Quotas and intersectionality: Ethnicity and gender in candidate selection

Introduction

There is growing scholarly attention on how equality policies affect diversity in politics. Much of this attention has been fuelled by intersectional theory, which conceptualises gender, class, ethnicity, race, age and sexuality as interrelated systems. Recognising such diversity and intersectionality challenges perceptions that gender equality is established when ‘women’ – without paying attention to ‘within-group’ differences – are deemed equal to ‘men’. Increasingly, it is acknowledged that gender equality is not fully realised when only the dominant group of women – in this case, white, highly educated, middle-class and heterosexual – are equal to their male counterparts (Collins, 1990; Crenshaw, 1991).

Intersectional analyses of candidate selection, election processes and political institutions are an emerging field (Hardy-Fanta, 2011; Mügge, forthcoming; Prins and Saharso, 2013; Squires, 2013). Research to date has focused on how intersecting identities influence the experiences of ethnic minority women in politics (Krook and Schwindt-Bayer, 2013: 568). Intersectionality, however, is not only about marginalised groups, but also concerns power relations between groups (McCall, 2005; Mügge and De Jong, 2013; Verloo, 2006) – a perspective that remains less developed in empirical studies (but see Hancock, 2009; Scola, 2007). This article evaluates the intersectional effects of gender quotas in the selection of ethnic minority candidates for political office in Belgium and the Netherlands. Our attention to gender quotas is rooted in the observation that gender equality in West European politics precedes demands for the political inclusion of ethnic minority groups (Bird, 2004). This raises the question as to whether gender quotas affect the inclusion of ethnic minority groups: do they foster the inclusion of ethnic minority women to the detriment of ethnic minority men? Or, are both evenly included when candidate lists are opened up to ethnic minority groups? Answering these questions will give us insight into how forms of affirmative action affect specific groups and how the representation of ethnic minority groups shifts when ‘new groups’ enter the political arena.

The extant literature on gender quotas raises contradictory expectations regarding the intersectional effects of quotas. A positive hypothesis suggests that gender quotas will increase ethnic diversity as quotas increase the chances of a more heterogeneous group of politicians coming to the fore. The terms for including one group (women) extend to other groups (ethnic minority groups). Especially for ethnic minority women, gender quotas also provide strategic opportunities to renegotiate their positions within parties. A contrasting hypothesis predicts that gender quotas will not increase the representation of (female) ethnic minority groups. As gender quotas cater to the interests of ethnic majority women, they will reinforce within-group inequalities.

Our study finds that the ethnicisation of candidate selection in Belgium and the Netherlands is highly gendered. The inclusion of ethnic minority groups has occurred especially through the integration of ethnic minority women, who are better represented than their male counterparts. The gender imbalance among ethnic minority candidates is most noticeable in Belgium, where there are gender quotas; nevertheless, the imbalance cannot be explained by gender quotas alone. Instead, gendered patterns of ethnic minority representation are best understood as part of a broader process of list formation – a complex ‘intersectional puzzle’ in which gender and ethnicity are among a candidate’s characteristics that gain relevance depending on the identities of incumbents. The presence of intra-party networks and party elites’ vote-seeking and power maintenance strategies are also key explanations for how the ethnicisation of candidate lists plays out in terms of gender.

The following two sections assess the literature and outline our methodology. A contextual section then presents an overview of the similarities and differences in ethnic minority representation in Belgium and the Netherlands. The empirical body of the article offers three explanations for the variation between the cases: list formation, intra-party networks and the strategies of party elites.

Gender quotas, diversity and intersectionality

The question as to whether gender quotas have an intersectional impact has generated contradictory expectations in the literature on quotas. Some scholars assume that gender quotas improve the odds that a more heterogeneous group of women, including ethnic minority women, will be elected. Researchers indeed show that gender quotas have benefitted rural women (Mehta, 2002), women with lower-status occupations (Bird, 2003), young women (Britton, 2005) and ethnic minority women (Hughes, 2011). There are multiple possible explanations for these trends. First, terms for including one group may be extended to other groups (Norris, 2004). Because women are significantly under-represented, any policy increasing their presence may enable a more heterogeneous group of women to be elected (Mansbridge, 1999; Paxton and Hughes, 2007). Second, gender quotas provide ethnic minority women with strategic opportunities to strengthen their candidacies (Paxton et al., 2006), especially as the women’s movement is now more inclusive than it was in the past (Weldon, 2006). Third, the election of more ethnic minority women through gender quotas may be related to ethnic majority men’s efforts to remain in power. Selecting ethnic minority women allows party leaders to ‘kill two birds with one stone’, preserving in the process white male dominance (Hughes, 2011). Ethnic minority women are also sometimes perceived to be easier to negotiate with as they can switch between group loyalties (Fraga et al., 2005), and easier to exclude from real power through ‘racing-gendering’ (Hawkesworth, 2003), which further bolsters the image of the ‘experienced white male’ politician (Meier et al., 2006: 50).

