29 Descriptive Representation of Marginalized Groups in Dutch Politics

I want a dyke for president … I want to have a president that had an abortion at sixteen … and has been unemployed and layed [sic] off and sexually harassed and gaybashed and deported … I want someone … who has been in love and been hurt … who makes mistakes and learned from them. I want a Black woman for president.

Leonard (1992), cited in Fahs (2020, p. 27)¹

Politicians who share citizens’ lived experiences are of particular importance for those who have no or few elected or appointed officials they can identify with, such as those living as a woman, Black, gay, with a disability, or a combination thereof. In democracies around the world there are sizable groups of people who have not yet been able to demonstrate that people like them have the ‘ability to rule’ (Mansbridge, 1999). (Former) prime ministers like Jacinda Ardern (New Zealand) and Sanna Marin (Finland), who go on maternity leave, breastfeed during a United Nations meeting, or like to party in their time off, have become global role models for girls and women. The massive attention their seemingly normal actions attract demonstrates how ‘atypical’ female political leadership still is.

Atypical role models are also a rarity in the Netherlands, where the first woman prime minister still has to be inaugurated and where White cisgender heterosexual abled men are the ‘norm’ among political representatives and executives (Ministerie van Binnenlandse Zaken en Koninkrijksrelaties, 2022; Mügge, Runderkamp, & Kranendonk, 2019). In this chapter we study the political representation of: women; people with a ‘migration background’; lesbian, gay, bisexual, trans (LGBT) citizens; and disabled citizens in the Netherlands. These four overarching categories of marginalized groups are at the centre of current so-called identity politics. The groups are politicized as part of the newer cultural axes of the party system (see also Harteveld & Van der Pas, this volume, on new cleavages and Otjes & De Lange, this volume, on the party system).² How have women, citizens with a ‘migration background’, LGBT citizens, and disabled citizens—descriptively been represented and how has this developed over time? What explains variation among the political representation of these various groups in the Netherlands?

The literature on descriptive representation takes its point of departure from the work of Pitkin (1967). Pitkin conceptualizes the presence of a recognizable group as the representative not acting for others, but ‘standing for’ them, based on resemblance or shared experiences. This literature is best developed on women’s representation, but studies increasingly include other groups that are structurally underrepresented in politics, such as ethnically/racially minoritized, LGBT, and disabled citizens (De Jong & Mügge, 2023; Evans & Reher, 2023; Sobolewska et al., 2018). In the past decade, an intersectional approach to descriptive representation has been introduced to reveal how mechanisms of political inclusion and exclusion do not work in the same way across groups (Smooth 2011). For instance, what works to increase the number of Black women politicians might not work for Latina politicians (Holman & Schneider, 2018; Reingold et al., 2020).

Looking at descriptive representation of historically disadvantaged groups from the Second World War until the present, we show that, while Dutch politics is still far from equal, it is gradually becoming more representative. Women’s representation has been campaigned for since the 1970s. Large-scale attention, also within political parties, to the political representation of ethnically minoritized, LGBT, and disabled citizens is more recent. At the same time, the opposition to (certain types of) diversity by a broad range of anti-feminist, anti-trans-, anti-Muslim-, and anti-immigrant groups has become more vocal, including in representative bodies.

In what follows, we review relevant explanatory factors for descriptive representation and describe the historical context of group representation in the Netherlands. Subsequently, we provide an overview of the descriptive representation of the groups under study and analyse representation per group over time. Then, we discuss explanations for descriptive representation at the individual, party, and national levels, the different phases of political careers, and intersectional variation. Finally, we present an agenda for future research.

