Interactions between Party and Legislative Quotas: Candidate Selection and Quota Compliance in Portugal and Spain

Abstract

About two decades after the introduction of party quotas, in the mid-2000s both Portugal and Spain enacted legislative gender quotas. The simultaneous implementation of party and legislative quotas raises questions about the potential interactions between two types of candidate quotas sharing the same goal – granting gender equality in political representation. Following a feminist institutionalist approach, this article aims at disentangling under what circumstances compliance with legislative quotas is greater. By looking at the different party institutional contexts in which candidates are selected, a double comparative framework is set. Firstly, we examine within country how legislative quotas affect political parties with dissimilar strategies to pursue equal gender representation. Secondly, we analyse across countries how they impact on political parties with differently institutionalized voluntary quotas. The article shows that legislative quotas are nested in political parties’ candidate selection process and that existing gendered practices and norms limit the effective compliance with such measures.

Party strategies are crucial for the nomination of more women candidates and eventually the election of higher numbers of women representatives (Lovenduski and Norris 1993). An increasing number of political parties have assumed gender quotas in the last few decades, be they ‘strong’ – codified in party rules – or ‘soft’ – non-written targets or recommendations. If effectively implemented, these quotas may allow parties to set a lead in women’s representation within their polity (Kittilson 2006). More recently, legal quotas have been introduced worldwide, especially in Latin America and post-conflict contexts. In Europe, however, their spread has been more modest (eight countries), and party quotas are the dominant strategy (Krook et al. 2009).

The simultaneous implementation of party and legislative quotas in some countries raises questions about the interactions between two types of candidate quotas sharing the same goal (Kunovich and Paxton 2005: 534) – that is, granting gender equality in political representation – which this article seeks to unpack. Specifically, we aim to explore under what circumstances compliance with legislative quotas is higher. The ideal comparative research design to undertake such a study is one in which the two kinds of candidate quotas coexist and where the main socioeconomic and cultural factors affecting gender equality as well as the most relevant systemic factors shaping candidate selection are rather similar. These requisites are met in the comparative analysis of Portugal and Spain developed in this article.

In these south European countries party quotas had already operated for almost two decades when legislative quotas were introduced in the mid-2000s. As regards socioeconomic and cultural factors, at the onset of their transitions from authoritarian rule, these third-wave parliamentary democracies shared a strong Catholic tradition, heavy inegalitarian socialization and a familialistic approach to welfare, which still persists (González et al. 2000). As to political factors, the combination of a proportional representation electoral system with a small to medium average district size has produced moderate multiparty competition, with one of the lowest effective numbers of parliamentary parties in Western Europe (Armingeon et al. 2014). Closed party lists have also allowed parties to control candidate selection tightly in a centralized manner (Hopkin 2009).

In Portugal and Spain the introduction of party quotas was led by social democratic parties while liberal and right-wing parties have been vocal opponents of positive action in political representation. Intriguingly for our comparative purposes, compliance with party quotas by social democratic parties presents significant cross-national differences. This allows us to set a double comparative framework to examine within country how legislative quotas affect parties with dissimilar strategies to equal gender representation and across countries how they impact on parties with differently institutionalized voluntary quotas. In shifting the traditional focus of the literature on quotas from adoption to implementation, our most similar systems comparative design shows that legislative quotas are nested in parties’ candidate selection process and that existing gendered practices and norms limit effective compliance with such measures.

The remainder of the article is structured as follows. After elaborating our theoretical expectations of the interactions between party and legislative quotas, we introduce the cases under examination and present the data and methods used in the empirical analysis. The next two sections examine candidate selection processes in different types of party institutional settings – party quotas versus no party quotas and highly institutionalized versus poorly institutionalized party quotas. The article thus focuses on the last stages of the ‘ladder of recruitment’ (see Lovenduski 2016, in this issue), that is, the demand for certain types of candidates and the outcome of elections. Lastly, we discuss the main findings and suggest further avenues for comparative research.

INTERACTIONS BETWEEN PARTY QUOTAS AND LEGISLATIVE QUOTAS

Compliance with legislative quotas might well vary across the diverse party institutional settings in which candidate selection processes are embedded. Indeed, party norms and practices largely account for distinct within-country and cross-country outcomes in quota reform processes (Verge 2013: 441). The implementation of legislative quotas is nested in party organizations that already used or did not use party quotas. Simultaneously, party quotas differ in their level of institutionalization. Below we draw our main theoretical expectations for each type of party institutional setting.