Another stream of literature suggests that gender quotas do not increase ethnic minority representation. First, dominant subgroup experiences (e.g. of white women) are often taken as the norm in affirmative action. Singling out one dimension of inequality thus reinforces within-group inequalities (Collins, 1990; Hancock, 2004; hooks, 2000; Mansbridge, 2005; Mink, 1998; Strolovitch, 2007; Weldon, 2011). Indeed, gender quotas generally cater to ethnic majority women, while ethnic minority quotas support ethnic minority men (Darcy et al., 1993; Krook and O’Brien, 2010). In theory, ethnic minority women could benefit from both gender and ethnic minority quotas. In reality, they likely benefit from neither (Hancock, 2009); their political representation would only improve through tandem quotas that explicitly target ethnic minority women (Hughes, 2011). Second, when gender quotas require elites to suddenly recruit large numbers of women, they tend to rely on existing networks. This results in the recruitment of elite women close to men in power (Cowley and Childs, 2003; Franceschet et al., 2012). Third, gender quotas are not specifically designed to address within-group inequalities. Research on gender quota adoption highlights the importance of women’s mobilisation (Krook, 2006, 2009; Krook and Schwindt-Bayer, 2013). However, ethnic minority women’s organisations generally do not participate in these campaigns. Other explanations given for quota adoption – elite strategies, political norms and international norm diffusion and learning (Krook, 2009) – similarly give no reason to expect that gender quotas would increase ethnic minority women’s representation. In sum, the intersectional effect of gender quotas remains puzzling. Through a comparison of Belgium (with legally binding gender quotas) and the Netherlands (without such quotas), we put the contradictory assumptions to the empirical test.

Methodology and case selection

Belgium and the Netherlands are ‘similar systems’. They share a history of: pillarisation (and de-pillarisation); mechanisms of power-sharing and civil society consultation in policymaking; multi-party systems; and proportional list electoral systems (Deschouwer and Lucardie, 2003). Although voters can cast preferential votes for individual candidates, political parties in both countries play leading roles in the recruitment and nomination of candidates. Because individual candidates need a large number of preferential votes to breach the list order, they rarely surpass better-placed candidates to get elected (Andeweg, 2008; Craeghs and Dewachter, 1998). The Netherlands and Flanders (a Belgian sub-state) also have comparable ‘multiculturalist’ citizenship models (although the backlash against multiculturalism in the Netherlands has caused an overall shift towards assimilationism). The other Belgian francophone sub-states lean more towards the assimilationist-republican citizenship model (Loobuyck and Jacobs, 2006). To strengthen the similar system design, we focus on candidate selection in Belgium’s Flemish parties.

Belgium and the Netherlands also address diversity issues in remarkably similar ways (Celis et al., 2012). Since the 1970s, gender concerns have been institutionalised in political parties, with several parties establishing intra-party women’s groups (Van de Velde, 1992; Van Molle and Gubin, 1998). One important difference between the two countries concerns their use of gender quotas. In Belgium, many parties have had voluntary quotas or targets since the 1970s, while the Belgian federal government installed legally binding gender quotas in 1994 and 2002 (Meier, 2000). The first quota law in 1994 stated that not more than two-thirds of candidates on electoral lists could be of the same sex. But as it did not target the top positions on the lists, it did not lead to a substantial increase in the proportion of female representatives. The second, more effective, quota law of 2002 introduced gender parity on lists, as well as an alternation of the sexes in their top spots. The Netherlands, in contrast, has been much more reluctant to apply structural measures to increase the number of women in parliament. Although the Greens, the Christian Democrats and the Social Democrats have adopted voluntary targets, legally binding gender quotas have never been applied (Mügge and Damstra, 2013). Despite these different trajectories, women’s descriptive representation at the national level in Belgium and the Netherlands is similar: women currently make up 38.7% of the members of the Dutch Second Chamber and 39.3% of the members of the Belgian federal Chamber of Representatives (for 2013, see: www.ipu.org).

