Opening Up the Black Box: Gender and Candidate Selection in a New Era

Abstract

Recent global developments, including the feminization of parliaments and the rise of gender quotas, have transformed the ways in which parties and legislatures operate. This introduction to the special issue ‘Candidate Selection: Parties and Legislatures in a New Era’ puts these recent developments in context, making the case for revisiting the ‘secret garden’ of candidate selection in light of this ‘new era’ in politics. It sets out a critical dialogue between party politics and gender politics scholarship and points to the need for more research on how political parties facilitate or block women’s access to political office. Building on the burgeoning research on gender and political recruitment, it outlines how a gendered and institutional approach allows us to retheorize candidate selection processes and opens up new avenues for empirically examining the pathways prior to election. The article then introduces the papers in this special issue and concludes by evaluating the main implications of gendering analyses of candidate selection and party politics more broadly.

The feminization of politics – that is, the political inclusion of women and women’s policy concerns – has transformed the social and political context within which parties and legislatures operate. Women activists have been key players in debates over political representation and constitutional and institutional design around the world and have mobilized in social movements at the local, national and global level, as well as within formal organizations such as political parties (Krook 2006). While women are still numerically under-represented in politics, the overall trend is upward, and dramatic jumps in women’s political presence have occurred in a diverse range of countries worldwide.

What explains these trends? A key factor has been the adoption of reform measures such as gender quotas, aimed at increasing the selection and election of female candidates. The global spread of gender quotas can be traced back to the Fourth UN World Conference on Women, held in Beijing in 1995. The resulting Platform for Action recommended that governments should ‘take measures to ensure women’s equal access to and full participation in power structures and decision-making’, while also increasing ‘women’s capacity to participate in decision-making and leadership’ (United Nations 1995). This represented a significant discursive turning point in national and international debates over women in politics, legitimizing women’s movement campaigns for gender quotas around the world. Indeed, post-Beijing, quotas have become a global trend – they are currently applied in over 100 countries and are now in their ‘second’ and ‘third’ generation in many cases (see Franceschet et al. 2012; Krook 2009 for comprehensive overviews).

Almost every pluralist democracy uses gender quotas, either in the form of (voluntary) party quotas or statutory quotas introduced by regular legislation, electoral system reforms or constitutional amendments. The inclusion of women in political decision-making is also increasingly seen as a prerequisite for democracy, with gender quotas introduced in post-conflict and transition to democracy contexts, as well as in hybrid regimes (Dahlerup 2006). Yet, while research in this area highlights the significant impact of these measures on levels of women representatives, it also demonstrates that gender balance is far from achieved. Gains remain slow and incremental, and many countries have seen either stagnation or reversals in their numbers of women parliamentarians (Dahlerup and Leyenaar 2013).

The worldwide diffusion of gender quotas establishes a ‘new era’ for political parties and parliaments. In most countries parties are the key gatekeepers to political office, in that they have almost exclusive control over which candidates are recruited and selected (Norris and Lovenduski 1995). Quota reforms, then, challenge the very core of the relationship between voters, parties and representatives (cf. Dahlerup 2006) and force political parties to revise their selection practices in light of gender criteria. Yet, while gender quotas and the broader feminization of party politics are a major global development, most studies of political parties continue to be notably silent on issues of women and gender. The relationship between gender and political parties has not been extensively theorized in mainstream studies of candidate selection specifically, or in party politics scholarship more generally. Meanwhile, though the importance of the candidate selection process is widely recognized in the gender and politics literature, there have been relatively few studies that directly examine the role of political parties in shaping women’s representation.

This article seeks to fill this gap by revisiting the ‘secret garden’ of candidate selection and investigating the central question of how political parties facilitate or block women’s access to political office. We first examine the existing divide between gender politics and party politics scholarship, highlighting the tendency of these two fields to talk past, rather than to, each other. Building on the burgeoning research on gender and political recruitment, we discuss the ways in which a gendered perspective opens up new directions for theorizing the dynamics of candidate selection and recruitment – focusing on how gender ‘makes’ parties, but parties also ‘make’ gender. We move on to consider how scholars might bridge the gap between gender and party politics scholarship and set out the main components of a ‘feminist institutionalist’ approach to studying candidate selection, one that is firmly focused on the gendered and institutional dimensions of the opportunity structures within parties. We conclude by introducing the articles in this special issue, outlining how they deal with various aspects of the relationship between gender, institutions and candidate selection, and highlighting the wider implications of the special issue for research on the critical pathways prior to political office.