Comparing Parties with Dissimilar Strategies on Equal Gender Representation

As highlighted by feminist institutionalist research, the parallel existence or even contradiction between ‘rules-in-form’ and ‘rules-in-use’ (Leach and Lowndes 2007: 186) and between ‘old’ and ‘new’ institutions (Kenny 2013) determine the outcome of quota reforms. The frequent mismatch between the proportion mandated by legislative quotas and gender outcomes indicates that their effectiveness largely depends on the willingness of party actors (Dahlerup and Freidenvall 2005). Consequently, gender quota reforms have been characterized as ‘layering’ processes in which some elements are renegotiated while others persist (Waylen 2014: 219). In fact, institutional reform efforts must contend with party actors often ‘remembering the old’ and ‘forgetting the new’ (Mackay 2014). Gendered institutional settings such as political parties (Lovenduski 2005) provide a fertile ground for examining how legislative quotas are affected by the gendered politics of distribution (Bjarnegård and Kenny 2016, in this issue) and for assessing the extent to which this measure can subvert male dominance (Verge and de la Fuente 2014).

Although the effectiveness of quotas depends on their design (Schwindt-Bayer 2011), including placement mandates and sanctions for non-compliance, political parties frequently use several gendered practices (Dahlerup and Freidenvall 2005; Murray 2010). Institutional innovation might be ‘actively resisted or passively neglected’ since the ‘old’ is not automatically displaced (Mackay 2014: 550). As argued by Mateo Díaz and Millns (2004: 301), legislative quotas ‘have little chance of producing effective outcomes where they operate within a climate of gender biased political structures and do not adjust to national electoral systems’. Whereas under plurality or majority systems candidates’ success depends on candidates fighting for safe seats, under proportional representation systems the rank order of candidates is crucial to getting elected (Norris 2004).

Quota reforms are combined with previous institutions producing good fit or conflict and tension (Krook 2009: 52–3). Where party quotas exist and ‘are followed’ (Matland and Montgomery 2003: 32), legislative quotas might be ‘complementary’ in that equal gender representation already operates as ‘party law’. Furthermore, legislative quotas might stimulate the introduction of further party measures and/or incentivize parties to surpass the legally imposed proportions (Meier 2004). Since in post-legislative quota settings latecomers on women’s representation might catch up with early adapters, parties already using positive action might seek to outperform their previous records in order to continue championing gender equality.

Conversely, in the absence of party quotas, legislative quotas need to be ‘accommodated’ to a party institutional setting in which this representational norm has not yet been assumed. This is particularly important since legislative quotas interact with parties’ goals, norms and strategies (cf. Siavelis and Morgenstern 2008: 28) and coexist with other formal or informal party rules on group representation – such as incumbency, interest groups or party factions – which might yield conflict (Reiser 2014: 56). When legislative quotas do not match existing party practices and norms, they may produce marginal effects on the number of elected women (Verge 2013: 449). In this vein, our first theoretical expectation posits that parties using voluntary quotas will show a higher compliance with legislative quotas. Thus, differences between parties with dissimilar commitments to equal gender representation will not necessarily vanish once quota laws are passed.

Comparing Parties with Voluntary Quotas but Different Levels of Institutionalization

When a behaviour pattern such as the consideration of equal gender representation criteria in list-building is taken for granted it becomes a routine-based performance enacted without further action needed. Otherwise an explicit reminder of the party rule and mobilization to reproduce it are required (Inhetveen 1999: 405). Highly institutionalized party quotas are routinely accepted and applied, whereas poorly institutionalized ones are regularly challenged by resistant party actors and heavily debated within parties (Cheng and Tavits 2011; Reiser 2014). Therefore, our second theoretical expectation establishes that the higher the institutionalization of party quotas the more compliant parties are likely to be with legislative quotas.

The institutionalization of party quotas is also likely to shape reactions to rule-breaking. Highly institutionalized party quotas may not only entail formal sanctions when party selectors do not act in accordance with the quota stipulations – despite them being a rare measure – but also bring about informal consequences such as ‘criticism, a loss of reputation of the responsible persons within the party, or media reports on inner-party conflicts about the quota’ (Inhetveen 1999: 406). This explains why highly institutionalized quotas are barely affected by parties’ electoral misfortune (Verge 2012: 406), while poorly institutionalized quotas are vulnerable to regress under unfavourable contexts, such as electoral swings, when the number of seats expected to win by the party shrinks and internal competition for safe seats or winning positions gets tougher (Murray 2010: 92).