Ethnicity has grown increasingly salient in understandings of diversity in Belgium and the Netherlands over the past decade(s). In this article, we define ‘ethnic minority groups’ as non-Western ethnic minority groups subject to immigration or integration policies, whose political representation has become salient and politicised. This mostly includes ethnic minority groups who, or whose parents, migrated from the former colonies (Surinamese and Antilleans in the Netherlands; Congolese in Belgium) or within guest worker programmes (Turks and Moroccans in both countries). To the best of our knowledge, no prior qualitative study has examined the effects of gender quotas on ethnic minority groups in Europe with migrant backgrounds. Existing work on gender quotas that address ethnic minority groups are large-N studies and largely focus on indigenous minority groups.

Our analysis draws on 23 semi-structured interviews with representatives of the major political parties in each country that have at least one ethnic minority Member of Parliament (MP) in the national parliament: the Green Party (the Dutch GroenLinks and the Flemish Groen); the Social Democratic Party (the Dutch PvdA and the Flemish Sp.a); the Christian Democratic Party (the Dutch CDA and the Flemish CD&V); and the Liberal Party (the Dutch VVD and the Flemish Open VLD) (see Appendix). In Belgium, the main party families typically consist of a Flemish and a francophone party; our interviews were limited to the Flemish parties to enhance comparability with the Dutch data. Respondents were selected on the basis of their expertise in recruitment and selection or diversity. References to interviews are indicated with a ‘N’ for respondents from the Netherlands and a ‘B’ for Belgian respondents. The interviews took on average 90 minutes and were conducted by the authors between March and June 2013. The first part of the questionnaire sought to identify the central actors supporting (or resisting) the recruitment and selection of ethnic minority groups, such as the party’s leadership or women’s organisations. A second set of questions addressed selection criteria to determine how the recruitment and selection of ethnic minority candidates is gendered, while the third part questioned how parties integrate diversity on candidate lists and into party structures. Finally, we inquired about patterns of competition or collaboration between identity groups.

Ethnic minority groups in parliament

The first ethnic minority representatives entered the Dutch and Belgian parliaments for the Social Democrats, in 1986 and 1995, respectively. Ethnicity since then has become steadily more important in party politics. The majority of our respondents underscored the importance of including ethnic minority groups, often for electoral reasons: ‘If you do not have any representatives from this particular group, you lose voters’ (N8). Indeed, the presence of ethnic minority groups in Belgian and Dutch national politics has gradually increased over the years. At the time of writing (2013), 15 out of 150 representatives in the Dutch Second Chamber (10%) have non-Western migrant backgrounds, close to their share of the population. Nine out of 150 representatives in the Belgian Chamber (6%) have non-Western migration backgrounds, below their proportion of the population (Wauters and Eelbode, 2011).

The recruitment of ethnic minority candidates seems most natural in left-wing parties. Both Social Democratic parties in our study argued that ‘diversity is part of our genetic make-up’ (N8, B3). In contrast, the Liberal parties explained that recruiting ethnic minority groups does not abide with liberal ideology: ‘It is not that we are hesitant to involve ethnic minorities in politics, but we consider people to be individuals and not members of a particular community’ (B8). The importance of ethnicity also depends on the ethnic composition of the population. Especially in larger cities, parties feel the necessity to include more ethnic minority candidates on their lists: ‘Elections in every large city will be won or lost with the ethnic vote’ (B2). In places with fewer ethnic minority groups, parties tend be more cautious: ‘The party was really very careful and very scared to play the ethnic card at the wrong time and place’ (N4).

Until the late 1990s, most ethnic minority MPs in both countries were male. This has changed since 1998/1999, when ethnic minority women entered the national/federal parliaments in Belgium and the Netherlands (see Figures 1 and 2). Nevertheless, the gender imbalance remains stronger in Belgium than in the Netherlands. In the Dutch parliament, 70% of ethnic minority representatives in the Second Chamber between 2002 and 2010 were women, while there has been an almost equal number of ethnic minority women and men elected since 2010. In Belgium in the period 1999–2003, the two ethnic minority MPs were both women, while women have constituted 80% of elected ethnic minority representatives since 2007. This raises the question as to whether gender quotas favoured ethnic minority women to the detriment of ethnic minority men and were used to preserve ‘white male dominance’. However, as the next sections show, gender quotas are neither the sole nor the dominant explanation for the representation of ethnic minority women.

OPEN IN VIEWER

Figure 1.

Ethnic minority MPs by gender in the Netherlands.

