Imagined constituents: Minoritized citizens’ evaluations of political representatives in Germany and the Netherlands

ABSTRACT

The presence of minoritized representatives is considered essential to advance the interests and inclusion of minoritized groups. But do minoritized citizens prefer politicians like them? Drawing on nineteen focus groups (N = 98) with minoritized Dutch and German citizens, I ask how citizens evaluate (descriptive) representatives. Citizens see descriptive representation as intersectional. They value politicians – regardless of descriptive status – who imagine their lived reality, politically judge, and act on that basis, and enhance emancipation and group legitimacy. Many see descriptive representatives as better able to do so. Yet, in practice, citizens examine if descriptive politicians share their political judgements and act for them within established politics. The findings suggest that increasing numbers is necessary but not sufficient. Whether citizens perceive representation depends on the space descriptive representatives have within the political system to articulate marginalized identities and concerns. Citizens’ evaluations reflect differences in self-identification, privilege/disadvantage, group histories, electoral systems, and representation legacies.

KEYWORDS:

• Political representation
• political equality
• identity
• ethnic/racial diversity
• the Netherlands
• Germany

Introduction

Western European parliaments have seen a slow increase in the number of racially minoritized politicians over the last decades. Yet, underrepresentation persists, raising concerns about whether politics truly represents the concerns of all citizens equally. Particularly for historically excluded groups, the presence of descriptive representatives – politicians who share their backgrounds (Pitkin 1967) – is considered key to be and feel represented. Minoritized politicians are thought to advance group concerns (Mansbridge 1999). Also, their presence signals inclusion and equal political standing (Phillips 1995). But, it is not self-evident that citizens and politicians who share social backgrounds identify in the same way, or favor the same politics (Lee 2008). Do minoritized citizens in Western European countries believe that minoritized politicians stand and act for them, and do they experience better representation through their presence?

Representation scholarship mainly studies minoritized representatives (Celis and Childs 2020). Electoral research shows that citizens vote for politicians who share their background (Bergh and Bjørklund 2011). This suggests that citizens prefer descriptive representatives. But other studies nuance this picture. It is not sharing the same demographic label that matters, but rather whether citizens self-identify with politicians (Van Oosten, Mügge, and Van der Pas 2024). Citizens do not look at politicians’ race/ethnicity alone, but its interaction with gender and other identities (Montoya et al. 2022). Additionally, evaluations differ between groups, and support for descriptive representation does not always translate to a vote for minoritized political candidates (Street and Schönwälder 2021; Spierings and Vermeulen 2023). Qualitative studies provide insight into minoritized citizens’ perspectives on representation (Akachar 2018; Bird 2015; De Jong and Mügge 2023; 2024). However, studies cover single groups or countries, which makes it difficult to disentangle general and context-specific views.

This study draws on nineteen focus groups (N = 98) and centers on the insights of minoritized constituents. I explore who citizens see as descriptive representatives. I contrast this to common measurements of descriptive representation: shared demographic between politicians and citizens. To explore how minoritized citizens evaluate politicians, and for whom, when and why descriptive representation matters, I combine classic political representation scholarship (Pitkin 1967), Arendt’s (1978) work on political judgement, and intersectionality theory (Crenshaw 1991). The study spans two political contexts: Germany and the Netherlands. It includes Turkish-, Resettler-, and African-Germans; and Turkish-, Moroccan-, and Surinamese-Dutch.

The findings problematize using single demographic labels as a proxy for descriptive representation. Most minoritized citizens see descriptive representation as intersectional and gradual. Political actors are more or less “like them” depending on shared experiences of disadvantage, like discrimination. These experiences pertain to specific political issues or contexts. A minority of participants, however, did not believe that the politicians they are associated with “stand for” them in politics. Findings underline the need to consider minoritized citizens’ self-identifications to accurately establish who descriptive representatives are. Combining Pitkin’s (1967) and Arendt’s (1978) concepts moves beyond the descriptive/non-descriptive binary and helps to articulate the steps citizens take to determine whether politicians represent them. Citizens positively evaluate politicians who imagine their lifeworld, and politically judge and act on that basis. Most see descriptive politicians as better able to do so. But, they critically examine their judgements and actions within established political parties. Nevertheless, most value descriptive representatives because they enhance emancipation and legitimacy. The intersectional approach reveals that group privilege, historical experiences, and ideological divisions feature in citizens’ views. Perspectives are similar across countries, though focus groups reflect context-specifics.