The degree of institutionalization can thus be operationalized into three indicators: acceptance, adherence and sanctions for non-compliance. It should be noted that the institutionalization of party quotas is shaped by some traits of the party organization. Firstly, the existence of detailed and standardized written rules on candidate selection strengthen the effective implementation of party quotas. On the contrary, selection rules that are unwritten (or written but rarely implemented) or not explicit enough so that their interpretation might change from one selection to another give party selectors more room to distribute personal patronage (Norris 1993: 322). Similarly, party quotas with specified rank orders reduce selectors’ capacity to rely on informal understandings of list-building that may mainly benefit particular candidates or members of their inner circles (Bjarnegård and Zetterberg 2011: 190–3).

Secondly, provided that party leaders are sympathetic to increasing women’s representation, centralized candidate selection processes – wherein candidate lists are approved and/or ratified by central party bodies – allow quotas to be enacted in a top-down fashion, giving central-level elites the ability to coordinate district-level nominations as if a national list were being drafted (Kenny and Verge 2013; Murray 2010). Thirdly, party women’s mobilization, predominantly through party women’s sections (Lovenduski and Norris 1993), plays a crucial role in the supervision of quota implementation. However, their influence depends both on the position they hold in the party hierarchy and on the strategies they are able to develop (Kittilson 2006).

OVERVIEW OF CASES, DATA AND METHODS

In order to examine the interaction between party and legislative quotas, we adopt a most similar systems design focusing on Spain and Portugal. Significant increases in the number of women representatives only occur when the largest parties ‘buy into’ gender balance. The social democratic parties, namely the Spanish Socialist Workers’ Party (Partido Socialista Obrero Español – PSOE) and the Portuguese Socialist Party (Partido Socialista – PS), adopted party quotas in the late 1980s. On the other hand, their main competitors, the right-wing Popular Party (Partido Popular – PP) in Spain and the centre-right Social Democratic Party (Partido Social Democrata – PSD) in Portugal,¹ have invariably rejected positive action. These parties have traditionally held about 80 per cent of seats altogether in their respective national parliament.² Since the gender outcomes produced by the largest parties are crucial for the representation of women, our empirical analysis will focus exclusively on them.³

In both countries increases in women’s representation are chiefly explained by party quotas. In 1988, the Spanish Socialist Workers’ Party adopted a 25 per cent quota for women and the Portuguese Socialist Party assumed a 25 per cent quota for either sex. Yet, as illustrated in Figure 1, women’s representation experienced sharper increases in Spain than in Portugal since party quotas were immediately adhered to and more rapidly expanded. In 1994, the Spanish Socialist Workers’ Party raised the quota to 30 per cent for women candidates and enforced it to winnable positions of lists. In 1997, a gender-neutral quota entitled either sex to at least 40 per cent of positions (Verge 2012: 398) and in 2013 ‘zipping’ was adopted – 50 per cent for either sex, with strict alternation of women and men throughout the lists. In Portugal, the Socialist Party only started complying with its own quota a decade after its adoption (Meirinho Martins and Pequito Teixeira 2005: 150), and it was not until 2003 that the quota was enlarged to a third of positions for either sex (Espírito-Santo 2006: 72).

Figure 1 Women Representatives in Portugal and Spain, 1976–2015 (%) Note: The data reported here as well as in subsequent tables captures the percentage of women representatives right after the respective national elections. Source: Own elaboration.

In both Portugal and Spain several quota bills were submitted from the late 1990s onwards but the conservative majorities in the respective national parliaments rendered them unsuccessful. In the mid-2000s, when the Spanish Socialist Workers’ Party and the Portuguese Socialist Party came back to power, quota laws informed by the respective party quotas were eventually passed. The Spanish Law for the Effective Equality of Women and Men (2007), which incorporates gender balance in political representation, and the Portuguese Parity Law (2006) present remarkable differences (Baum and Espírito-Santo 2012; Verge 2012). In Spain, party lists must include a minimum of 40 per cent of candidates of either sex – a proportion to be met too in each stretch of five candidates. In Portugal, party lists must include at least a third of candidates of either sex and neither sex shall occupy more than two consecutive positions. In the Spanish case non-compliant party lists are withdrawn, whereas in Portugal financial penalties are applied. Parties’ public subsidy for the electoral campaign is reduced by 25 per cent when the proportion of either sex falls between 20 per cent and 33 per cent and halved when lists contain less than 20 per cent candidates of either sex or the placement mandate is not followed. Legislative quotas did not produce an overnight change. In Portugal, women’s representation was boosted from 21.3 per cent in 2005 to 27.8 per cent in 2009, decreased to 26.5 per cent in 2011 and increased again in 2015 to 33 per cent. In Spain, it stagnated at pre-legislative quota levels around 36 per cent in the 2008 and 2011 elections and reached 39.4 per cent in 2015 (see Figure 1).

