‘To the surprise of absolutely no one’: Gendered political leadership change in Northern Ireland

The office of political party leader remains one that women rarely occupy. In the largest comparative study of party leadership to date, only 10.8% were women. One region which has made significant advances in this area is Northern Ireland. Since 2015, the two largest parties, Sinn Féin and the Democratic Unionist Party, have experienced a rapid feminisation of their leadership. Such a development is particularly remarkable given Northern Ireland’s historically poor record on gender equality. This article explores the puzzle of gendered leadership change in Northern Ireland to reveal that the transition was primarily facilitated through the parties’ informal practices rather than embedded structural change. In doing so, it demonstrates the relative importance of party- and system-level factors on women’s political presence. As a power-sharing democracy, this case also provides comparative insights for those interested in addressing persistently low levels of female representation in post-conflict settings.


Introduction
The office of political party leader remains one that women rarely occupy. In the largest comparative study of party leadership to date, only 10.8% were women (Wauters and Pilet, 2015). As such, the 'law of increasing disproportion' holds true in most cases -that is, the proportion of women in party offices is inversely related to the importance of the position (Putnam, 1976). One case where this law has been broken, however, is Northern Ireland. Since 2015, the region has witnessed a feminisation of party leadership, with the two largest parties, Sinn Féin and the Democratic Unionist Party (DUP), electing female leaders. 1 Arlene Foster was elected DUP leader in February 2015, while Michelle O'Neill and Mary Lou McDonald were appointed Sinn Féin 'leader in the north' and President in January 2017 and February 2018, respectively. 2 Such a development is all the more remarkable given Northern Ireland's poor record on political gender equality and resistance to broader national and international trends affecting women's rights (Galligan, 2013;Moon et al., 2019). 3 Over 30 years of violent conflict has also instilled an enduring masculinised political culture within Northern Ireland's power-sharing institutions (Mackay and Murtagh, 2019).
This article addresses the puzzle of gendered leadership change in Northern Ireland, providing an original assessment of the factors underpinning the selections of Foster, O'Neill and McDonald. These include the primary endogenous and exogenous factors identified by scholarship on women's political presence: party ideology; party organisation; party performance; institutional context; and post-conflict opportunities. These factors are outlined below before being systematically examined in respect to the Northern Ireland case. Our analysis is based on a range of primary and secondary sources, including party constitutions and rulebooks, news publications, and existing scholarship on the DUP and Sinn Féin. By combining these sources, we are able to provide both a detailed account of the selection of the three women leaders and, most importantly, interpret the meaning of decisions taken by two highly secretive organisations (see Bevir and Rhodes, 2003). While these five factors feature in different ways for each party, we find that the outcomes in all three cases owe more to the parties' informal practices and ad hoc strategies than formal modifications or deeply embedded structural change at the system level. As a result, we conclude that the Northern Ireland 'success story' should be celebrated with caution. If the parties' strategic priorities or organisational approaches change, then such rapid progress could be easily reversed. The Northern Ireland case, therefore, provides a wider lesson for those interested in why gendered leadership change occurs and how such advances in political gender equality can be sustained.
Scholars interested in gender equality in other post-conflict societies or multi-level settings should also find the Northern Irish case to be instructive. Prolonged violent conflict during the 'Troubles' not only influenced gender dynamics in the region, but also provided the backdrop in which the DUP and Sinn Féin emerged as political parties (Braniff and Whiting, 2017). The institutional solution to that conflict, in the form of a consociational power-sharing arrangement set within the wider devolved political settlement of the United Kingdom, also saw the DUP and Sinn Féin overtake their more moderate rivals to become the largest parties within their respective communal blocs of unionism and nationalism in 2003. An in-depth analysis of the Northern Irish case, therefore, provides important insights for future comparative research on gendered party leadership change in other similar cases.

Explaining gendered leadership change
Understanding the gendered nature of political parties and institutions, including the dynamics that structure and interact with the process of leadership selection, requires acknowledging how both formal and informal rules and cultures can hinder or facilitate women's political representation (Annesley and Gains, 2010;Bjarnegård and Kenny, 2016;Kenny and Verge, 2016;Krook, 2010). Helmke and Levitsky (2004: 727) define these informal dynamics as 'socially shared rules, usually unwritten, that are created, communicated, and enforced outside of officially sanctioned channels'. For researchers, informal norms and practices are harder to uncover due to them being the unquestioned way of operating and seen as 'natural and immutable', if individuals are even aware of them (Chappell and Waylen, 2013: 600). In contrast, formal 'rules and procedures . . . are created, communicated and enforced through channels widely accepted as official', such as law, regulation, and in the case of political parties, the recognised constitution and party rule book (Helmke and Levitsky, 2004: 727). It is only by examining both formal and informal norms that it becomes possible to gain a clearer insight into why institutional reform does not necessarily result in the outcomes anticipated. For example, while formal processes may be in place at party level to provide opportunities for greater gender equality, constraints on women's agency are typically entrenched in informal norms and practices that prop up masculinised political cultures (Childs and Krook, 2013;Verge and de la Fuente, 2014). Research on institutional change and gender equality notes how the informal may hinder reform in some cases, yet in other instances can strengthen and support the formal institutions they interact with (Azari and Smith, 2012;Grzymala-Busse, 2010). Therefore, it is not only important to examine the rules, norms and procedures at both a formal and informal level, but the interplay between the two. Research focusing on the gendered dimensions of political opportunity structures is, therefore, required to look beyond the codified party rules to chart women's access to high office and comprehend the interaction between formal and informal dimensions (O'Brien, 2015;Sahin-Mencutek, 2016;Verge and Claveria, 2018).