OPEN IN VIEWER

Figure 2.

Ethnic minority MPs by gender in Belgium.

List formation as an intersectional puzzle

The ethnicisation of Dutch and Belgian politics occurred mostly through the integration of ethnic minority women. This is surprising, as several studies have suggested that women of colour experience ‘double barriers’ or ‘double burdens’ in society and politics (Hull et al., 1982; St. Jean and Feagin, 1998). One explanation for the similar results in Belgium and the Netherlands relates to their electoral systems and processes of candidate selection. Candidate lists in proportional electoral systems are more likely to include traditionally marginalised groups and to reflect the population in its full diversity than majoritarian systems (Leyenaar, 2004; Norris, 2004). Indeed, all respondents stated that the recruitment of candidates and the formation of lists is a complex puzzle in representing gender, ethnicity, regional background, age and expertise:

Our first concern is not whether we have enough male or female candidates. Our first question is: who do we have among the incumbents and where do we have room for new people?… If we have three female incumbents and one male incumbent, we include a male newcomer. The profile of this new male candidate ideally complements the profiles of the four incumbent candidates…. It’s a puzzle and candidates’ gender and ethnic origin are part of this puzzle. (B8)

The forming of lists is an intersectional puzzle in which gender and ethnicity are but two characteristics. Ideal candidates do not represent a particular segment of society. To appeal to different groups simultaneously, a list as a whole needs to represent a combination of social characteristics. Hence, the profile of the incumbents largely determines the profile of the other candidates – with the latter cashing a ‘complementarity bonus’. This means that a young female ethnic minority candidate is not selected simply because there are few women or ethnic minority candidates. She is selected because, on top of her competence, her specific profile – being a young ethnic minority female – complements those of incumbent candidates, who are often senior white men.

If list formation is an intersectional puzzle, this may increase competition between groups. There is general competition between candidates for strategic places on the list, the so-called ‘eligible spots’. Depending on the size of the party – and during the campaign, through polls – parties are able to estimate the number of seats that they are likely to win. The number of these strategic places on the list may be few. This means that there is stiff competition for relatively few spots in which a large number of characteristics must be represented. Contrary to expectations generated by the literature, our interviews revealed that this competition does not occur between different groups (for instance, between women or ethnic minority groups, or between ethnic minority women and ethnic majority women). Instead, competition plays out between candidates with similar profiles: between ethnic minority women within the same party, between male candidates from the same region, and between women with similar socio-economic backgrounds. For example, several Dutch interviewees emphasised fierce competition between ethnic majority women:

There is always competition in politics, but women are not particularly nice when push comes to shove. Men have their old boys’ network … women have the cat’s box [krabbenmand] which is the opposite…. If one woman stands up, the others pull her down. (N10)

The selection processes work in surprisingly similar ways in both countries. Gender quotas did not reproduce a strict hierarchy of gender over ethnicity in the candidate selection process in Belgium. The search for female candidates after the implementation of the 2002 quotas temporarily favoured ethnic minority women as they were needed to comply with the quotas and complemented incumbent ethnic majority profiles. But this effect will arguably disappear when incumbent candidates in Belgium are increasingly gender-balanced (and when ethnic minority men complement incumbent ethnic majority women at the top of the lists). Although less dramatically, a similar pattern can be observed in Dutch parties that voluntarily implemented gender targets (Mügge and Damstra, 2013).

The role of intra-party networks

Alongside the ‘complementarity bonus’, ethnic minority representation can benefit from the institutional recognition of ethnic minority groups within political parties. Such intra-party groups pressure party leaderships to recruit ethnic minority candidates and select ethnic minority groups for the top list positions. The recognition of ethnic minority groups within parties differs in important ways across the two cases. In the Netherlands, ethnic minority groups were set up in the Social Democratic and Christian Democratic parties in the mid-1980s and within the Greens in the 1990s. Ethnic minority women’s groups followed in the mid-1990s. These networks, while formally recognised, did not enjoy the same status as the women’s sections. Ethnic minority women’s groups have been particularly successful as they could use the institutional channels of the women’s sections (Mügge and Damstra, 2013). In Belgium, in contrast, no party except the Greens has ever established an intra-party diversity group. The Belgian interviewees repeatedly alluded to reluctance to integrate ethnicity on a structural basis in the parties’ organisation. Some parties feared that a separate section would essentialise ethnic minority groups and put too much emphasis on their ethnic background: ‘We believe that a person is more than his/her ethnic origin…. It is not the only thing that connects people. Even if you have a migrant background, you do not want to be pinned down to it’ (B5). Other parties envisioned a horizontal rather than a vertical integration of ethnic minority groups: ‘We only have such separate sections for women, youth and the elderly. We do not have them for ethnic minorities, because we want attention for this latter group to be mainstreamed’ (B4).