GENDER AND PARTY POLITICS SCHOLARS: TALKING PAST EACH OTHER?

Candidate selection is at the core of what political parties stand for and what they do (cf. Ranney 1981: 103) – it ‘influences the balance of power within the party, determines the composition of parliaments, and impacts on the behaviour of legislators’ (Hazan and Rahat 2010: 3). Yet, while the important role of parties in shaping access to political office is widely recognized, there have been surprisingly few systematic studies into the ‘secret garden’ of candidate selection and recruitment. Meanwhile, those classic studies of candidate selection that do exist were largely undertaken before the global surge in gender quotas from the mid-1990s onwards (see, for example, Gallagher and Marsh 1988; Lovenduski and Norris 1993; Norris 1997; Norris and Lovenduski 1995).

Access to political office has long been a core research area within gender politics scholarship, with much of the literature in this area seeking to understand why women are under-represented in political institutions. However, until rather recently, this literature has tended to focus on legislatures rather than on political parties (Childs 2013: 81) and has over-emphasized the importance of political, socioeconomic and cultural variables in explaining cross-national variations in levels of women representatives (see, for example, Inglehart and Norris 2003; McAllister and Studlar 2002; Reynolds 1999; Tripp and Kang 2008). In doing so, these studies have often missed significant within-country variations between political parties, overlooking the fact that individual parties differ in the number of female candidates they nominate and the proportion of women they send to parliament (Caul 1999; Kittilson 2006).

This work is supplemented by a burgeoning body of research on the origins and impact of gender quotas, as the most visible and direct mechanisms used to increase women’s political presence (see, for example, Dahlerup 2006; Franceschet et al. 2012; Krook 2009). Scholars working in this area have offered several explanations for variations in quota adoption and implementation, including, for example, the importance of women’s mobilization inside and outside parties (Kittilson 2006); cultural norms and understandings of equality and representation (Davidson-Schmich 2006); and pressures from international organizations (Krook 2006). Yet, while electoral quotas are intended to counter gender biases in the distribution of political positions, the literature on gender quotas has tended to underplay the importance of internal party dynamics, the arena in which the politics of distribution is ultimately played out (Kenny 2013: 182; see also Verge and de la Fuente 2014). As several scholars have noted, the relationship between political parties and gender quotas has not been extensively theorized, and the intraparty mechanisms that explain how quotas are effectively adopted and implemented in practice remain largely unexplored (but see Davidson-Schmich 2006; Kenny and Verge 2013; Murray 2010; Threlfall 2007).

As posited by Baer (1993: 562), ‘for a party scholar, the concept of political party in women and politics research is missing where one would most expect it – in studies of recruitment and public office holding’. We can similarly argue that, for a gender politics scholar, the concept of gender is also missing where one would most expect it – in mainstream studies of candidate selection and intraparty power struggles. When party politics scholars have looked at the crucial pathways of candidate selection and recruitment, they have generally failed to engage with issues of women and gender. For example, power struggles within the party organization are usually identified with regard to leadership contests and factional disputes (Boucek 2009; Harmel and Tan 2003), or the division of power between the central party and its regional and/or local branches (Hough and Jeffery 2006; Swenden and Maddens 2009). In both cases, the logics of office-seeking versus policy-seeking are applied as motivations for taking over or influencing leadership contests, as well as for preserving or expanding decision-making autonomy in matters of policy, strategy and candidate selection.