Alongside an awareness of the hidden life of institutions, any analysis of gendered opportunity structures also requires sensitivity to context (Kenny and Mackay, 2009). Comparative research on female party leadership reveals significant variation in women's access to positions of power and accounts for the impact that institutional and cultural variations may have (Caul, 2001;Kittilson, 2006;O'Brien, 2015). Political parties are subject to various endogenous and exogenous factors that can either restrict or open up women's access to positions of leadership. In-depth context-specific analysis provides an opportunity to interpret the role of historical, social and institutional context in shaping gendered pathways and also account for how these evolve over time (Kenny, 2013: 62; see also Lovenduski, 2005).
Existing scholarship on women's political presence identifies common factors that intersect to shape women's access to positions of political power. Before turning to the case study of leadership change in Northern Ireland, we outline those factors which are both endogenous and exogenous to political parties. Taking endogenous factors first, party ideology and organisation are worthy of scrutiny.
In terms of ideology, left-wing political parties have traditionally been presented as more receptive to calls for greater gender equality and interventionist measures to facilitate the selection of women candidates than those on the right (O'Neill and Stewart, 2009). However, a classic left-right distinction obscures other ideological factors and, therefore, fails to sufficiently explain women's political recruitment (Norris, 1995;O'Brien, 2015). The influence of party ideology is further complicated in cases with ethno-national divisions, as in Northern Ireland. Research on the Kurdish Party of Turkey, for instance, demonstrates that while its left-wing ideology was a contributing factor to women assuming positions of power and the adoption of gender quotas, its association with a wider ethno-national movement also facilitated women-friendly opportunity structures (Sahin-Mencutek, 2016). The relationship between nationalist parties and women's political presence and mobilisation is, therefore, multifaceted and complex.
Scholarship on nationalism and gender has traditionally focused on the incompatibility of gender equality with nationalist agendas (Enloe, 2014). Ethno-nationalist discourses are accused of framing men as 'protectors of the nation', who publicly serve this role within military and political organisations, while women are assigned maternal and domestic roles, restricting their access to the public sphere (see Anthias and Yuval-Davis, 1989). The construction and persistence of these gender binaries has been used to explain the dearth of women entering the political arena in Northern Ireland (Ashe, 2007). Research exploring the relationship between nationalism and gender equality also reveals a diversity of experiences, from women reporting feelings of alienation and hostility, to participation and affiliation with national projects (Vickers, 2006). Identifying the factors that have influenced such variation is, therefore, fundamental in understanding how nationalism can limit, or facilitate, women's access to political leadership.
Second, party organisation is a significant endogenous factor for women's political presence. Political parties are gendered institutions where both intra-and inter-party behaviours affect male and female politicians differently (O'Brien, 2015(O'Brien, : 1036. It is, therefore, not possible to fully comprehend women's participation and presence without assessing party-level rules and structures (Kenny and Verge, 2016). One notable trend within contemporary political parties is the increased concentration of power and authority in the hands of central party elites (Poguntke and Webb, 2007). Crucially, such centralization in respect of candidate selection has been deemed conducive to achieving higher levels of gender diversity, as it enables party elites to more easily and effectively implement equality procedures and initiatives, such as quotas (Norris, 1995). Comparative research on candidate selection has also found more exclusive selectorates to be advantageous for under-represented groups, including women (Rahat et al., 2008). In terms of leadership selection, however, Wauters and Pilet (2015) caution against viewing that process in the same way as candidate selection. They find little evidence to support the claim that exclusive selectorates are more beneficial for the selection of female party leaders. Instead, they conclude that it is not the institutional rules that matter but the structural context in which parties operate and the extent to which they are generally 'open' to women (Wauters and Pilet, 2015;89-90). Again, this reminds us that focusing solely on the formal processes governing selection -overlooking other aspects, including informal organisational norms and cultural practices -will provide only a partial picture of a complex reality. While some recent organisational changes by the DUP and Sinn Féin have been identified as significant for gender equality (Braniff and Whiting, 2015;Matthews, 2014), the role of party rules in facilitating gendered political change in Northern Ireland remains understudied.
The impact of endogenous factors such as party ideology and organisation can then play an influential role in determining women's access to, and promotion within, the formal political arena. However, such party-level analysis also needs to consider the impact of external dynamics. These exogenous factors, which include party performance, institutional context and conflict, are discussed below.
First, party performance is presented as a key exogenous factor in determining women's access to party leadership. A strong electoral performance increases the appeal of the leadership post and reduces the willingness of selectorates to deviate from the 'usual mould' in terms of political leadership. O'Brien (2015) states that existing patterns of leadership selection are more likely to persist when a party possesses a 'winning strategy'. Electoral failure, on the other hand, makes the position of leader less desirable and may encourage parties to change tack (Beckwith, 2015). Opportunity structures are, therefore, more likely to 'open up' to women in minor parties, those in opposition and those losing seat-or vote-share (Bashevkin, 1993;O'Neill and Stewart, 2009). Previous gender equality initiatives by parties have also been related to a broader strategy to alter their 'image' in the eyes of the electorate and, as a result, boost their levels of support (Childs and Webb, 2012). Selecting a female leader can represent a visible break from a 'toxic' male-dominated past and symbolise a 'new direction' for a party (Murray, 2010). Gendered leadership change in the 'Männerparteien' (Mudde, 2007) of the radical right, for example, is credited with broadening their electoral appeal by symbolising a distinct change in leadership style and presenting a 'softer' party image (Mayer, 2013). In a more acute sense, moments of party scandal or crisis can also facilitate women's selection as party leaders, framing women, who are more often perceived as operating 'above politics', as ideal candidates to reconcile differences or oversee internal 'house-keeping' (Trimble and Arscott, 2008).