In its most basic definition, ‘descriptive representation’ refers to the number or proportion of bodies of a particular group in a political layer, such as a parliament or municipal council. The term was coined by Pitkin (1967) in her book The Concept of Representation, in which she explains how descriptive representation is related to the representation of the interests of a particular group, so-called substantive representation. Feminist work argues that female politicians are best suited to represent the interests of women due to their shared experiences as women, and that representation is important in and of itself for reasons of social justice, parity, and countering of negative stereotypes (Mansbridge, 1999; Phillips, 1995). Moreover, descriptive representation is vital for the functioning of democracy, since a political body is only truly representative if it corresponds adequately to the whole nation (Pitkin, 1967). Accordingly, empirical studies demonstrate that descriptive representation of women (and other social groups) has substantial effects on decision making, policy outcomes, trust, social cohesion, and the acceptance of election results (Haider-Markel, 2007; Reynolds, 2013; Ruedin, 2013).

Scholarship that explains variations in descriptive representation is best developed on women. This work distinguishes micro-, meso-, and macro-level explanatory factors (Wängnerud, 2009, p. 59). Macro-level factors refer to the total cultural and institutional packages of norms and rights, such as the year of women’s suffrage, the gender equality culture, the welfare state system, and the electoral systems. Meso-level factors comprise party ideology, party organization, party gender quotas, and the strength of a women’s movement. Micro-level factors include voting preferences and motivations of women to run for office.

Additionally, descriptive representation is understood as a pipeline process encompassing four phases: (1) aspirants, (2) candidates, (3) elected, as well as (4) (non-)retention (Gouglas et al., 2018; Lovenduski, 2016). Aspirants are those citizens who develop an ambition to be elected. Candidates are the portion of aspirants who are placed on an election list by a political party. The elected are the portion of candidates who win an election. Retention is about those who stay and run for election again. These phases are directly related to a combination of macro-, meso- and micro-level factors. For instance, the available pool of aspirants depends on the gender equality culture and whether girls and boys develop political interest in similar amounts. Recruitment and selection of women is influenced by, for example, gender targets and the availability of a strong women’s network, while whether candidates are elected depends on list positions, the electoral rules, and political campaigns. Finally, whether women politicians move up, stay, or drop out depends on the party and the political culture as a whole, which can be hostile and not gender sensitive (Childs, 2016). Research shows that similar factors play a role for, among others, ethnically minoritized citizens (Bloemraad & Schönwälder, 2013). Finally, an intersectional lens enables scholars of descriptive representation to analyse variation within and between groups (Mügge et al., 2018). Scholars have empirically revealed how explanatory factors are gendered and racialized/ethnicized in all aspects (Brown, 2014; Hawkesworth, 2003; Mügge, 2016).

Given the absoluteness of voting rights (i.e. formal or judicial representation) as a macro-level explanation of descriptive representation, we first discuss the voting rights of the different groups under study. Then, we turn to the groups’ presence in various legislative bodies and over time.

In 2022, 93% of the Dutch adult population—18 and older—had national voting rights (Statistics Netherlands, 2022). This figure has not always been so high. Dutch women obtained passive voting rights in 1917 and active voting rights in 1919. Since 1986, non-nationals residing in the Netherlands for at least five years have had voting rights in local elections. Dutch citizenship is required to vote in national parliamentary and provincial elections (Rijksoverheid, n.d.). The majority of the 7% of the adult population that have no voting rights have no Dutch nationality (see Van der Kolk, this volume).

Voting rights have been more ambiguous for citizens with a (mental) disability and LGBT citizens. Citizens who were diagnosed as ‘mentally incapable’ and had a guardian had no voting rights until a constitutional change in 2008 (DeNederlandseGrondwet.nl, n.d.). Belonging to a sexual minority was never a ground for withholding one’s right to vote. However, until 1971, sodomy was under certain conditions reason for incarceration. The constitution allowed the removal of voting rights as a punishment in addition to a minimum sentence of one year’s deprivation of liberty. Of the estimated 5,000 people who were persecuted based on the anti-sodomy law, roughly half were found guilty and received average sentences of three to six months’ incarceration. Others were forced into an ‘asylum for psychopaths’ for two years or longer (IHLIA, 2011). While the purpose is undocumented to our knowledge, this was probably a tool to strip homosexual citizens of their voting rights. Today, taking away voting rights is restricted to those convicted of a ‘serious violation of the foundations of the Dutch constitution’ (DeNederlandseGrondwet.nl, n.d.; Kiesraad, n.d.).