In most of this research, those who define the conditions for action and who participate in party factions and coalition-building – the activist, the middle-level cadre, the leader, the aspirant, or the selector – are assumed to be gender-neutral and disembodied actors (cf. Verge 2015). In other words, the question of ‘who is present’ in these deal-making processes is assumed not to matter. Accordingly, party politics scholars have generally failed to consider the extent to which female members are integrated and have influence and power within the party structure, and the resulting implications for wider understandings of party characteristics and party change (cf. Childs 2008, 2013; Young 2000). Meanwhile, those few works that do take women into account tend to fall back on the dichotomous variable of sex, rather than integrate gender more systematically into their analysis (see, for example, Katz and Mair 1995; Webb et al. 2002). In doing so, most party politics scholars have failed to acknowledge the extent to which parties’ ideologies, organizational structures, procedures and practices are ‘saturated with gender’, as well as the ways in which the experiences of actors within political parties also vary according to both sex and gender (Childs 2008: xix; cf. Lovenduski 2005). We expand on this argument in the following section.

HOW GENDER MAKES PARTIES AND HOW PARTIES MAKE GENDER

As Carver (1996) reminds us, gender is not a synonym for women. Rather, gender can be understood as a constitutive element of social relations based upon perceived (socially constructed and culturally variable) differences between women and men, and as a primary way of signifying (and naturalizing) relationships of power and hierarchy (Scott 1986: 1067). From an analytical perspective, gender can therefore be seen as a useful ‘category’ to examine and identify the socially constructed (rather than natural or given) institutional roles, identities and practices conceived of as ‘masculine’ or ‘feminine’ in particular contexts (Beckwith 2005: 131). Yet, gender not only operates at the interpersonal level, but is also a feature of institutions and social structures, and a part of the symbolic realm where gender meanings are constructed (Kenny 2013: 37). To say that an institution is ‘gendered’ then means that constructions of masculinity and femininity are intertwined in the daily culture or ‘logic’ of political institutions, rather than ‘existing out in society or fixed within individuals which they then bring whole to the institution’ (Kenney 1996: 456; see also Kenny 2013). Gender can therefore also be conceived as a ‘process’ through which structures and policies may have a differential impact upon women and men, while also providing different opportunities to female and male actors seeking favourable gendered outcomes (Beckwith 2005: 132).

These wider theoretical insights have been taken up by recent studies on gender and political recruitment, where the focus has shifted from studying ‘women in’ to ‘gender and’ political parties (see, for example, Bjarnegård 2013; Hinojosa 2012; Kenny 2013; Kittilson 2006; Lovenduski 2005; Murray 2010; Verge and de la Fuente 2014). In doing so, many of these scholars have drawn upon the insights of ‘feminist institutionalism’, an emerging variant of new institutional theory which seeks to ‘include women as actors in political processes, to “gender” institutionalism, and to move the research agenda towards questions about the interplay between gender and the operation and effect of political institutions’ (Mackay et al. 2010: 574). Institutions, in this view, are not gender neutral, rather they reflect and reinforce power inequalities (cf. Kenny 2007). This is particularly the case for political parties, which have historically been dominated by men and which are, as a result, characterized by traditional (and often unacknowledged) conceptions of gender relations that generally disadvantage women. Indeed, as Lovenduski (2005: 56) notes: ‘If parliament is the warehouse of traditional masculinity . . . political parties are its major distributors’.

The ‘institutional turn’ in research on gender and candidate selection is rooted in the pioneering scholarship of Pippa Norris and Joni Lovenduski, evidenced in their classic (1995) work Political Recruitment: Gender, Race and Class in the British Parliament. This text represents one of the first (and only) systematic studies into the ‘shadowy pathways’ prior to political office in either the gender politics or party politics fields, and has yet to be replicated on the same scale. In Norris and Lovenduski’s framework, the outcome of particular parties’ selection processes can be understood in terms of the interaction between the supply of candidates wishing to stand for office and the demands of party gatekeepers who select the candidates. In attempting to systematically theorize the political recruitment process, Norris and Lovenduski emphasize that political parties do not operate within a vacuum; rather, supply and demand play out within a wider framework of party recruitment processes, which are shaped and structured by the broader political system (see also Norris 1997). Supply and demand are therefore interactive processes that produce specific gendered outcomes, namely the under-representation of women as aspirants, candidates and eventually elected politicians.