Political systems are also an important exogenous factor in determining the likelihood of women occupying executive positions. Political parties do not operate in a vacuum but within institutional structures, which themselves can either hinder or facilitate women's political presence. For example, women are more likely to be elected to the position of party leader where multiple actors share in political power and placed in executive positions where responsibilities are dispersed (Jalalzai and Krook, 2010). While instituting a form of consociational power-sharing, the Northern Ireland Assembly also operates within the United Kingdom's devolved political system. Sinn Féin, while not taking its seats in Westminster, is also represented in the Republic of Ireland. For parties operating in multi-level settings, where several 'nodes' of power exist, male political elites are less reluctant (or more willing) to support a woman for party leader (Beckwith, 2015: 722).
Finally, the impact of conflict on women's political presence has been multifaceted. First, the utilisation of gendered stereotypes is particularly evident following episodes of violent conflict. Within traditional gender roles, women are considered to be 'unifiers' and 'nurturers' of the family and the nation (Campus, 2013). In some post-conflict cases, female leaders have also been perceived as 'political outsiders' untainted by corruption or abuse of power (Jalalzai, 2008: 212). Such gendered ideologies have played an important role in furthering women's political careers; Ellen Johnson Sirleaf's election in postconflict Liberia, making her the first female head of state in Africa, is cited as one notable example (see Adams, 2008). Second, the creation or reform of political institutions following conflict can provide opportunities to embed gender equality measures within 'new' structures and wider citizenship rights. On an international level, since 2000 the Women, Peace and Security Agenda (formalised in various United Nations Security Council resolutions) was seen as an important milestone for women's inclusion in postconflict institutions (Thomson, 2018). However, efforts of 'gender mainstreaming' by the international community are critiqued for being 'grafted on' to existing political structures and, therefore, fail to adequately capture the institutional inequalities that enforce power relations (Willett, 2010). More specifically, a tension exists between powersharing as a dominant approach to post-conflict institutional design and the inclusion of women and provisions for gender equality (Mackay and Murtagh, 2019).
Existing research demonstrates how understanding gendered leadership change requires acknowledging the political, social and historical influences on both formal and informal party rules, norms and behaviours. For researchers concerned with the gendered nature of political institutions, informal elements, which can obstruct or facilitate women's political recruitment, are more difficult to unearth. As Norris (1997: 8) notes, the journey to political power typically involves 'shadowy pathways'. Using the framework of endogenous and exogenous factors outlined above, discussion now turns to the case study of gendered leadership change in Northern Ireland to shed light on these important elements.

Party ideology
The distinction between right-and left-wing parties in respect of their record on gender equality exists in Northern Ireland to some extent. For example, Sinn Féin, which sits on the left of the political spectrum, has had the highest level of women elected to Stormont in every election since 1998. Although not specified in the party constitution, Sinn Féin places an informal commitment to achieving a minimum target of 30% for female candidates contesting winnable seats, overturning local decisions to fulfil its quota if necessary (Matthews, 2014: 626). The DUP on the other hand is situated on the political right and has historically suffered from a dearth of women as party representatives and in leadership positions. Indeed, from its formation in 1971 until the end of the twentieth century, the promotion of women in the DUP was described as being through the personal connections of blood or marriage (Cochrane, 1997: 47-48). In response to this poor track record, the DUP introduced a centralised selection process in 2011 which allowed the party to by-pass traditional, male-dominated, branch-level politics (Braniff and Whiting, 2015: 102). 4 As a result, it returned its highest number of female candidates in the 2011 Assembly election and became the party with the second highest number of female MLAs, a position maintained in subsequent contests (Tonge et al., 2014). The growth of female elected representatives and Foster's appointment as leader, therefore, demonstrates how left/right distinctions do not adequately explain the evolving opportunity structures impacting women's access to party, and national, leadership positions (Norris, 1997;O'Brien, 2015). In this instance, changes to party policy through the centralisation of selection, rather than ideology, were crucial in promoting women within the DUP.
The ethno-national dimension also dilutes left/right distinctions in Northern Ireland and demonstrates the importance of considering different nationalist identities, their relationship to the state and proximity to power. For example, while women of dominant national groups tend to rely on state structures and institutions to protect their position and identity, women of minority national groups have historically taken more personal responsibility, making their active involvement with such national projects more common (Vickers, 2006). In the Northern Ireland context, the DUP represents the dominant British unionist community. Historically, protecting unionist interests occurred through state institutions and structures such as political parties, policing and Protestant fraternal organisations. Traditionally, these organisations were led, and dominated, by men and served to reinforce unionism through patriarchal networks (Ashe and McCluskey, 2015). Where women did organise, this was very much separate, and subordinate to, unionist men (Hennessey et al., 2019). The DUP's formation reflected the desire for a more assertive political force to protect Protestant values and defend Northern Ireland's place in the Union. As party leader, Ian Paisley's fusion of politics, evangelicalism and ethno-nationalism underpinned much of the DUP's appeal which simultaneously consolidated a social conservatism that restricted women's access to political or public office (see Racioppi and Sullivan See, 2001: 98). Such gendered assumptions continue to be reinforced by a persistent social conservatism within the DUP. Shortly after Foster became party leader and First Minister, DUP MLA Edwin Poots commented, it would be the 'second most important job that she will ever take on. Her most important job has been, and will remain, that of a wife, mother and daughter' (BBC News, 2016). In a similar vein, MP Sammy Wilson expressed admiration for his new leader's handling of political pressures 'along with all the domestic things she has to do' (Meredith, 2017). Despite Foster's rise to the highest political position in both her party and the state, the gendered divisions between public and private spheres persisted.