The next section outlines my conceptual framework. After, I discuss the case selection and methods. The results detail who minoritized citizens see as descriptive representatives, and how descriptive representation relates to interest representation and to symbolic inclusion.

Studying how minoritized citizens evaluate representatives

A longstanding question in representation scholarship is whether politicians from historically excluded groups champion group interests. At the heart of this question is the concept of “descriptive representation”. For Pitkin (1967), descriptive representation means shared social backgrounds between politicians and constituents, like region. Yet feminist political theorists (Phillips 1995) move away from essentialist notions of resemblance based on characteristics alone. They argue that “shared experience […] lies at the core of descriptive representation” (Mansbridge 1999, 629). Empirical studies, however, rarely measure a sense of shared experiences between politicians and citizens. Researchers often establish descriptive representation based on predefined ethnic/racial demographic labels (Celis and Mügge 2018), like “immigrant background” in Europe.

Shared demographic labels may insufficiently capture descriptive representation. Citizens and politicians who share demographic labels do not necessarily share political identifications or goals (Lee 2008). Additionally, identities are heterogeneous (see Walter and Glas 2024). Intersectional scholarship (Crenshaw 1991) emphasizes that multiple, interacting social categories, like gender, ethnicity/race, and social class, inform people’s identifications and experiences in context-specific ways. Intersectionality complicates the idea that politicians with immigrant backgrounds “stand for” citizens with immigrant backgrounds in every respect. Politicians may resemble citizens of immigrant background, but differ in gender, or other identities. Moreover, Pitkin's work (1967) assumes stable and transparent groups and interests that politicians simply represent. But constructivist scholars (Saward 2010) emphasize that groups are not objective or fixed. In real-world politics, representatives claim to speak or act for groups, like “immigrant-origin citizens”, constructing group identities in this process. Representatives need not be elected. Activists or advocacy groups also make representative claims.

Considering critiques of predefined categories to measure descriptive representation, constructivist scholars (Saward 2010) promote shifting the focus to constituents’ perspectives. Citizens should judge how they are represented. Celis and Childs (2020, 63) suggest that: “who counts as a descriptive representative is not read off from ‘known’ or ‘objectively observed’ ascriptive identities” but should be studied empirically, requiring “recognition on behalf of the represented”. Taking up this call, I explore how minoritized citizens perceive descriptive representation. How do citizens with immigrant backgrounds self-identify, and who resembles them in politics? Empirically, I examine how citizens construct identities and discuss who descriptive politicians are in group settings. Individuals rarely cover diverse group experiences (Weldon 2011). By discussing instances of sexual harassment, for instance, women discover shared, but also diverging experiences based on ethnicity/race or class.

To study how minoritized citizens evaluate (descriptive) representatives, I depart from classic theories. For Pitkin (1967), resemblance ultimately did not matter. Citizens should judge politicians based on whether they advance their interests: substantive representation. Regarding how politicians should act, Pitkin (1967) distinguishes between “trustees” and “delegates”. Trustees follow their own judgement when taking political decisions, while delegates follow constituents’ wishes. Later theorists stress the importance of descriptive representation for historically excluded groups (Phillips 1995). They argue that who politicians are (descriptive) informs their actions (substantive). Besides what minoritized politicians do, their presence carries symbolic weight (Runderkamp et al. 2022). Visibility signals that members of minoritized groups belong and are “able to rule” (Mansbridge 1999). I refer to representatives who positively change the social meaning associated with minoritized groups as “symbolic representation” (Lombardo and Meier 2014). Symbolic representation happens through presence, but also when representatives adopt inclusive language or symbols. Symbolic representation has received limited attention (Lombardo and Meier 2014), but could matter to minoritized citizens whose identities are often invisible or stigmatized.