What is the value-added of a feminist institutionalist approach to the study of candidate selection and recruitment? To answer this question, we adapt Vickers’ (2013) theoretical framework of how ‘gender makes institutions’ and ‘institutions make gender’ to political parties. Figure 1 illustrates the key components of the processes through which gender makes parties and parties make gender. We argue that gender makes parties with regard to both the ‘supply’ and ‘demand’ factors underpinning candidate selection processes and broader party organizational dynamics. On the supply side, due to wider systemic factors such as the public–private divide, the sexual segregation of the workforce and patterns of gender socialization, we might expect aspiring women candidates to have less time, money, ambition and confidence than their male counterparts to run as candidates (Norris and Lovenduski 1995). For example, a notable recent study finds that women in American politics – despite sharing similar qualifications and experiences to their male counterparts – are much less likely than men to consider running for office or to put themselves forward as candidates (Lawless and Fox 2010).

Figure 1 How Gender Makes Parties and Parties Make Gender

Ingrained social gender ideologies also shape the direct or imputed prejudices of party selectors, leading to social bias in parliaments, as demand-side explanations suggest (Norris and Lovenduski 1995: 106–8). In other words, parties institutionalize ideas about politics that have gendered implications (Lovenduski 2005). Some studies find that party elites typically list stereotypically masculine characteristics when asked to describe a ‘good leader’ or a ‘good candidate’ (see, for example, Tremblay and Pelletier 2001). Others demonstrate that male party elites tend to recruit fewer women than men because of an out-group effect, basing their candidate evaluations on stereotypes about women as a whole, and as a result, generally perceiving female candidates to be less competent than their male counterparts. In contrast, male candidates are seen as in-group members and are assumed to be politically competent (Niven 1998). Women’s exclusion from the distribution of rewards (political positions) therefore often builds on psychological considerations deeply rooted in society about gender roles that make men see other men as more likeable and reliable peers when recruiting candidates for public office (Bjarnegård 2013). In a similar vein, the late-hour meetings typically entrenched in party culture clearly reflect male schedules or wifely support, which makes individuals with fewer family responsibilities – due to the current sexual division of labour, those would typically be men – more likely to join and to be active participants of political parties (Franceschet 2005; Verge and de la Fuente 2014).

Equally important, we argue that parties make gender. This is not to say that political parties only represent male interests and always operate to oppress women, or that all men are in dominant positions in party politics while all women are in subordinate ones (cf. Kenny 2013). Rather, the historical exclusion of women from political parties, institutions and public life has ‘permitted a set of male-centred institutional practices to evolve without comment or protest’ (Lovenduski 2005: 27) – practices which have palpable effects. One of the most fundamental (and often invisible) by-products of men’s historical domination of political parties (including party membership, party leadership and candidate selection committees) has been their ability to set the institutional ‘rules of the game’ (Lovenduski 2005: 27). This, in turn, produces gendered outcomes, shaping the opportunities for both men and women to be selected and elected and to advance through the party ranks. While these gendered barriers are, on the one hand, a reflection of wider disadvantages that women face in society at large, they are compounded by party cultures that have ‘institutionalized codes of behaviour that make discrimination against women both possible and acceptable’ (Lovenduski 2005: 81).

In the early stages of the candidate selection process, for example, parties have significant influence over who decides to run for office – often informally encouraging particular candidates to run, or more indirectly, sending signals about what kind of candidates would ‘fit in’ with the party (Cheng and Tavits 2011: 467; see also Niven 1998). The evidence suggests not only that male party leaders prefer, but also that they actively support and promote the nomination of male candidates, with women much less likely to be approached to run for office than their male counterparts (Carroll and Sanbonmatsu 2013; Dittmar 2015; Lawless and Fox 2010). Many of the qualities sought by candidate selectors also indirectly favour men over women, with formal and informal selection criteria on the demand side – such as party service, resources, experience and leadership traits – shaping the supply of candidates along gendered lines (Kenny 2013; Verge 2015). Once women decide to stand for office, there is ample evidence that they face both direct and indirect discrimination from party selectors in the candidate selection process – ranging from gendered assumptions regarding women’s traditional roles to explicit sexual harassment (see, for example, Lovenduski 2005). Finally, once women are selected, political parties are much more likely to run them as ‘sacrificial lambs’ in unwinnable seats or positions on party lists (Luhiste 2015; Murray 2010). This, in turn, may reinforce the notion that women are not ‘suitable’ for political office, while further discouraging prospective female candidates from entering the political arena (Ryan et al. 2010).