In resistance to the legitimacy of state security and political apparatus, minority nationalist positions on the other hand tend to take more responsibility for protecting their national identity, making female involvement in wider mobilisation strategies more likely (Sahin-Mencutek, 2016: 471;Vickers, 2006: 92). 5 Therefore, in contrast to unionism, the celebration of women as political activists and 'freedom fighters' fits more comfortably within nationalist and republican discourse (Rolston, 2018). These mobilisation networks across the republican movement allowed historic female figures such as Dolours Price and Martina Anderson to emerge across the spheres of politics, civil rights and armed resistance. For Sinn Féin, the promotion and active involvement of such women fits their wider equality agenda ('Ireland of Equals') that accompanied a pragmatic shift away from the armed struggle for Irish freedom towards gaining equality for nationalists and republicans within the North of Ireland. Gender plays a central role in this equality discourse, as stated by Gerry Adams in his final address as Sinn Féin President: 'We cannot have the freedom of Ireland without the freedom of women' (Irish Times, 2017). This historical context allows Sinn Féin to celebrate notable women in the republican movement alongside the current 'equality for all' agenda.
The wedding of gender equality to a wider political strategy makes the changes in Sinn Féin's leadership unsurprising, and in some ways inevitable. For the DUP, however, the appointment of Arlene Foster was less predictable, particularly considering the continuing levels of social conservatism within the party. The diverse interactions between nationalist projects, left/right ideologies and women's political mobilisation demonstrate the need to go beyond examining ethno-nationalist parties as one distinct category and highlight the importance of understanding the historical context in which political parties emerge, particularly on party culture and organisation.

Party organisation
Each of Foster, O'Neill and McDonald emerged 'victorious' from distinctly noncompetitive processes, conceived and managed by the party centre. Accounting for the party organisational dynamics which enabled these coronations is central to understanding gendered leadership change in Northern Ireland.
On paper, Sinn Féin's procedure for leadership selection is inclusive and decentralised, with a low threshold for candidacy and selection by delegates at conference (Matthews, 2016). 6 These qualities could conceivably frustrate efforts by the party centre to control the process and dictate outcomes. If the goal of the party centre is to select a female leader, the formal process neither precludes men from standing nor compels delegates to vote for a woman candidate in a competitive contest. Leaving such a significant appointment to chance, however, does not tally with existing accounts of how Sinn Féin conducts itself. The party has a long-established reputation for possessing a top-down culture, including strict central management of internal selection processes (Matthews, 2016;Whiting, 2016). One account describes Sinn Féin as being 'run on Bolshevik lines with a politburo type leadership enforced with army style discipline' (Gormley-Heenan, 2007: 61); traits which have been attributed to its origins as the subordinate political wing of a militant group, the Irish Republican Army (IRA), and which have been acknowledged by party representatives, including Mary Lou McDonald herself (Finn, 2018;Ó Broin, 2009: 304-305).
Sinn Féin's reputed predilection for long-term strategizing also points towards the O'Neill and McDonald successions being carefully choreographed, handpicked by the central party leadership in advance of the 'official' selection process. McDonald was, for example, long touted as Adams' probable successor and the selections of both women were presented by several Sinn Féin officials as forming part of 'a 10 year project of generational change', suggesting a clearly formulated plan (BBC News, 2017). Both candidates were also endorsed by their predecessors. Adams, for example, openly declared that he had personally 'recommended' O'Neill's candidacy to the party's executive, a recommendation subsequently ratified and announced at a press conference prior to the selection convention (Emerson, 2017). O'Neill also explained how, 'Martin and Gerry spoke to me and asked me to take on the position' (Clifford, 2017). The irony of two coronations in a republican party was not lost on Sinn Féin's critics, who also raised questions about McDonald and O'Neill's authority, on account of their being 'handpicked' by other leadership figures (see Clifford, 2017;McIntyre, 2017;O'Hanlon, 2018).
If Sinn Féin's organisational culture facilitated the appointment of women to leadership positions, a similar dynamic was also evident in Arlene Foster's selection as DUP leader, albeit in a more circuitous way and with a stronger formal institutional element in play. The DUP adopts a highly exclusive approach to leadership selection, with the leader selected by the party's parliamentary group before being ratified by its executive body (Matthews, 2016). This process has remained unchanged over the party's lifetime and is in keeping with widely held perceptions of its internal structure, which suggest a centralised and top-down order (Tonge et al., 2014). Recent revelations from the public inquiry into the 2016 Renewable Heat Incentive debacle shed even further light on the DUP's inner workings, revealing a powerful central party officer team exercising clear and unrivalled decision-making authority (McBride, 2019). It is likely then that the party's approach to leadership selection is even more exclusive than the formal process outlined in its rulebook, with the exercise carefully managed by party headquarters.