Voting rights shape who can vote and who can be elected. The most recent Dutch Parliamentary Election Study (Jacobs et al., 2021) demonstrates that there are hardly any differences between the turnout of men and women in first-order elections. Homosexual men and women seem to be slightly more likely to vote than heterosexual or bisexual citizens. Those who report bad, and particularly very bad, health—as a proxy for disability—vote considerably less. The ‘Minority Integration Survey’ (Sociaal en Cultureel Planbureau, 2020) demonstrates much lower turnout rates among citizens with a migration (i.e. Caribbean-Dutch, Iranian, Moroccan, Somalian, Surinamese, or Turkish) background (see Den Ridder & Ostaaijen, this volume).

How are the groups under study descriptively represented across legislative institutions and over time? Table 29.1 provides an overview of the most recently available percentages. It provides the data for the national, regional, and local levels based on the following categories: women; migration background; Muslim; lesbian, gay, bisexual, trans, intersex, queer, asexual (LGBTIQ+); and those with a disability. Figure 29.1 portrays the evolution of the representation of women and of citizens with an (ethnic-minority) migration background at the national and local levels. Sufficient data were only available for these two groups. Figure 29.2 shows the current parliamentary representation by party ideology.

Dutch women are descriptively not equally represented and never have been. Overall, there has been a slow upward trend since the 1960s. Yet, there is considerable variation longitudinally and across the legislative bodies. In rare spatiotemporal locations (e.g. Nijmegen, Utrecht, and Renkum), we even find parity or overrepresentation. However, in the last parliamentary cycles, the percentage of women in the House of Representatives hovered between 35% and just over 40%. In the monthly ranking of ‘women in parliament’ of the Inter-Parliamentary Union (Inter-Parliamentary Union, 2023), the Netherlands currently ranks twenty-sixth, between Austria and Ethiopia.

Women’s representation in parliament is higher among postmaterialist parties and left-wing parties (e.g. Democrats 66 (D66), GreenLeft (GL), Labour Party (PvdA), Socialist Party). In far-right parties, fewer than one in four members of parliament (MPs) are women. The (orthodox) Political Reformed Party (SGP) has never had a woman MP and excluded women candidates until recently. Following a court ruling in 2006, the SGP had to admit women to membership. Since 2013, women have been allowed to run for office for the party. Shortly thereafter, the first SGP woman councillor was elected.

When the Rutte IV cabinet was inaugurated in 2021, it was almost gender equal—with 48% women cabinet members—for the first time in history. At its most senior level, that of ministers, the cabinet was gender equal. Percentages are much lower across subnational executive bodies, and women’s representation in regional parliaments and municipal councils varies enormously. Among the 300-plus municipal councils, the numbers vary between 5% and 61% (see Table 29.1). More generally, there is a positive relation between municipality size and women’s representation: the representation of women is higher in bigger municipalities with a population of more than 100,000 inhabitants (Ministerie van Binnenlandse Zaken en Koninkrijksrelaties, 2022).

The representation of citizens with a migration background and/or ethno-religious minoritized groups is far from unambiguous: categorizations are problematic and self-identification and how others ‘read’ politicians is not always accounted for in formal figures. The different columns and footnotes in Table 29.1 reflect this. Moreover, descriptive representation is often reported for ethnically minoritized citizens as a homogeneous group, while degrees of representation differ considerably across and within those ascribed groups. Citizens from one ethnic-religious minoritized group may only partly identify with ethno-religious minoritized politicians more generally.

Since the most recent parliamentary elections, the proportion of MPs who were born or had at least one parent born in Africa, Asia, or Latin-America³ (i.e. ‘non-Western’) is currently roughly equal to their proportion of the population, and compared to other European immigration countries, citizens with a migration background are relatively well represented (Mügge, Van der Pas, & Van de Wardt, 2019). But there is variation across groups. In particular, Muslim (Table 29.1), Turkish-Dutch, and Moroccan-Dutch citizens (Van der Zwan, 2017) are somewhat overrepresented in parliament in the last few cycles. The numbers increased from the start of our data, reflected in Figure 29.1, until the turn of the millennium, then stabilized until 2018, after which numbers increased again.