Differences between Belgian and Dutch parties in their institutional embedding of ethnicity had important consequences for the political representation of ethnic minority candidates (both men and women). Despite their status as advisory bodies, ethnic minority groups within Dutch parties also lobby for the nomination of ethnic minority candidates. The first candidates promoted by these groups were men, largely because ethnic minority groups initially worked separately from women’s groups. Soon, however, women’s groups and ethnic minority groups sought ways to cooperate rather than compete over the nomination of candidates. The founder of the CDA ethnic minority group referred to its reciprocal relationship with the CDA women’s group: I made sure that … I supported the majority women of the CDA women’s group, and then I asked them to support my candidates as well’ (N3). This proved to be a winning strategy as both female candidates put forward by the ethnic minority groups were well ranked on the party list.

Ethnic minority women’s groups in Dutch parties supported the candidacies of ethnic minority women and provided institutional resources once they were elected, organising networking opportunities and preparing them for their jobs as representatives (N9). These groups were most powerful in the 1990s and the beginning of the 2000s, which may explain the relatively strong presence of ethnic minority women in Dutch politics at the time (see Figure 1). The intra-party groups gradually lost their clout in the anti-multicultural climate of the 2000s; at the time of writing, most ethnic minority networks have been dissolved. The Christian Democrats and Greens still host such networks but resist organising on the basis of culture or ethnicity: ‘colour’ is the new principle. This may explain the narrowing of the gender gap among ethnic minority groups since 2010 (see Figure 1).

In Belgium, political parties did not witness the development of intra-party ethnic minority organisations. Some parties maintained contact with civil society organisations for policy-related reasons, but the majority of Belgian respondents resented the idea that ethnic minority organisations would influence the selection process and intervene in internal party matters. Autonomous immigrant organisations sometimes pushed their own candidates, especially targeting left-wing parties:

I received a lot of phone calls from these organisations, lobbying for an alderman with a Turkish background. I was very surprised that they plucked up such courage to do so. I cannot imagine that I would lobby for a female candidate. (B3)

Diversity sections are likewise absent in the women’s groups of the Flemish parties. While the Christian Democratic, Socialist and Green parties claim that the women’s groups support their parties’ attempts to bring in more ethnic minority groups, this support has thus far not led to any structural solutions. Given the absence of groups lobbying for their presence, the descriptive representation of ethnic minority women depends on other political opportunities, such as the presence of progressive party leaders and ethnic minority incumbents.

According to the interviewees, the adoption of binding gender quotas in Belgium did not present a window of opportunity for ethnic minority women. The quotas obliged parties to actively recruit women, for which they turned to their women’s groups. However, due to the absence of ethnic minority bodies and the limited integration of diversity measures in the women’s groups, the promoted female candidates were mostly ethnic majority women. Some parties, especially the Socialists, did try to place several ethnic minority women on the top of the list. However, overall, gender quotas did not boost the presence of ethnic minority women in the same way that they did for ethnic majority women. Although there is no direct causal link, the absence of intra-party networks seems to have rendered the political representation of ethnic minority groups more difficult in Belgium. Several ethnic minority respondents mentioned that they sometimes felt isolated and abandoned by their parties.

The strategies of party elites

The previous section showed that ethnic minority representation in Belgium mostly involves women, but that this was not the result of intra-party networks. Contrary to our expectations, gender quotas also fail to explain the gender imbalance in ethnic minority representation. According to our interviewees, the distorted gender balance in Belgian ethnic minority representation is due to party leaders’ anticipation of voters’ response to ethnic minority candidates and to leaders’ efforts to maintain the balance of power within their parties.