The DUP's history of leadership change also suggests that Foster's appointment was not serendipitous. The only other instance of leadership succession, with Peter Robinson replacing Ian Paisley in 2008, bears clear similarities to the Robinson-Foster transition: a long-mooted candidate anointed by the outgoing leader in a carefully stage-managed process. The one notable difference with Foster's selection, however, concerns the deferral of a male rival for the leadership. Foster formed one part of Robinson's favoured plan for a 'split leadership ticket', with Nigel Dodds, the DUP's deputy leader, assuming the position of leader and nominating Foster as First Minister (Clarke, 2015). With Dodds eventually declining to stand for the leadership, the path was cleared for Foster to assume leadership of both offices. While the succession plan may not have unfolded quite as the central party leadership originally intended, they remained in control of the process throughout, with Dodds and Foster the only viable candidates.
The DUP and Sinn Féin's conception of leadership, in an organisational sense, also appears significant when accounting for gendered leadership change. While both are unquestionably monolithic, top-down, hyper-centralised parties, they are not solely led or dominated by an individual leader. Rather they practice a form of collective leadership, with a dominant leadership group; in Sinn Féin's case a party executive and in the DUP's a party officer team (Matthews, 2016). As such, the existence of other prominent powerbrokers may have reduced the perceived 'costs' or 'risk' of selecting a female leader. When discussing her appointment, O'Neill, for example, stressed the importance of 'collective leadership' in Sinn Féin. Tellingly she offered this as reassurance to those with reservations about her experience and suitability for the role, promising to work closely with other 'leaders' and senior members in the party (Breen, 2017). The presence of other powerful leadership figures likely proved a propitious organisational context for the promotion of women within both parties.
On the evidence above, the significance of party organisation in the selection of women leaders in Northern Ireland is clear. All three cases were examples of 'managed intra-party democracy', where, in a process of 'precursory delegation', selection was managed by another internal agent or agency before it reached the 'official' selectorate (Aylott and Bolin, 2017). Indeed, with powerful leadership figures determining outcomes in advance, what essentially occurred was an Irish version of 'dedazo'. 7 As Langston (2016) demonstrates, the practice of an outgoing leader controlling succession is intended to minimise intra-regime conflict and maintain a party's hegemonic position within the party system. The selection processes of Sinn Féin and the DUP were likely motivated by similar concerns of control and dominance. The one unknown, of course, is the extent to which the unofficial selectorates in both parties were driven by concerns of gender; did the DUP and Sinn Féin central leaderships purposively select women? An evaluation of certain exogenous factors underpinning leadership feminisation can help address this question.

Party performance
Sinn Féin has essentially enjoyed an unbroken run of electoral success in Northern Ireland since the establishment of devolved power-sharing in 1998, emerging as the dominant party in the nationalist bloc. Prior to the appointment of McDonald and O'Neill the overwhelming impression was therefore of a party either bolstering its dominant position (in the case of European and local government contests) or breaking records (as with the 2017 Westminster election). However, while Sinn Féin possessed a winning electoral strategy in Northern Ireland, its performance in the Republic of Ireland was much more modest, with returns falling short of expectations. Crucially, one of the foremost explanations for Sinn Féin's failure to engineer a southern electoral breakthrough of real significance at national level concerned Gerry Adams' leadership and the party's historical ties with the IRA. In stark contrast to his star power north of the border, Adams was widely regarded as an electoral handicap for Sinn Féin in the south. Much was made of his weak grasp of policy detail on the campaign trail, not least Sinn Féin's fiscal programme; a particularly acute failing in a political system focused primarily on the economy and austerity. The latter years of Adams' tenure were also marked by several controversial episodes concerning both his family and, more significantly, his alleged involvement in historic IRA operations. In an era of intense electoral focus on party leaders (Poguntke and Webb, 2007), the controversy surrounding Adams could be regarded as damaging for Sinn Féin more generally, placing a ceiling on its support in the south.
In direct contrast to Adams, McDonald was deemed a potential electoral asset for Sinn Féin. A prominent theme in coverage of her accession was that it formed part of a strategy to broaden her party's appeal in the south. As with other cases of party renewal or 'modernisation', the selection of a woman leader contributed to the 'detoxification' of the party 'brand' in the eyes of a traditionally sceptical electorate. 8 Such a reading tallies with how Sinn Féin itself framed the leadership change, with senior figures consistently referring to McDonald's appointment as one step in a long-term process of 'renewal' or 'generational change' (see BBC News, 2018;McGee, 2018). A former Fianna Fáil member, middle-class, universityeducated, Dubliner, with no IRA record -McDonald was Sinn Féin's 'Heineken politician' (O'Doherty, 2017), who could appeal to a middle Ireland 'turned off by the whiff of cordite from the party's old IRA links' (O'Dowd, 2017). As one Sinn Féin activist explained at the time, the crucial difference between McDonald and Adams is that '[she] will have the space to focus without distraction on other issues such as housing and health, without having to have one eye looking at the past' (McGee, 2018). Adams' replacement by McDonald, therefore, represented the removal of one especially virulent toxin and the continuation of a more wholesale decontamination and modernisation effort by Sinn Féin in the south.