When we look at migration background and gender combined, we see that women with a migration background have been better represented than their male counterparts (Mügge, 2016). But it was only in the 2021 elections that the first Muslim woman with a headscarf was elected to parliament. Until the 2023 elections, two women in parliament wore a headscarf, both representing postmaterialist parties. On the basis of both their gender and religion, and the politicization thereof, these politicians are highly visible.

Parties that were founded and are led by ethnically minoritized MPs (Think/Equal (Denk) and AsOne (BIJ1)), and which explicitly advocate racial, ethnic, religious, and intersectional diversity, are represented only by ethnically minoritized MPs (see Figure 29.2). Postmaterialist and left-wing parties are the traditional champions for diverse candidates. In the run-up to the 2021 parliamentary election, four Black women candidates from the PvdA, D66, GL, and Denk formed the ‘Dutch Squad’. Inspired by the Democratic US Squad, they encouraged Black citizens to vote and increase the number of Black MPs.⁴ Far-right parties have either no descriptive MPs from the groups under study, or the lowest proportion.

In regional legislative bodies, citizens with a ‘non-Western’ migration background are overall underrepresented. Locally, descriptive representation in absolute numbers is relatively high in the progressive and large municipalities, but not in the rest of the country. Specific longitudinal data per municipality are lacking, but for Amsterdam, Mügge and Van Stigt (2018) show a trend very similar to the representation of MPs with a ‘non-Western’ migration background. The proportion of municipality councillors with a ‘non-Western’ migration background was increasing until the turn of millennium, then stabilized, and has been increasing again since 2018.

Although ethnically minoritized MPs are relatively well represented in parliament, this success is not reflected in the most powerful political positions, such as that of (junior) minister, party leader, or mayor (Table 29.1). At the time of writing, the position of (junior) minister has been held by only eight ethnically minoritized citizens. A Surinamese-Dutch woman, Philomena Bijlhout, appointed in 2002 for the Pim Fortuyn List, was the first junior minister. She resigned within 24 hours due to a scandal related to her support for the military dictatorship in Suriname in the 1980s. Since then, the PvdA, D66, and the People’s Party for Freedom and Democracy (VVD) have appointed cabinet members with ethnic minority backgrounds. Among them women with roots in Turkey are particularly well represented (Aydemir & Mügge, 2024).

Systematic data on openly LGBT(IQ+)⁵ politicians are rare and there are no solid data on the proportion of LGBT citizens in the population. Nevertheless, from our own calculations for the House of Representatives, parity or overrepresentation of the ‘rainbow community’ at large is rather common. This effect is intersectional: homosexual men are particularly well represented. With the start of the 2021 parliamentary cycle, ten LGBT(IQ+) MPs were inaugurated: four women (including one transgender woman) and six men. Regarding the representation of trans* people, one transgender MP would imply proportional representation. This is currently the case, as in 2021 the first transgender MP entered parliament: Lisa van Ginneken (D66).

Comparatively and historically, the Netherlands is a front-runner in the descriptive representation of LGBT citizens. In a 1976–2011 overview (Reynolds, 2013), the Netherlands had the most LGBT MPs as a proportion of the number of seats in parliament. The Netherlands also claims the first openly gay MP worldwide: Coos Huijsen (see Huijsen, 2016), first for Christian-Democratic Appeal (CDA) and later for the PvdA.⁶ Proportionally, LGBT MPs mostly run for the postmaterialist and left-wing parties. Of the right-wing parties, the VVD has a relatively strong LGBT presence. Among ministers, LGBT citizens are currently overrepresented. While systematic historical data are lacking, we observe two openly homosexual women and men (junior and senior) ministers for the CDA, including the first (Wijn in 2006; Verburg in 2007).⁷

The figures for (larger) municipalities initially suggest an overrepresentation of LGBT citizens, but since LGBT people often migrate to these bigger cities, it is hard to draw robust conclusions. For transgender citizens, this is more complex. While in both the municipal councils of Nijmegen and Utrecht one transgender person was elected in the 2018–2022 cycle, no transgender representatives have been counted since the 2022 municipal elections, countrywide.