One possible way to surmount these barriers, as advanced by some scholars, is to increase the number of women in party leadership positions (see, for example, Cheng and Tavits 2011). Yet this is no guarantee of positive change; there are many cases where the numerical feminization of party executive bodies has not led to the feminization of party practices. Indeed, several studies point to how power in male-dominated political spaces may shift over time from formal to informal institutions, or to different institutional arenas, in order to counteract women’s increased access and presence in formal decision-making sites (Bjarnegård and Kenny 2015; Hawkesworth 2005; Kenny 2013). Power inequalities in regard to access to resources and recognition often continue to be maintained through more informal ‘rules-in-use’ – including gendered norms or expectations of a ‘good’ leader, gendered rituals in party meetings that keep men conversationally dominant, and the subjection of women party officials to enhanced levels of scrutiny (Franceschet and Piscopo 2014; Puwar 2004; Verge and de la Fuente 2014).

Political parties must then be regarded as institutional spaces where gender is both produced and reproduced through a myriad of practices and norms – many of them informal – and that may also shift over time – for example, when gender quotas are adopted (Kenny 2013; Lovenduski 2005; Verge and de la Fuente 2014). Simultaneously, the conceptualization of gender not only as a category but also as a process (Beckwith 2005) allows us to pay attention as well to the ways in which gender-biased institutionalized frameworks can be challenged. Indeed, as the introduction of quotas shows, gender can be a source of party change and, as such, is likely to be resisted by organizations such as political parties that have historically proven to be largely conservative when it comes to change (Panebianco 1988) and rather good at protecting their (male-dominated) cultures and procedures (Lovenduski 2005).

BRIDGING GENDER POLITICS AND PARTY POLITICS RESEARCH

Looking at party politics through a gendered lens inevitably changes the way we think about and analyse candidate selection. It fundamentally tells us that party politics is a site of power relations. The central questions of candidate selection and party politics scholarship – Who can be selected? Who selects the candidates? Who should select the candidates? Who participates? – cannot be fully addressed without reference to gender. Answering these questions requires an evaluation not only of women’s positional power within party structures, but also an evaluation of their power relative to that of their male peers (Childs 2013: 93). For example, individuals’ interests and strategic calculations in selection processes cannot be generalized since they derive from their social, economic and political position, including the ways in which their private lives operate. The amount of resources at their disposal is also likely to shape their capacity for and types of collective mobilization (see Lovenduski 1998).

Acknowledging the impact of unequal gender relations, in turn, will allow party politics scholars to provide a more complete and accurate account of internal reform processes and intraparty power struggles. An example of this can be found in current debates over internal party democracy, the increasing trend among parties in Western democracies aimed at democratizing and decentralizing internal structures and candidate selection processes (see, for example, Cross and Katz 2013; Hazan and Rahat 2010; Rahat and Hazan 2001). We would argue that any analysis of these trends is partial at best if attention is not paid to the discrimination faced by some groups with regard to their access and opportunities to participate effectively within the party organization. Indeed, as Sarah Childs (2013: 93) asks, ‘can a political party be judged internally democratic’ if women continue to be under-represented at all levels of the party?

More specifically, a gendered lens raises deeper and more critical questions about not only what constitutes internal party democracy, but also whether it has gendered effects (Childs 2013). Young and Cross (2002), for example, chart the rise of the ‘plebiscitary party’, in which parties have responded to a decline in conventional political participation by reforming their internal practices to include more direct involvement in decision-making (see also Seyd 1999). Yet this shift from group to individual representation has a gendered impact – party commitments to increasing the numbers of women and other marginalized groups in politics rely on ‘the acceptance of sex as a basis of group identity’ (Childs 2013: 90; see also Krook and O’Brien 2010). Indeed, Young and Cross (2002: 689) find that women, ethnic minorities and young party members are less supportive of an undifferentiated conception of membership that denies the salience of group differences.

