.Rather, it denotes a move that tests the readiness of a local party to engage and, based on clues gleaned from previous antagonistic experiences, temporarily backing off.
In the words of one male former MP: 'Sometimes you just can't win the argument, can't win.I tried in [constituency], couldn't win it.I tried in [constituency] and I backed off because you couldn't win.'The repeated refrain of this quote illustrates the struggle of walking with and a decision made to temporarily step away.Informing this was a pragmatic concern for keeping Labour representation in a seat -'the last thing we want is to lose the seat' was a familiar tropebut also of avoiding an association between AWS and pain or loss.
Walking with was therefore a subtle practice of pre-emptive engagement, maintaining a path of agonistic engagement that challenged gender norms and that drew on previous antagonistic experiences as a means of encouraging dialogue.But it was also a practice that involved a temporary backing off from AWS imposition when officials and senior politicians felt that a practice of walking with would be impossible in that place at that particular point in time.

| DISCUSSION
We offered two research questions and stated that we would address them empirically through interleaving agonistic theory and the gender de/segregation literature.We first asked: 'How does gender desegregation unfold in political and contested ways?' We have shown how radical forms of desegregation in organizations can be understood through seeing gender as a political practice (Mouffe, 2005;Pullen, 2006).In the case of politicized practice in our study, we see how quota implementation cannot be dissociated from other manifold identifications and practices of organization.This shows that dissociating notions of gender from lived practice seems naive, as they are one and the same thing: deeply embedded and rearticulated daily.Such an insight chimes with claims made in some of the literature on quotas, which emphasizes the distinction between changing the numbers of women in senior positions and changes to the everyday norms of organizations (Cockburn, 1989;Krook & Norris, 2014).While one may help lead to another, we cannot take such assumptions for granted and can instead view the desegregating of organizations as deeply political work.Hence proponents of quotas faced opposition from men and women (McDonald, 2016;Tiessen, 2004), recalling experiences of dealing with taken-for-granted sexism and misogyny (Ali & Syed, 2017;Johansson & Ringblom, 2017).Viewing desegregation as a matter of organizational language (Johansson & Ringblom, 2017), an instrumental win-win (Ness, 2012) or as simply requiring senior leadership commitment (French & Strachan, 2015), is limited and risks perpetuating a politics-as-usual approachhence the need felt by our participants for agonistic practice.While the quotas policy analysed here did have the commitment of senior leaders, which undoubtedly helped drive change, over-reliance on this aspect of desegregation could result in the neglect of deeper and more systemic contestations.Further, prioritizing senior leader commitment might also overlook grievances felt towards these leaders.
The political practice perspective helps us see the articulation of chains of equivalence for and against desegregation as crucial.It is unrealistic to seek a perfect compromise or a technical and procedural solution for 'smoothing over oppositions' (Islam et al., 2017, p. 16), as desegregation efforts will likely disrupt and challenge taken-for-granted views about the everyday values of an organizationsuch differences in Mouffian language are 'ineradicable' (Mouffe, 2013, p. 14).A sense of dispossession, injustice and hostility, whether directed towards an organization's leadership team, managers or other members, can become entangled with any form of radical change, serving as a further obstacle to women's equal participation, and progression, within an organization (Calás & Smircich, 1996).
Illuminating the political unfolding of gender desegregation, we have shown how gender's political practice occurs at the level of ontology, disrupting and contesting what it means to be within an organization.This is most obviously witnessed in our study in the two competing and perhaps irreconcilable perspectives of quota opponents feeling that equality and fairness were founding principles of the party but that these had only peripheral relevance to the representation of women in elected politics, whereas proponents felt that belief in these values was incompatible with a gender-segregated senior leadership cadre.It is intuitive that such signifiers are central to conceptions of being in an organization that claims to represent left and centre-left political views.In other contexts, the empty signifiers used to articulate such a sense of being might differ.One of the key propositions of post-foundational theory is that the world is political, in the sense of being grounded in an unavoidably contingent (S.Smolovi c Jones et al., 2016)and therefore contestedweb of signifiers and identifications.The empty signifiers of less formally political organizations will therefore also be hotly contested.Masculine working norms appearing as a common sense of an organization (Ali & Syed, 2017), combined with the additional pressures of reproductive labour that disproportionately fall on women (Raz & Tzruya, 2018), may even make contestations around desegregation in other organizations more complex and intense.
Our second research question asked: 'How might we learn from politicized experiences of desegregation to inform future attempts and initiatives?'We analysed two practices evident from our data analysis.'Standing up' meant approaching gender as a political practice, following the tenets of an agonistic approach, asserting difference and insisting on the generative role of conflict in forming new identifications and practices.Standing up was a bold approach that involved confronting one another's ontological sense of organization, being willing to sacrifice time and body to assert the justness of a particular position.Acts of standing up made visible embedded and previously taken-for-granted gender inequalities in a way that otherwise might have remained an unnoticed form of politics-asusual.In Mouffe's terms, standing up could be equated with the realignment of chains of equivalence, with opponents removed from within the chain to an antagonistic outside.Both sides remained committed to defeating the other through democratic means, an agonistic approach, yet the approach occasionally veered into antagonism, with opponents cast as enemies to be silenced or eradicated.