Sinn Féin's strategy of using leadership change to signify party 'renewal' was also evident in O'Neill's appointment as 'leader in the north'. As with McDonald, much was made of O'Neill's non-militant background, situating her selection within a broader process of 'generational change' (BBC News, 2017). Commentators described O'Neill as a 'new leader for a new era' and a representative of a 'younger generation of post-conflict republicans' (Manley, 2017). While she boasted strong republican roots, O'Neill had no IRA record, hailing from the 'post-ceasefire' cadre of Sinn Féin activists. These credentials were all the more obvious when set against the military record of her predecessor, Martin McGuinness, and main rival for the northern leadership, Conor Murphy, a former IRA volunteer and prisoner. Reports suggest that Murphy's IRA credentials jarred with Sinn Féin's plans for the leadership transition to symbolise a 'changing of the guard' (Moriarty, 2017). As one republican commentator explained at the time, 'Leadership is not just being passed from the old to the young, but also from the military to the civilian' (McIntyre, 2017).
In addition to their non-militant backgrounds, gender featured prominently in the presentation and reception of both O'Neill and McDonald's selections. In her acceptance speech, McDonald exclaimed that Sinn Féin is 'probably the most exemplary party when it comes to girl power at this stage in Irish politics' (Roberts, 2018), citing her and O'Neill's leadership as proof positive. Both McDonald and O'Neill's addresses at Sinn Féin's 2018 party conference, their first as a leadership duo, contained standalone sections on 'a new Ireland for women', while the event itself closed with the Aretha Franklin anthem, 'Sisters are doing it for themselves' (Finn, 2018). Again, the party's southern electoral strategy is key here. The issue of women's rights was high on the Irish political agenda at the time, with an imminent referendum on removing a constitutional ban on abortion. The selection of two women leaders, therefore, appeared in-step with the increasingly progressive zeitgeist of Irish politics and positioned Sinn Féin to attract a greater level of support from women, a demographic with whom it traditionally struggled to appeal. If leadership change formed part of a Sinn Féin modernisation strategy, to decontaminate itself and open up new electoral frontiers in the process, then it was clearly both non-militant and gendered in nature.
There is a less obvious electoral or strategic explanation for Arlene Foster's selection as DUP leader. As the dominant party in the unionist bloc and the largest party in Northern Ireland, the DUP appeared in rude electoral health. On closer inspection, however, there were signs, prior to Foster's appointment, that the DUP's electoral dominance was weakening, as it faced challenges from rival unionist parties and other established and emergent cross-community parties. In the 2014 local government elections, for example, the DUP's vote-share dropped by 4.1%, with the chief beneficiaries being the Traditional Unionist Voice (TUV) and Ulster Unionist Party (UUP). When accounting for Foster's accession to the DUP leadership, however, a more significant feature of that election was the emergence of a new party, NI21, which stood on a socially liberal and left-of-centre unionist platform. This new vehicle was both emblematic of the seeming direction of travel in Northern Irish politics at the time -towards a more 'normalised' or moderate space (Garry et al., 2017) -and a manifestation of a political strategy advocated by DUP leader Peter Robinson from 2011 onwards (Tonge and et al., 2014: 180-186). To remain electorally dominant and, by extension, preserve Northern Ireland's place within the union, Robinson advocated that the DUP 'readjust', to attract votes from the growing Catholic community in Northern Ireland. To do so meant espousing a more civic, moderate, unionist programme, akin to that of NI21, renewing (if not reimagining) the DUP in the eyes of 'floating' Catholic voters in Northern Ireland.
Given her virtual coronation by Robinson, the question is whether Arlene Foster was viewed as the candidate best placed to continue that project to modernise, or rather pragmatise, the DUP, to detoxify its brand and broaden its appeal? Foster's political background is significant here. As a former UUP member, she was associated with a less hard-line variant of unionism. Contrary to her subsequent actions as party leader, she was believed to occupy the DUP's moderate wing and, along with other high-profile 'UUP arrivistes', was regarded as part of a modernising vanguard in senior positions of leadership (Gomez and Tonge, 2016;Irish Times, 2015). Her acceptance speech as party leader was certainly Robinsonian in tone, pitch and content, where, in a clear nod to the changing demographics of Northern Ireland, she spoke of her determination to make the DUP 'home' to 'people of all religious persuasions, from all social backgrounds' (The Guardian, 2015).
If Foster's selection did form part of a continued modernisation effort by the DUP, then gender did not constitute an intentional element of that strategy. Neither the party nor Foster herself focused -or played -on her gender to any significant degree. When pressed on her views on political gender equality in the past, Foster emphasised the importance of merit for promotion (Kane, 2015) and, shortly after her appointment, declared that, 'leadership transcends masculine or feminine labels' (Ganiel, 2016). The aforementioned rather ham-fisted musings of other DUP figures on the significance of Foster's gender also suggests that a well-considered strategy to capitalise on her identity as a woman was not in place. The central party leadership's original plan for a Dodds-Foster dual leadership also further undermines the idea. Interpreting Foster's appointment as a case of feminisation symbolising a 'new direction' for the DUP is, therefore, difficult to validate.