Similarly, trends to democratize and decentralize candidate selection procedures – including the introduction of broad-based membership ballots and primaries – have gendered consequences. However, those party politics scholars who have looked at the differential impact of primaries on women and men have contented themselves with highlighting that this selection method usually produces fewer women candidates and with providing relatively superficial explanations such as the underlying coordination problems of inclusive selectorates (Hazan and Rahat 2010) and the conservative attitudes of party members towards women candidates (Rahat 2007). This focus on sex rather than on gender fails to notice the gendered processes embedded in primaries: this selection method tends to benefit incumbents, who are mostly men; inclusive selectorates (citizens or party members) typically vote for the most popular candidates, whose popularity cannot but reflect the vertical segregation within parties and institutions, as well as asymmetrical access to party patronage networks (Franceschet and Piscopo 2014; Verge and de la Fuente 2014). More crucially, it also misses, as previously highlighted, the fact that women are less likely to be encouraged to stand by party selectors (Carroll and Sanbonmatsu 2013; Lawless and Fox 2010).

Adopting a gendered lens also allows us to evaluate critically the ideals and goals of intraparty democracy. Gender quotas, for example, are in many ways the antithesis of internal party democracy – in that they are usually highly centralized ‘top-down’ party initiatives that run counter to traditional conceptions of ‘democratic’ candidate selection. Yet quotas are intended to counter gender biases in the distribution of political positions – indeed, as Childs (2013: 94) notes, preventing selectorate discrimination is their raison d’être. As already highlighted, the existence of discrimination against women candidates is well established, and even the adoption of formal rules such as gender quotas does not necessarily overcome all institutionalized forms of male bias, as informal party practices may undermine formal rule change (Bjarnegård and Kenny 2015; Kenny 2013). A gendered lens, then, raises the question as to whether there must necessarily be limits on intraparty democracy (as it is traditionally understood in the literature) in order to ensure the wider ‘good’ of ‘system-level democracy’ (Childs 2013: 99).

We can see an example of these tensions in the multitude of ways in which parties often (formally) comply with the letter of gender quota laws while (informally) violating their spirit. In cases of quotas that do not have placement mandates, parties typically place women in hopeless seats or list positions, keeping the lion’s share of safe seats and winnable positions for male incumbents or male newcomers, which erodes the transformative potential of gender quotas (Murray 2010; Ryan et al. 2010). When quotas do have placement mandates, parties often fulfil the minimum requirements of the rules in terms of women’s representation, but go no further than this (Verge 2013). Political parties have also developed expertise in exploiting the legal loopholes of electoral and quota rules. Mexico’s ‘first generation’ quota law, for example, exempted parties from applying gender quotas in direct votes or primary elections. Parties made extensive use of this ‘escape clause’, which resulted in strongly masculinized candidacies but also meant that they were able to avoid scrutiny as to how open these primary elections really were (Baldez 2007; see also Johnson 2016 and Piscopo 2016, both in this issue). In some cases, parties have gone so far as to directly break quota laws – for example, by putting men candidates as alternates of women candidates and have the latter resign once elected but before taking office, as has happened in Mexico (Hinojosa 2012; see also Piscopo 2016, in this issue), or by entering male candidate names under female names, as in the case of Bolivia (Albaine 2009).

More research is needed, then, to open up the ‘black box’ of candidate selection and to identify and empirically investigate the ways in which parties facilitate or block women’s access to political office. In one way or another, all of the contributors in this special issue seek to fill this gap, bridging the intersection between gender politics and party politics scholarship. Starting from the question ‘what does a gendered perspective tell us about party politics?’, the special issue points to several new directions for theorizing and empirically examining the dynamics of candidate selection. On the one hand, given that selection processes do not take place in a vacuum, the party organization as an institutional setting that shapes gendered patterns of candidate selection must be more thoroughly examined. On the other hand, empirical analyses must take into account informal rules and their interaction with the formal rules of the selection process and how these enable or constrain men’s and women’s political participation, an issue of paramount importance after the global expansion of gender quotas.