Our empirical analysis proceeds in two steps. Firstly, we look at the interplay between formal and informal institutions in the candidate selection process by examining the list-building carried out before and after the introduction of legislative quotas by the two main parties of each country, which present dissimilar commitments to equal gender representation. Legislative and party quotas are considered the main ‘rules-in-form’. To identify the ‘rules-in-use’ operating at the last stages of the ‘ladder of recruitment’ we have scrutinized the lists presented by the four parties since the first post-transition elections up to the most recent elections. By comparing proportions of women and men candidates and by assessing the effect of rank order on candidates’ chances of being elected, we seek to unfold how the gendered politics of distribution plays out at the last stages of candidate selection.

Secondly, to delve further into the gendered dynamics that may limit the effective compliance with quotas, we examine social democratic parties in greater detail. Their formal architecture is captured by written party rules on candidate selection. Semi-structured interviews with key party informants (see Appendix below) have provided us with valuable insights to identify how the selection process works in practice and to gauge the degree of institutionalization of party quotas, which have been operationalized, as already mentioned, into three indicators − acceptance, adherence and sanctions for rule-breaking. The interviews have also helped us to assess the level of formalization and centralization of candidate selection processes and to identify the intervening actors, paying particular attention to the influence of party women’s sections.⁴

DISSIMILAR COMMITMENT, DIFFERENT RESULTS

In Spain, the introduction of the party quota in 1988 by the Spanish Socialist Workers’ Party set it well apart from the main electoral competitors with regard to the feminization of party lists. From the 1989 elections, the social democrats nominated a much larger proportion of female candidates than the Popular Party, as shown in Table 1. In the 2004 elections women accounted for 44.1 per cent of Spanish Socialist Workers’ Party candidates whereas the Popular Party only reached 35.8 per cent. In 2008, the first post-legislative quota elections, female candidates experienced a 30 per cent growth rate in the Popular Party compared with 9 per cent in the Spanish Socialist Workers’ Party, but the latter’s advantage was still kept. Further increases have been modest in subsequent elections (2011 and 2015), with percentages already close to 50 per cent of either sex. The threat to have lists withdrawn has made non-compliance extremely costly for parties, so lists have always met the required proportions.

Table 1 The Gendered Politics of Distribution in List-building, Spain (% women)

Note: To calculate the winnable positions in election t we have used the seats obtained by parties in each district in t−1 as the benchmark.

Source: Own elaboration.

In Portugal, the socialists have also traditionally nominated more women candidates than their main rival, but differences were rather small until 1999, when the Socialist Party first complied with its own party quota, as shown in Table 2. While in 1995 the Socialist Party nominated 14.3 per cent women candidates and the Social Democratic Party 11.7 per cent, in 1999 the former increased this proportion to 24.8 per cent and the latter to 16.1 per cent. The difference between the two parties widened in 2005 following the expansion of the Socialist Party quota to one third for either sex in 2003. As in Spain, the legislative quota brought about more change in parties with no previous use of quotas. While the proportion of female candidates between 2005 and 2009 barely changed in the Socialist Party (around 34 per cent), the Social Democratic Party jumped from 17.8 per cent to 31.7 per cent. Thereafter, slight increases were observed in the Socialist Party (37.8 per cent of women candidates in 2011 and 2015), while the Social Democratic Party followed an upward trend, fielding 33.4 per cent of female candidates in 2011 and 39.6 per cent in 2015 (in coalition with the Democratic Social Centre (Centro Democrático Social – CDS)). Although the Portuguese Parity Law establishes a softer sanction for non-compliance – a financial penalty – all political parties represented in parliament have complied with the quota dispositions.⁵

Table 2 The Gendered Politics of Distribution in List-building, Portugal (% women)

Note: *In 1979, 1980 and 2015 the Social Democratic Party coalesced with other parties, so data on winnable positions and number one positions cannot be calculated. For these elections, we report the share of female candidates of the coalition’s lists and the proportion of women representatives it elected.