Institutional context
The institutional and political context in which Sinn Féin and the DUP operate may also have facilitated the emergence of female leaders. Both parties organise and compete on multiple levels. As a result, they comprise distinct teams of representatives. The DUP possesses Assembly and Westminster groupings, and for Sinn Féin the key distinction is between a southern operation in Dáil Éireann and a northern operation encompassing Stormont and Westminster. The existence of these different 'nodes' of power (Beckwith, 2015), responsible for leading the party in different contexts, potentially helps explain why gendered leadership change faced little internal resistance or opposition within both parties. As with their shared organisational approach of 'collective' leadership, this dispersal of authority across multiple levels meant that the risks, if any, of selecting a woman leader could be more readily tolerated or offset by other powerful male elites. The DUP provides perhaps the clearest example of this dynamic. As leader of the DUP in Westminster, Nigel Dodds -Foster's main rival for the leadership -commanded a significant power-base. Crucially, his position of influence was not threatened by Foster's elevation to the leadership. Indeed, as would transpire, with the collapse of power-sharing in Northern Ireland in January 2017 and the DUP entering into a confidence-and-supply arrangement with the Conservative Party following the 2017 UK general election, Dodds effectively became the de facto party leader.

Post-conflict opportunities
As well as being organised across multiple levels, there are also important structural factors to consider in relation to the impact of Northern Ireland's post-conflict power-sharing institutions on gender diversity and pathways to political leadership. The creation or reform of political institutions following conflict can provide opportunities to embed gender equality measures within 'new' political structures and wider citizenship rights. In Northern Ireland, the 1998 peace agreement requested that parties 'affirm in particular . . . the right of women to full and equal political participation', while public authorities would, 'promote equality of opportunity in relation to religion and political opinion; gender; race; disability; age; marital status; dependents; and sexual orientation' (Northern Ireland Peace Agreement, 1998). Despite these early promises within a 'new' Northern Ireland, the reality over two decades later reveals the persistence of essentialised gender roles where debate, policy decisions and party competition continue to be framed by ethno-nationalism. This reflects a key criticism of the variety of consociational powersharing established in Northern Ireland; that due to a 'freezing' of communal divisions along ethno-national lines, political success relies on party appeals to ethnic loyalties rather than cross-cutting appeals focused on class, sexuality or gender (Murtagh, 2008).
In contemporary Northern Ireland, this 'freezing' manifests in persistent tensions on how to manage ethno-cultural expressions, such as commemorations and parading, as well as other contentious issues such as same-sex marriage and abortion. 9 With the DUP and Sinn Féin occupying opposite ends of the cultural-moral spectrum (see Garry et al., 2017), debate and policy decisions are tainted by ethno-national divides rather than being discussed on the grounds of women's reproductive rights or those of the LGBTQ+ community (Moon et al., 2019). These debates are described as perpetuating the 'war by other means' and proved particularly contentious during discussions to restore the Assembly after its collapse in 2017 (Hayes and Nagle, 2018: 466). This period of political inertia also saw the expiry of the Gender Equality Framework, the executive's 10-year strategy to address 'the lack of women's representation in political and public life' (Gender Equality Unit, 2010). To date, the framework has not been superseded and demonstrates how gender mainstreaming strategies can get lost in the demands of an ethno-national party system (Northern Ireland Assembly, 2019).
On top of criticisms that power-sharing in Northern Ireland has sidelined or distorted concerns around political gender equality, the institutions themselves also lack mechanisms to ensure greater female representation. In cases of political reform following inter-ethnic conflict, gender quotas often accompany efforts to balance power between communities (see Guariso et al., 2018). Yet, despite recommendations, no such features have been implemented in Northern Ireland's post-conflict political institutions. 10 As a result, parties play a crucial role in facilitating women's political representation and in determining party, and national, leadership. 11 The absence of institutional or system-level measures to facilitate political gender equality reaffirms, once again, the importance of party-level factors to the phenomenon of gendered leadership change in Northern Ireland. More specifically, any systematic analysis and theorising around the gendered paradox of post-conflict power-sharing needs to account for the role played by political partiestheir origins, 'rules', networks and cultures -in shaping gendered patterns of advantage and disadvantage from grassroots to leadership level.

Conclusion
As a rare case of a system which has experienced substantial feminisation of party leadership, Northern Ireland holds some important lessons for those interested in both gendered leadership change and the relationship between political gender equality and intra-party democracy more broadly. Surveying the endogenous and exogenous factors that shape women's political presence reveals some striking similarities between the DUP and Sinn Féin, despite occupying opposite ends of a deeply divided party system. They both emerged in a context of ethnonational conflict, possess histories of masculinised leadership and participate in multiple political systems. More significantly, they were also both engaged in a process of modernisation or 'rebranding', albeit with clearly differing emphases on gender. While Sinn Féin explicitly framed the appointments of McDonald and O'Neill as one facet of a broader equality agenda, the DUP engaged sparingly (and clumsily) with Foster's identity as a woman. Nevertheless, political success -encompassing electoral fortunes, crisis and renewal -appears a significant performance-related incentive for gendered leadership change in the Northern Irish case.
Where the comparison between the DUP and Sinn Féin is perhaps most revealing, however, is the importance of party organisation in the appointment of all three women. Both parties clearly operated in a way that facilitated gendered leadership change. A curious finding from the Northern Ireland case, however, concerns the relationship between formal and informal organisational elements in producing such change. Feminist scholars stress the importance of tracing both the formal and informal 'rules of the game' within political parties, to determine whether and how they affect and structure opportunities for women. A common observation is that formal processes or mechanisms regarded as favourable to women candidates (including those intentionally designed to achieve gender equality) can be undermined by masculinised norms, cultures and informal ways of operating. In the case of leadership selection in Northern Ireland, however, informal norms and cultural practices have, in certain respects, worked to the advantage of women. Both the DUP and Sinn Féin possessed a hypercentralised and top-down organisational culture, which enabled the effective appointment of a winner ahead of a formal selection process. While comparative research on gender and leadership selection suggests that parties institutional rules do not matter in any case, the DUP and Sinn Féin opted not to leave things to chance. Elite control of the process was deemed more favourable to exposing the process to the whims and potential prejudices of a wider selectorate, be that the DUP's elected representatives or Sinn Féin's party activists.