Ten to 15% of the Dutch population has a disability, but there are very few politicians of whom it is known that they have a disability, and the Ministry of the Interior observed that very few citizens with a disability are active in politics (Ministerie van Binnenlandse Zaken en Koninkrijksrelaties, 2019). We know of six MPs who have spoken openly about their disabilities and impairments. Tara Oedayraj Singh Varma (GL, 1994–2001)—also the first ethnically minoritized woman MP—left parliament due to health issues. She announced that she had cancer, but later stated that she was dealing with mental health issues. Otwin van Dijk (PvdA, 2012–2016) served in parliament in a wheelchair. His work obliged the government in 2016 to make buildings and services accessible for citizens with a disability.

In the 2021 parliament, four MPs are known to have had a disability or impairment. Fleur Agema (Freedom Party) shared that she has multiple sclerosis and Sylvana Simons (BIJ1) that she suffers from rheumatism; Daan de Kort (VVD) is blind and Lucille Werner, a former television moderator (CDA), has had a physical disability since birth. She is widely known for Mis(s)verkiezing, a television pageant for women with a disability. These examples pertain to relatively visible and physical disabilities; the less visible disabilities are largely unknown and receive little attention. Data on the subnational level are hardly available beyond anecdotal information (e.g. in Nijmegen a council member is a former international wheelchair field hockey player; the recently appointed mayor of Leiderdorp is blind).

What explains the different patterns of descriptive representation of the groups under study? First, they are all systematically underrepresented at least at certain levels of politics. Arguably, LGBT citizens, and particularly gay men, are the exception, having reached equal to overrepresentation across the board. Some groups with a migration background have also been equally or overrepresented in some periods in parliament; Turkish-Dutch women do exceptionally well as cabinet members.

Secondly, compared to other democratic systems in western Europe and Anglo-Saxon countries, descriptive representation is relatively strong in the Netherlands. At the same time, women’s representation is falling behind. The most recent parliament marked a new era with the inauguration of the first Muslim women MPs wearing a headscarf, the first trans woman and the first Black woman party leader, while men and women MPs with disabilities are visible role models. Time will tell whether or not this intersectional diversity will increase even more.

Thirdly, over recent decades representation has increased steadily. Looking at trends, there have been jumps at different moments in time. Since the turn of the millennium, the increases in representation seem to have levelled off for women. For ethnically minoritized citizens, a new upward trend started in the late 2010s.

Fourthly, descriptive representation is mostly a postmaterialist and left-wing party affair. The Christian, conservative-liberal, and far-right parties score below average (except on LGBT representation). Recently, mainstream right-wing parties seem gradually to be including women and ethnically minoritized citizens. Fifthly, there is considerable variation across legislative levels. The national level is most diverse, while the local level lags behind with some exceptions in the larger municipalities.

Sixthly, apart from data on women there is a lack of systematic data at all legislative levels. Intersectional data are largely absent. Available data on ethnically and religious minoritized groups are still based on the outdated ‘non-Western migration background’ classification. There are no data on disability, while existing databases on LGBT MPs are not complete or include MPs who were not ‘out’ in their days as MP.

Micro-, meso-, and macro-level factors help further explain the development of descriptive representation, although these levels are also intricately intertwined.

An intersectional lens shows how little systematic information is available on intersectional descriptive representation in Dutch politics. An evident pattern is that ethnically minoritized women are better represented than ethnically minoritized men, which seems to go against the standard gender patterns. Part of the explanation is the beginning of the pipeline at the meso level: ethnically minoritized women have been supported by strong women’s networks as well as migration networks (Mügge, 2016). For LGBT MPs, an intersectional lens suggests that the gender division among LGBT MPs is rather similar to that among heterosexual and cisgendered MPs. Considering, however, that LGBT MPs mainly run for parties that list more women, LGBT representation is skewed towards gay men.