Source: Own elaboration.

While candidate gender outcomes show the extent to which parties comply with the letter of the law (required minimum proportions), to grasp whether the spirit of the law (producing parity representation) is also followed we must look at levels of elected women. In Spain, the Spanish Socialist Workers’ Party clearly outperformed the Popular Party before the enactment of the quota law. For example, in 1996 the Spanish Socialist Workers’ Party nearly doubled the proportion of women compared with the Popular Party (27.7 per cent vs. 14.3 per cent). In 2004, both parties significantly increased women’s presence in their parliamentary groups. The Spanish Socialist Workers’ Party elected 46.3 per cent women and the Popular Party levels rose to 28.4 per cent. This comparative advantage was maintained once the legislative quota came into force in 2008 (42.3 per cent in the Spanish Socialist Workers’ Party vs. 30.5 per cent in the Popular Party), though it was almost lost in 2011 (38.2 per cent vs. 36.6 per cent) due to the severe electoral loss suffered by the Spanish Socialist Workers’ Party. In 2015 the difference widened again, with the Spanish Socialist Workers’ Party electing 45.6 per cent women and the Popular Party stagnating at 35.8 per cent.

In Portugal, the Socialist Party has consistently elected more women than the Social Democratic Party since the 1990s, especially between 1999 and 2005 as a result of adherence to party quota rules. However, after the implementation of the legislative quota in 2009, cross-party differences were notably reduced, with the Socialist Party electing 29.1 per cent women representatives and the Social Democratic Party 26.9 per cent. In 2011, the Socialist Party (24.3 per cent) was even surpassed by the Social Democratic Party (28.7 per cent), partially due to the latter’s better electoral performance. Likewise, in 2015 the Social Democratic Party (together with the Democratic Social Centre) elected more women than the Socialist Party (33.6 per cent vs. 31.4 per cent).

As already mentioned, in proportional representation systems rank orders in lists crucially shape candidates’ chances of getting elected. The gendered allocation of winnable positions is a pervasive informal practice entrenched in list-building that hinders the success of quota reforms – that is, it is possible to respect the letter of the law while unfolding strategic discrimination against women candidates. This practice interlocks with electoral and party system dynamics and can be very harmful for women’s representation under small district magnitudes – half the Spanish (27 out of 52) and Portuguese (10 out of 22) districts elect five deputies or fewer. Were parties to field only men in the top positions of their lists, women candidates would find it hard to obtain representation in small districts.

In the Spanish case, as shown in Table 1, the Spanish Socialist Workers’ Party has clearly outperformed the Popular Party since 1989. Women candidates have occupied over 40 per cent of winnable positions in Spanish Socialist Workers’ Party lists since 2004 and reached 49 per cent in 2015, compared with 31.7 per cent in the Popular Party in 2004 and 37 per cent in 2015. Hence, party differences have not vanished after the introduction of the quota law. In Portugal, as can be seen in Table 2, the proportion of winnable positions allocated to women candidates has been higher in the Socialist Party than in the Social Democratic Party since 1987, with remarkable differences found in 2005: 27.1 per cent vs. 10.5 per cent. However, once the legislative quota was introduced, differences were eroded. In fact, both parties tend to place women every third position on lists, a practice that constitutes a clear barrier to their election, especially in small districts.⁶ Whereas the Social Democratic Party raised the proportion of women in winnable positions to 25.3 per cent in 2009, 27.1 per cent in 2011, Socialist Party levels stagnated around 29–30 per cent between 2009 and 2015.

Among winnable positions, heads of lists are the ones with the highest chance of getting elected. In this case, cross-party differences have been fairly small in both countries. The lion’s share of these positions tends to be allocated to men, usually party officials with a strong hold in the district (predominantly the district-level party leader) or an incumbent with prominent public visibility nationwide. The norm underlying the selection of candidates for the number one position suggests that the gendered distribution of candidacies mirrors the gendered distribution of party offices,⁷ thus reinforcing patterns of exclusion for women and dominance of men in both party and elective posts. In Spain, though, the gender gap has steadily decreased (see Table 1). In 2015, 35 per cent of Popular Party lists and 50 per cent of Spanish Socialist Workers’ Party lists were headed by women candidates. In Portugal (see Table 2), the proportion of women as heads of lists stabilized at around 9 per cent between 2005 and 2011 and significantly increased in 2015, with 27.3 per cent of Socialist Party lists headed by women.