Two problematic issues with standing up persist.The first is empirical, in that candidates for parliamentary seats have been selected via an AWS without antagonism or real political contestation.We can discuss this as connected to statements made to us by quota proponents that contestation was more likely in poor and former industrial areas, less likely in urban seats with more middle-class members.Rather than claiming this as an issue of economic determinism, we prefer a more agential explanationthat opponents of quotas were expressing general dissatisfaction with a neoliberal consensus that left many people in their towns bereft of hope.Unfortunately, these legitimate concerns crossed over into a general suspicion of centrally driven change proposalsand sometimes into sexism and misogyny.We see here the germ of a problem that continues to haunt progressive politics, that of being cast as a liberal elite, interested in pursuing a 'woke' cultural agenda at the expense of a more 'authentic' account of the concerns of the working class, a similar discursive division that has fuelled the rise of populist right movements globally.To further the analogy, many of our opponent participants spoke to one degree or another of a loss of control over their organizationusually framed in terms of a democratic deficit and the centralizing of power amongst the organization's leadership elite, similar refrains to those that have characterized populist movements globally.Mouffe (2005Mouffe ( , 2009) again provides us with some guidance in how to address such discursive positioning through her refusal to conceptually privilege one progressive identification over another, instead insisting on the primacy of democratic participation in maintaining an equivalence between a range of demands.Such an approach does not mean abandoning commitments to gender desegregationrather, equality initiatives might be accepted as more legitimate if proposed within a vibrant democratic milieu.
The second problem is that standing up will be challenging in many organizations, particularly those with a profit motive or with statutory duties to fulfil, as the disruption to work can be significant.Quotas are very widely used in politics around the world; they are however much rarer in conventional, individual workplaces (Noon, 2012).Capital-P Political organizations such as parties are certainly more accustomed to political debate and conflict, yet also, similarly to more conventional organizations, are wary of being viewed as too riven by conflict.Perhaps one lesson from this study is that although conflict seemed ferocious at times, the bottom-line damage caused can yet be viewed as marginalfor example, with Labour largely maintaining hegemony in Welsh politics since implementing its desegregation initiatives.Avoiding conflict and disruption to protect the bottom line may be a misplaced concern, therefore, particularly at the expense of radically improving the diversity of an organization.However, we are wary of generalizing from our case to other contexts and acknowledge the need for further, more qualitative research.This brings us finally to walking with.This practice is more clearly rooted in agonism, facing up to poor gender representation and confronting underlying causes in a way that seeks to keep conflict within the boundaries of the organization, rather than allowing it to spill out into public displays and acts of antagonism.That said, we are reluctant to view such an approach as an easy solution.We hold open the possibility that many organizations may require a passage through standing up, to surface schisms and experience the embodied hardships generated by antagonistic clashes.This is shown in part in the tendency of participants to reference past negative experiences of antagonism as justification for seeking an alternative approach.Hence walking with could sometimes lead to a temporary 'stepping away', which in some ways has more in common with Mouffe's (2005) notion of politics, or, in our rewording, 'politics-as-usual'.It was a position informed by bitter experiences, a felt need to regroup and reflect before further action.Mouffe's (2005) equation of politics to the taken-for-granted and the ontic seems like an unsatisfactory description when politics is pursued as a conscious and deliberate strategy by people who have lived through a more confrontational approach.Rather, we note a more nuanced approach in practice, of experimenting and living through conflictual engagements and drawing on such experiences to better judge how and where to pursue desegregation initiatives.

| CONCLUSION
We began this article with the presentation of a problem, the lack of political bite to studies of desegregation, which have tended to elide the messy and conflictual processes of implementing initiatives.We developed a politicized practice view of gender desegregation, analysed as offering a radical challenge to the politics-as-usual status quo.
Doing so means decentring the category of gender, approaching it as a practice shaped by social norms but whose status and primacy can be challenged and rearticulated.Our political view of gender desegregation requires scholars and practitioners to grapple with gender as enmeshed within an ontology of organizationof women's inequality either being regarded as peripheral to organizational ethics or central to them.Identifications with commonly held but differentially articulated notions such as fairness and equality, we state, can only be challenged through conflictually engaging with people's foundational reasons for identifying with an organization, the kind of work that constitutes both 'standing up' and 'walking with'.Such work cuts all ways, requiring a degree of critical self-reflection and commitment to democratic forms of engagement from multiple organizational subjects and stakeholders prepared to engage with the alienations perpetuated through an organization's historical practice: in our case of women's experiences of marginalization and also an organization's grassroots feeling ignored and disempowered.
Finally, our study demonstrates the centrality of conflict to progressive forms of organizational change.Without such a conflictual approach to desegregation, it is unlikely that meaningful change would have occurred, with a politics-asusual approach perhaps eventually leading to a gradual and inadequate reform of an organization lethargically taking its lead from broader social norms.Conflict can be painful but, we posit, is also vital.