Two macro-level explanations are relevant: (1) the sociopolitical culture and (2) the political system influencing all stages of a political career and recruitment and selection. Regarding Dutch sociopolitical culture, the Netherlands is a relative front-runner in terms of valuing diversity, emancipation, and self-development (Inglehart, 1997). With the strong presence of postmaterialist and far-right parties, the cultural axis is relatively dominant in Dutch politics (see Otjes & De Lange, this volume). Since the 1980s, the underrepresentation of women has received considerable attention by the media and in societal debate. The Netherlands has had a strong LGBT movement (Holzhacker, 2012). More recently, #MeToo and Black Lives Matter campaigns kept women’s and minorities’ issues on the political agenda (Berthet & Kantola, 2020). In contrast, disability has not been politicized and the debate about decolonization has only taken off fairly recently (Hendriks, 2023). Since the turn of the millennium, the public debate along the cultural axis has focused more on ethnically minoritized citizens and Muslims (De Lange & Mügge, 2015). The presence of far-right parties has increased, which helps to explain the stagnation in average diversity.

The Dutch electoral system is relatively open and proportional (see Walter & Van Praag, this volume), which partly explains the levels of descriptive representation. The low electoral threshold, for example, facilitates the entry of new parties. Examples of these are parties founded and led by ethnically minoritized leaders, such as Denk (since 2017) and BIJ1 (since 2021). While these parties deliver descriptive representatives for citizens, the impact of their presence on perceived representation by the political elite at large remains open to question (De Jong & Mügge, 2023).

Preferential voting seems particularly beneficial for candidates with minority characteristics that are visible on the ballot—for example, through their name (Nagtzaam, 2019). Surveys show that between 6% and 11% of voters pick another candidate than the party leader because the other candidate is a woman (Jacobs et al., 2021). This stimulates parties to list minoritized candidates and women to appeal to a larger electorate alongside a majority of male candidates in eligible spots. Since 2017, preferential voting for women in elections has been actively promoted during election campaigns by a women’s organization, Vote for a Woman (Stem op een Vrouw), and following their success others established Color the Parliament (Kleur de Kamer). For other groups, such initiatives do not exist. Yet, even in the late 1970s the left-wing gay community in Amsterdam initiated the campaign ‘Vote for a Fag in the Council’. The leading LGBT interest group, COC, initiated the ‘Rainbow Voter’ in 2010, which lists ‘rainbow candidates’—that is, openly LGBT candidates—as well as straight, cisgender allies.

Another institutional way to increase descriptive representation is the implementation of quotas in politics, but legally binding quotas do not exist in the Netherlands. This partly explains why other countries from around the world have overtaken the Netherlands in international rankings. Altogether, the relatively postmaterialist political culture and institutionally open system in the Netherlands explain its relatively high levels of descriptive representation, particularly of women, LGBT citizens, and ethnically minoritized citizens.

Some parties have active diversity networks that play an important role in mobilizing and training candidates and in mobilizing party members to vote for them (Mügge & Damstra, 2013). Women, LGBT citizens, and Moroccan- and Turkish-Dutch are rather well organized and also relatively well represented descriptively. Other ethnic and postcolonial minoritized communities and disabled citizens are less so. The degree of organization within parties, but also outside parties with movements such as Black Lives Matter, seems crucial. Research on informal and formal rules within parties has shown that these have serious gendering effects. Examples of such rules are that to be selected for a party list, you need to be an active party member and already have political experience (Van Dijk, 2023). Such rules make it hard for women to be selected in the party context, even when voters prefer women (Spierings et al., 2021).