Overall, in both countries, in the two post-legislative quota elections, parties with voluntary quotas outperformed those with no such measure in their share of women candidates. Concerning women representatives, in Spain, the Spanish Socialist Workers’ Party repeatedly surpassed the Popular Party thanks to successive increases in the proportion of women in winnable positions. In Portugal, the legislative quota blurred the difference between the Socialist Party and the Social Democratic Party, despite their dissimilar commitment to equal gender representation, due to an analogous gendered distribution of winnable positions. Thus, our first theoretical expectation is confirmed in the Spanish case but only partially validated in the Portuguese case. This leads us to turn our attention to the varying degrees of institutionalization of party quotas.

SIMILAR COMMITMENT, DIFFERENT DEGREE OF INSTITUTIONALIZATION

The Spanish Socialist Workers’ Party and the Portuguese Socialist Party have used party quotas for almost 30 years but their degree of institutionalization presents remarkable differences. The first indicator under examination is the level of acceptance of party quotas. According to the Spanish Socialist Workers’ Party interviewees, quotas have long been regarded as ‘formal rules’ and the implementation of quotas has been taken for granted.⁸ As one interviewee put it, the party takes pride in ‘having enriched Spanish democracy with its commitment to gender equality in representation’.⁹ If anything, the main contribution of the legislative quota to the institutionalization of the party quota has been the vanishing of the residual opposition of local party selectors when drafting lists for municipal elections.¹⁰ In the Portuguese Socialist Party, although none of the interviewees expressed overt opposition to quotas, some of them depicted this measure as ‘a necessary evil’¹¹ and suggested that equal gender representation should be addressed instead by strengthening ‘women’s skills and capacity to hold political positions’.¹² Party selectors still see quotas as an ‘obligation’.¹³ As one interviewee pointed out, ‘people are resigned to it’.¹⁴

Acceptance may owe much to the ways in which quotas were introduced. In the Spanish Socialist Workers’ Party quotas were adopted by the 1988 party conference after a membership ballot had supported their introduction in 1986 – 67 per cent of positive votes (Verge 2012: 398). The successive party leaders have vocally supported the gender quota and urged all party levels to embrace it. In sharp contrast, in the Portuguese Socialist Party, the party quota introduced in 1988 was largely seen as the personal initiative of the secretary-general at the time, lobbied by some key female figures (Baum and Espírito-Santo 2012: 326). Indeed, lack of compliance for over a decade instilled the belief that party quotas were not adopted until the mid-1990s, a belief even shared by some of the former leaders of the women’s section.¹⁵ Quotas still remain an unpopular measure among many men and among women public officials elected before party quotas were effectively enforced in 1999.

Adherence to party quotas constitutes a second indicator of their degree of institutionalization. One might argue that if the composition of the most exclusive party decision-making body systematically meets the party quota, a clear example is set for the whole organization. The comparison discloses that party quotas have only been routinely applied by the Spanish social democrats. As Table 3 shows, in the Spanish Socialist Workers’ Party the composition of the Federal Political Commission has always met (and even surpassed) the party quota in use – bar in 1990. Alternatively, in the Portuguese Socialist Party the National Secretariat has never reached the minimum proportions established by the party quota, although the gap was marginal between 1999 and 2002 and in 2014. Concerning candidate lists, as presented in the previous section, compliance with the party quota was immediate in the Spanish Socialist Workers’ Party and was also respected when it was extended to winnable positions in 1994 and enlarged to 40 per cent in 1997. In Portugal, however, the Socialist Party needed over a decade to adhere to its own quota rules. A placement mandate to secure that at least one woman was included every three positions was only adopted in 2010 – in line with the legislative quota. While exceptional cases of non-compliance may be authorized by Socialist Party central party bodies, no exception has ever been contemplated in the Spanish Socialist Workers’ Party – far from it, regional branches are actively encouraged to outperform the quota provisions.¹⁶

Table 3 Women’s Representation in National Party Executives

Note: *A positive difference indicates that the percentage of women in the national party executive overcomes the minimum proportion set by the quota, whereas a negative number means that the minimum proportion was not reached.