First, political parties in Belgium assume that ethnic minority women are more attractive to voters. The female politician of ‘the first generation of ethnic minority representatives’ was typically highly educated and well integrated in Belgian society, did not have any ties to ethnic minority civil society organisations, and did not wear the Islamic veil or headscarf: ‘Ethnic minority women are the softer face of the emancipation of ethnic minority candidates’ (B8). The selection of Muslim women was supposed to lessen the threat of Islam: ‘A Muslim woman is perceived as a woman, less as a Muslim’ (B1). Ethnic minority men, in contrast, were less attractive to party selectors, who expected voter rejection based on racist stereotypes about ethnic minority men being criminals. Ethnic minority women were expected to attract the ethnic vote and at the same time to not alienate the existing electorate. Indeed, they provided an extra reason to vote for ethnic minority women: sympathy with repressed Muslim women. Complementing the view that ethnic minority women are oppressed is the view that ethnic minority men are the oppressors:

The image of the ethnic minority woman is that of an oppressed woman, in need of help, deserving a voice…. Voters prefer to cast their vote for ethnic minority women … that was often a sympathy vote: we support women’s emancipation. It was meant to be a positive thing, but the side effect was that in case of ethnic minority men, one thought: they want to keep their wives at home. (B7)

Second, according to our respondents, ethnic minority men are perceived as being more threatening to the internal balance of power, that is, to the power held by ethnic majority men. Ethnic minority male candidates play the ‘political power game’ like ethnic majority men and are therefore a greater threat to white men in power: ‘This particular masculine culture in politics … ethnic minority men inscribe themselves more easily into it … than women. Going to the pub after the meeting is still a bit more difficult especially for women with children’ (B2).

In contrast, ethnic minority women, especially younger women, are seen as complementary to incumbents who are predominantly white, older, middle-class men. In the search for complementary candidates when forming lists, young ethnic minority women indeed benefit more often from a ‘complementarity bonus’ than ethnic minority men.

Both of these reasons for why political parties prefer ethnic minority women were mentioned by interviewees in the Netherlands, albeit to a lesser extent. Representatives of Dutch parties did not unanimously argue that ethnic minority women are preferred over ethnic minority men because they are expected to attract more voters. Only one interviewee explicitly stated that ethnic minority women are perceived as less threatening than their male counterparts, who may suffer from the ‘angry young man stereotype’ (N9). Another argued that ethnic minority women are a ‘safer choice’: ‘A woman uses different instruments to give a nice impression, to be good: it’s always safer. A man is more focused on the content. So if there is little room for diversity, a woman is a safer choice’ (N6).

Conclusion

This article has assessed how legally binding gender quotas affect the representation of female and male ethnic minority MPs. By comparing the selection processes of ethnic minority candidates in Belgium and the Netherlands, we have shown that although gender quotas do not obstruct the election of ethnic minority women, they are not the key explanation for their electoral success. Concurring with the general literature on how quotas affect the representation of women, it is the interaction between quotas – or voluntary target numbers – and other factors that is decisive. In the current study, these factors include the intersectional puzzling behind the formation of candidate lists, the degree to which gender and ethnicity are institutionalised within parties, and the strategies of party elites to maximise electoral support.

Our analysis revealed that the intersection of gender and ethnicity has multiplying effects. Ethnic minority men were disadvantaged in both countries, suffering from negative stereotypes in Belgium and from mixed intra-party networks of ethnic minority groups having weaker positions than ethnic minority women’s groups in the Netherlands (due to the fear of exclusive ‘ethnic’ lobbies). The intersection of two disadvantaged positions thus turned into an advantage for ethnic minority women in both countries, who could embody the demand for women and ethnic minority groups in a single candidate. In the Netherlands, intra-party ethnic minority networks influenced recruitment and selection processes, while ethnic minority women profited from the infrastructure of women’s networks and committees created by ethnic majority women within political parties. In Belgium, in contrast, the party leadership left next to no room for the lobby of ethnic minority or women’s networks.

Moving beyond the experiences of specific groups, we saw that gender and ethnicity are part of a broader intersectional puzzle of forming candidate lists. Selectors aim for lists in which numerous identities are represented by a limited number of candidates with whom the electorate can identify. Given the electoral demand to include ethnic minority and female candidates and their under-representation among incumbents, ethnic minority women were able to cash a ‘complementarity bonus’. In their drive to include ethnic minority groups, political parties in both countries favoured women, resulting in a gender imbalance among ethnic minority representatives. However, the leading positions on the list – the safe choices – generally went to white men: the way that the intersectional puzzle was completed depended on what was missing from their profiles. The party specificity in this game is for future research to unravel.

The similarities in the gendered selection of ethnic minority candidates in Belgium and the Netherlands speak against attributing too much importance to gender quotas. While the gender imbalance among ethnic minority candidates was more pronounced in Belgium, our analysis did not find indications that ethnic minority women were more strongly favoured due to quotas. Much of the imbalance can be traced to how political parties manage ethnic diversity.