Finally, the party- or meso-level differences show how the proportionality and openness of the system can also have indirect negative effects, as it has facilitated the rise of far-right parties. These parties are generally anti-diversity and also glory in fielding White, cisgender male, heterosexual politicians. Their rise coincides with a halt in the increasing diversity in politics. Research demonstrates that parties with more restrictive positions on integration and migration nominate fewer ethnically minoritized candidates, and if they do include them, they are placed low on the list (Linders et al., 2022; Van der Zwan et al., 2019).

In sum, at the meso level, it is the relative strength of different parties, their ideology, and the success of intra-party networks, interlinked with social movements, as well as competition between left-wing and progressive parties that explains variation in descriptive representation. Additional factors are the politicization of identities, the relative size of a minoritized group and people who care about it, and institutional arrangements, such as preferential votes.

Descriptive representation is by definition a macro-level phenomenon, but in the end, it is grounded in the micro-level or individual actions of many different actors—actions that are rooted in a cultural and political context. For instance, being active in self-organizations increases the chance of being recruited by intraparty networks.

Two more specific issues—probably not unique to the Dutch case—deserve notice. First, citizens who have less internal political efficacy and who think politics is not for them are not likely to become active politically. The most recent Dutch Parliament Elections Studies show that women and people with a migration background (particularly from Africa, Asia, and Latin America) are still far less politically interested and talk less about politics, whereas the inverse holds for homo- and bisexual, compared to heterosexual, respondents (Jacobs et al., 2021; Lubbers et al., 2022). A lack of role models, who surface as a result of the presence of politicians one can identify with, and whether and how they are represented in media, might contribute to this pattern. The media ordinarily focus more on leadership traits of male than of female politicians (Aaldering & Van der Pas, 2020). In the most recent campaigns, women political leaders received systematically less attention, and the only visible ethnically minoritized male political leaders also received disproportionally little attention in media (Aaldering & Van der Pas 2021). Intersectionally, the Dutch media still paint non-norm politicians as ‘outsiders’ (Runderkamp et al., 2022).

A second individual factor that is crucial pertains to the decision to stay in or leave politics. Systematic research for the Netherlands is largely absent, but recently more attention has been given to violence against politicians. It is a wonder that MPs who deviate from the norm in many ways, such as Kauthar Bouchallikht (GL), are still standing for election given the sheer volume of hateful Tweets they receive (Saris & Ven, 2021). Worryingly, violence and aggression in politics is also rising each year (I&O Research et al., 2020).

Based on our review of descriptive representation, we outline three avenues for further research. First, the data gap needs to be addressed. Most knowledge and data are available at the national level and about women. Data at subnational levels, about other groups, and on intersections should be collected systematically too. The collection of data at the subnational level is crucial, given that provincial and municipal councils and executives show lower levels of descriptive representation than national-level politics. This discrepancy is worrisome because subnational politics is often a stepping stone to the national parliament. Moreover, and in light of Pitkin’s ‘standing for’ logic, an important avenue for data collection would be the inclusion of questions about self-identification in surveys (Celis & Mügge, 2018). Secondly, relatively little is known about the retention of ‘descriptive representatives’ of different groups. Scholars, parties, and policymakers alike focus on the influx of elected officials, but to understand underrepresentation we need to ask why elected officials belonging to marginalized groups in politics drop out or have shorter political careers. Data are needed on incumbency, running for multiple terms, and re-election. This also requires data and research on specific groups. For instance, a handful of reports and studies point to structural sexism and violence against women in politics (Krook, 2020; Mügge & Runderkamp, 2023), which might be a reason why they are underrepresented.

Lastly, at both the political and policy level, descriptive representation is currently a battlefield. There is more positive attention to diversity and ways to increase descriptive representation as part of a strongly functioning liberal democracy. However, counter-movements, backlash, and attacks by the far right, as well as religious groups, are growing, particularly in the form of anti-feminism, anti-Muslim, and anti-LBGTQI. While we surmised that the growth of the far right is hurting descriptive representation of several marginalized groups directly due to the candidates listed, it remains a question how descriptive representation as a battlefield, inside and outside of representative bodies, will influence actual descriptive representation in the longer term.