Source: Own elaboration based on Espírito-Santo (2006) and Verge (2006) and updated by the authors using media and party reports on party conferences. The data reflect the composition of the party executive right after the respective party conference.

As discussed in the theoretical section, the formalization and centralization of candidate selection rules are intervening variables in the institutionalization of party quotas. Both parties have formal written rules on candidate selection but these are more detailed in the Spanish Socialist Workers’ Party, wherein party statute rules are complemented with specific written regulations detailing how the selection of public officers must be made. These rules standardize candidate selection throughout all districts (Kenny and Verge 2013). Conversely, in the Portuguese Socialist Party the candidate selection process is rather informal (Freire and Pequito Teixeira 2011). The written but vague rules leave more room for party selectors to distribute personal patronage. In fact, practices underlying candidate selection vary across districts and over time, depending on who the party leader is. Candidate selection is equally centralized in both cases but the intervention of the central level differs in both the early and the late stages of list-building.

In the Spanish Socialist Workers’ Party, candidate selection is initiated by district-level selectors in coordination with regional branches. The draft lists are voted by members in each district and thereafter submitted for approval to the National Electoral Committee (composed of senior officials elected at the party conference), which holds veto power and can invalidate the lists or modify the rank order of candidates. Finally, the highest party executive (the Federal Committee) ratifies the lists. In the Portuguese Socialist Party, according to the party statutes, the National Political Commission, following the secretary-general’s instructions, can nominate candidates for 30 per cent of the winnable positions in each district (with the number of incumbent deputies per district used as benchmark) and also indicate their rank order. These candidates are usually placed in the top positions of lists, invariably including the number one position.¹⁷ For the remaining two-thirds of candidates, decisions are generally made at the district level, with a varying role for the respective party executive, president and members.¹⁸ The district-level executive approves the draft list, which is thereafter submitted to the central level (the National Political Commission) for ratification. In practice, throughout the whole process informal contacts are made between the presidents of the district-level branches and the central party leader or the secretary for the organization of the party (a sitting member of the National Secretariat).¹⁹

The institutionalization of party quotas is also dependent on women’s agency. In the Spanish Socialist Workers’ Party, the leader of the Secretariat for Equality has sat ex officio on national party bodies since the mid-1980s and in electoral committees at all party levels since the mid-1990s. So, the women’s section can effectively supervise the implementation of quotas at all stages of list-building and has the capacity to push central party bodies to address gender imbalances whenever they occur, especially in the allocation of winnable positions.²⁰ The women’s section has also sought to institutionalize party quotas through periodical reports on progress on women’s representation,²¹ intra-party awareness-raising campaigns and the mainstreaming of parity into the political education courses that new members receive (Verge 2012: 400).

In contrast, in the Portuguese Socialist Party, although the leader of the women’s section also sits on the main party bodies at all party levels and has been charged with supervising the implementation of the party quota since 1998, as mentioned in party statutes (Espírito-Santo 2006: 48), she does not participate in the crucial informal arrangements between the presidents of the district-level branches and the central party leader who initially draft the lists. Subsequently, the lists are merely rubber-stamped by the National Political Commission (see below for details). So the women’s section has no influence in candidate selection,²² which is corroborated by this piece of anecdotal but illustrative evidence. When the personal assistant of the leader of the women’s section was informed that our interview would deal with candidate selection, her answer was: ‘What does it have to do with us? Do you know you are calling the Women’s Department?’

Thirdly, the institutionalization of party quotas can also be measured by the sanctions that non-compliance may entail. In the Portuguese Socialist Party, all interviewees stated they had never heard of lists being rejected nor rank orders being modified by the National Political Commission. While the 30 per cent prerogative would allow central party bodies to address gender imbalance in candidate lists, the data presented above (see Table 2) suggest that they have chosen not to do so. When anticipating electoral decreases, district-level selectors have conveniently ‘forgotten’ the party’s commitment to equal gender representation, and central party leaders have turned a blind eye. Also, as noted by our interviewees,²³ the consensual practice agreed by both district and central party leaders, and accepted by some leaders of the women’s section, is to place women candidates every third position on lists, thus fulfilling the minimum requirement of the quota law but failing to secure a significant proportion of women representatives.