The consequence of the internal organisational culture across both parties, in delivering three women leaders, demonstrates, once more, the complex and imperfect relationship between intra-party democracy and gender (see Childs, 2013). What might strike us, at first glance, as a highly undemocratic scenario, with an omnipotent party centre and little meaningful participation by grassroots members, activists or, indeed, elected representatives, may in fact produce better outcomes from a gender equality standpoint. As with election candidate quotas, if the goal is to promote women to the position of party leader then exclusionary selection practices may need to be tolerated or even encouraged. To quote Sartori's (1965: 24) famous dictum in this respect, 'democracy on a large scale is not the sum of many little democracies'. That said, the Northern Ireland case does invite us to reflect on the permanence of feminised political leadership where change is primarily brought about by party organisational and electoral-strategic factors. As Wauters and Pilet (2015: 89) warn, when it comes to electing women as leaders, 'it is not the rules of the game that matter but rather the openness to women of the wider context . . . [G] ender balance within parties remains first of all a structural problem'. Their focus was exclusively on the formal rules of selection, neglecting informal organisational practices, which we have shown to be significant in the Northern Ireland case. Yet, it is still worth noting the importance of the wider context, particularly how 'open' political parties and society are to women in general. The absence of deeply embedded structural change in Northern Ireland -at both party and system level -means that backsliding on women's political presence in high office remains a very real possibility. As this case study has demonstrated, significant gendered obstacles exist across society in Northern Ireland, as do the persistent conservative political cultures, attitudes and norms that reinforce them. This is most stark at the level of local government; an important launch pad to higher office which remains heavily patriarchal in culture and practice. If the contingencies of political parties' inner structures, cultural dispositions and electoral-political machinations are not reinforced or supplemented by other more robust formal institutional mechanisms aimed at achieving gender equality elsewhere (e.g. candidate quotas), the stark feminisation of party leadership in recent times may prove a one-off episode.
Finally, scholars interested in political gender equality in other post-conflict societies should also find our insights from the Northern Irish case instructive. The consociational institutions established in Northern Ireland share many features with those found in other similar cases, such as Bosnia and Lebanon. While these cases possess their own unique historical and political contexts, political gender inequality, particularly at the level of party or national leadership, is a common and persistent trait. As we note, understanding the gap between the promise of formal frameworks established to resolve a conflict and actual outcomes for women is essential to addressing the gendered paradox of powersharing. An in-depth analysis of one specific case, of the kind we have attempted here, can therefore help others shed light on the shadowy pathways and institutional dynamics that these different cases and contexts have in common.

Notes
1. The phenomenon of gendered leadership change in Northern Ireland extends beyond the two main parties. Women have also been elected to leadership positions in three of the smaller parties: Naomi Long (Alliance), Margaret Ritchie (Social Democratic and Labour Party) and Clare Bailey (Green Party). 2. O'Neill was also elected Sinn Féin Vice-President in January 2018. 3. The Republic of Ireland's record on political gender equality is also a poor one. McDonald is only the third woman to lead a major party. 4. In instances where the party was targeting more than one seat in a constituency, the additional candidate was selected by the central executive as opposed to branch level. 5. It is important to note that while arguments for gender equality fitted more easily within a broader nationalist/republican equality agenda, this did not equate to these women operating on an equal footing to their male counterparts. 6. Any party member can stand for President or Vice-President, so long as they receive endorsement from one internal body and ratification from the party executive. 7. Dedazo, which loosely translates as 'big finger', was the informal process which allowed outgoing Presidents of the Institutional Revolutionary Party in Mexico to hand-pick their successors. The preordained nature of leadership selection in Northern Ireland was noted by many commentators, including O'Hanlon (2018), whom the paper's title is attributed. 8. There is reasonably compelling evidence for this assumption. Sinn Féin was consistently the most 'toxic' option in the Irish party system (Sunday Independent, 2018). One poll on the eve of McDonald's appointment, for example, showed over a fifth of all Irish voters would be more likely to vote for Sinn Féin with her as party leader (Irish Times, 2018). Leadership toxicity was also offered as an explanation for Sinn Féin's breakthrough success in the 2020 Irish general election (see McIntrye, 2020). 9. On 24 July 2019, the UK Parliament passed the Northern Ireland (Executive Formation etc) Act 2019 which, due to the failure to restore the Northern Ireland Executive, allowed the UK government to extend same-sex marriage and lawful access to abortion services to Northern Ireland.

A UN Convention on the Elimination of all Forms of Discrimination Against Women report stated '
There is a pressing need for additional Government actions to strengthen institutional mechanisms for gender equality, including women's equality, in Northern Ireland' (Equality Commission for Northern Ireland, 2018). 11. In Northern Ireland, national leadership is determined at party level. The roles of First Minister and deputy First Minister are a consequence of being elected as representatives of the largest and second-largest parties in the Assembly.