Adopting a broadly ‘feminist institutionalist’ approach – focusing on the gendered and institutional dimensions of the opportunity structures within parties – the ‘secret garden’ of candidate selection is surveyed in a comparative perspective, either across parties or across countries, covering a broad range of empirical case studies, including Western Europe (where most work on recruitment has traditionally focused), but also Latin America, Asia and Africa. Specifically, contributors examine how gender shapes the structures, practices and rules of political recruitment (both formal and informal), including its intersection with other axes of power such as race/ethnicity where applicable. In doing so, they not only shed light into the gendered politics of advantage and disadvantage within parties, but also provide insights into how these power hierarchies and structures might eventually be dismantled.

Elin Bjarnegård and Meryl Kenny’s (2016, in this issue) agenda-setting piece sets out the theoretical and methodological challenges of researching candidate selection in comparative perspective and explores these dynamics by revisiting original in-depth research in two very different contexts: Thailand and Scotland. Subsequent contributions provide in-depth empirical analyses of the pathways prior to political office and highlight a number of strategies that researchers might use to surmount the comparative challenges identified by Bjarnegård and Kenny – for example, comparing multiple parties within a country over time; situating the findings of individual cases within existing research and regional patterns to pull out similarities and differences; or carrying out structured small-n cross-country comparisons. Niki Johnson’s (2016, in this issue) study of candidate selection in Uruguay highlights the ways in which male power monopolies are discursively constructed and then practically implemented and sustained over time in different political parties in spite of gender quota reforms. Meanwhile, Tània Verge and Ana Espírito-Santo’s (2016, in this issue) study of Spain and Portugal unpacks the interactions between legislative and party quotas and identifies the practices and norms that hinder effective compliance with these measures. Francesca Refsum Jensenius’ (2016, in this issue) contribution tackles the issue of intersectionality in candidate selection and exposes the gendered political opportunity structures that make reserved seats in India more accessible to women candidates than regular seats. Elin Bjarnegård and Pår Zetterberg (2016, in this issue) look at the role of party gatekeepers in hybrid regimes, showing how the uneven playing field in which political parties compete affects the implementation of reserved seats for women in Tanzania.

While all of the contributions highlight the active and ongoing ways in which gender is reinscribed in political parties, they also point to possibilities for these institutions to be ‘re-gendered’ in a positive direction. In this vein, Jennifer M. Piscopo’s (2016, in this issue) analysis of gender and candidate selection in Mexico demonstrates how women’s agency through informal means such as cross-party networks in conjunction with state regulators has produced significant advances in gender quota implementation. Finally, the last of the seven original articles, by Joni Lovenduski (2016, in this issue), offers a more reflective and personal look at the development of the field of gender and political recruitment, focusing particularly on the theoretical and methodological underpinnings of the dominant framework in the field, the supply and demand model. Lovenduski reminds us that as we look forward to new and exciting directions for work on gender and political recruitment, we should also remember to look backwards, building on earlier feminist scholarship in the field, while also pushing and developing this work further in order to advance comparative research and improve data collection.

CONCLUSIONS

This article has identified the ‘missing link’ between party politics and gender politics scholarship, arguing that they must speak to, rather than past, each other. Bridging the intersection between both strands of literature allows scholars to reflect on how power is played out, to expose its reproduction, regeneration and maintenance and to unpick how, why and with what effects the institutions and processes surrounding political recruitment and candidate selection are gendered (as well as their relationship with other axes of inequality). A gendered lens is therefore crucial if we want to fully open up the ‘black box’ of candidate selection – it provides a more comprehensive account of the practices, goals and outcomes of the candidate selection process; the dynamics of internal reform processes and intraparty power struggles; and the general and gendered mechanisms of institutional continuity and change. As Lovenduski (2016) reminds us in the concluding article of this special issue, ‘it is not enough for us to know that parties may discriminate against women; we need also to know via what mechanisms and configurations the discrimination operates and what is at stake’. The contributions to this special issue take up this challenge, pointing to new directions for theorizing the relationship between gender and political parties, and generating a range of new data and insights into the ‘secret garden’ of candidate selection around the world.