In the Spanish Socialist Workers’ Party, since the late 1990s any strategic discrimination of resistant actors in list-building has been systematically addressed at its final stage by central party bodies through the modification of rank orders. Despite the existence of this formal mechanism to secure quota compliance, on some occasions, Spanish Socialist Workers’ Party women have resorted to informal means, as described by our interviewees, taking advantage of the party’s strong public commitment to gender equality.²⁴ In 2008, before lists had been ratified by central party bodies, a group of female incumbents leaked to the press that the brand-new quota law would only marginally increase women’s representation because parties, including the Spanish Socialist Workers’ Party, were abusing the minimum proportion for either sex (40 per cent) as if it was a quota for women (see El Pais 2008). To avoid undermining the party’s credibility, central party bodies modified rank orders to include more women in winnable positions.

Similarly, in 2011 it was leaked that the proportion of Spanish Socialist Workers’ Party women representatives was at risk of being halved. In effect, district-level selectors, in the face of the party’s potential electoral collapse, ‘remembered the old’ by disproportionately protecting male incumbents and male party officials at the lists’ top positions. As instructed by the secretary-general, the proportion of women as heads of lists was augmented to the highest party record ever (38 per cent) (see El Pais 2011, 2014). Still, the failure to reach at least 40 per cent women representatives (38.9 per cent) was perceived as a dramatic event, which pushed the Spanish Socialist Workers’ Party to adopt zipping as the new party quota in 2013.²⁵ With this reform, the Spanish Socialist Workers’ Party showed its willingness to keep championing equal gender representation in Spain.

Ultimately, party quotas are highly institutionalized in the Spanish Socialist Workers’ Party but poorly institutionalized in the Portuguese Socialist Party. A formally centralized selection process and an empowered women’s section have largely contributed to routinizing the implementation of both party and legislative quotas in the Spanish Socialist Workers’ Party, whose successive leaders have consistently supported equal gender representation and enforced it when deviations from strict compliance have occurred. To the contrary, in the Portuguese Socialist Party less formalized rules have trumped the coercing power of a similarly centralized candidate selection process. The party leadership has not enforced quota rules and the women’s section plays no role in list-building. Given that a higher institutionalization of party quotas leads parties to be more compliant with legislative quotas, our second theoretical expectation is confirmed.

CONCLUSIONS

The examination of the interactions between party and legislative gender quotas in Portugal and Spain indicates that legislative quotas are nested in parties’ candidate selection processes. For some parties, these are ‘new’ rules needing accommodation, whereas for those already using voluntary quotas they are, a priori, complementary to ‘party law’. In the former, existing ‘old’ gendered practices and norms limit the effective compliance with legislative quotas. In the latter, whenever the ‘old’ revives it can be counteracted provided that party quotas are highly institutionalized, meaning that they are widely accepted, routinely applied and enforced through sanctions whenever necessary.

The within-country comparison shows that, while compliance with legislative quotas has been higher among parties with voluntary quotas with regard to the proportions of women candidates in party lists, the introduction of quota laws has only pushed the innovator party to keep championing women’s representation in Spain. In this case, sustained disparities in the proportion of women candidates in winnable positions have not blurred the differences in the share of women representatives between parties with dissimilar commitments to equal gender representation. Conversely, in Portugal the legislative quota has produced rather similar outcomes regardless of prior party commitments, mainly due to the strongly gendered allocation of winnable positions by the two main parties.

The cross-country comparison of social democratic parties has further unpacked the interactions between party and legislative quotas. Party quotas present a much higher level of institutionalization in Spain. The empowerment of the women’s section in a genuinely formally centralized selection process and the strong commitment of party leaders have been pivotal to party quotas being taken for granted in the Spanish Socialist Workers’ Party. Consequently, legislative quotas have also been routinely enacted, with both formal and informal party sanctions guaranteeing effective implementation. Conversely, in Portugal the irrelevant role of the women’s section in a rather informally centralized selection process, coupled with a superficial commitment to quotas by most party leaders, have yielded a weak institutionalization of party quotas in the Portuguese Socialist Party, which has led to a poor compliance with the spirit of both party and legislative quotas.

In a nutshell, political parties are gendered organizations whose institutional setting matters for the success of quota reforms. Our findings suggest that a few elections under legislative quotas and even some decades of party quotas do not suffice to eliminate the gendered patterns that are deeply entrenched in candidate selection processes. Further comparative analyses are thus needed to tease out how gender quotas are layered in the broader party organization and to identify the practices and norms that hinder effective compliance. As such, we call on scholars to further examine the conditions under which party women’s sections or women’s agency more broadly can find the most advantageous setting to promote women’s representation in different party institutional settings.