I complement Pitkin’s framework with concepts developed by Hannah Arendt (1978).¹ Arendt is not preoccupied with political representation per se. She writes about mental activities, like thinking and judging. Yet for Arendt, judging is intimately connected to political action, rather than a contemplative activity (Steinberger 1990).² Together with Pitkin, Arendt’s concepts help to articulate how participants in this study evaluate representatives.

Based on Arendt’s work, I break down how representatives act – substantively, symbolically, or otherwise – into three steps. First, Arendt talks about “political imagination”. This means that political actors should put themselves in others’ shoes. They should imagine citizens’ differing “standpoints”, which means “the conditions they are subject to, always different from one person to the next, one class or group to another” (Arendt 1978, 258–265). Thus, politicians need thorough knowledge of the lived conditions of different groups. Second, by imagining others’ conditions, and reflecting on their possible views, political actors interpret and judge a political issue (Arendt 1978, 258). To judge does not imply simply adopting the political views of others. This would amount to passively exchanging “their prejudices for the prejudices proper to my own station” (Arendt 1978, 258). Instead, political actors should consider and weigh the positions of different groups to form a judgment. This holds a middle ground between the trustee-delegate distinction. Politicians judge and act independently while relying on a thorough understanding of how differently positioned constituents would be affected by, and judge a political issue. Third, representatives act on their judgements, substantively, symbolically, or otherwise. Judgement and action are distinct. Politicians often face constraints which prevent them from acting on their judgements (Arendt 1978, 260–261). In such a situation, politicians should not renounce their original judgment, even if they cannot act on it.³

Arendt’s (1978) concepts complement Pitkin’s (1967) work in several ways. To start, Arendt’s “imagination, judgement, and action” breaks down substantive or symbolic “acting for” in three steps. Rather than perceiving representation in binary terms (being represented or not), this helps to identify how citizens experience representation on some aspects and not on others. Minoritized citizens could experience some substantive representation by politicians who understand their lived reality. But they may disagree with their judgements or actions. Another advantage is that, “political imagination” describes a quality, rather than a static descriptive/non-descriptive identity. Both descriptive and non-descriptive representatives can be more or less able to imagine citizens’ lived condition, judge, and act. Political imagination also fits well with constructivist thinking. Here, constituents do not belong to stable groups with objective interests that representatives simply know or “read off” (Saward 2010). Yet based on real-world experiences, representatives imagine politically relevant groups and their interests. This aligns with representation as a “creative process” in which representatives make “allegations about identity and interest” (Saward 2010, 168).

Applying Arendt’s terminology, descriptive representatives, who have experiential knowledge, can be thought better able to imagine minoritized citizens’ realities. This could matter in situations of “uncrystallized interests:” new political issues on which politicians did not previously state their position (Mansbridge 1999, 644). Intersectional and ideological differences likely matter. Minoritized politicians from elite milieus may be less able to imagine the lifeworld of poor minoritized citizens. For this reason, Dovi (2002) argued that descriptive representatives who have strong ties with disadvantaged subgroups are preferable. Minoritized politicians and citizens may disagree over political judgements and actions (Weldon 2011). Therefore, some scholars have shifted the investigation to which “critical actors” represent minoritized group interest, regardless of their identity (Celis and Erzeel 2015). From this perspective, men or majority politicians can also represent marginalized groups. Dominant group members could acquire political imagination through reading, conversation and living with minority group members (Mansbridge 1999, 635). Ideals like ethnic/racial justice could motivate them to represent minoritized groups (Harder 2023).

The few studies that investigate minoritized citizens’ perspectives on representation show that country-specifics matter (Bird 2015; Schildkraut 2015). I compare several immigrant-origin groups across two countries to investigate similarities and context-specifics in citizens’ views. I particularly pay attention to immigrant histories, integration regimes, electoral systems, and legacies of minoritized politicians.

Immigrant histories, including postcolonial or labor, can affect citizens’ relationship to politics (Bird, Saalfeld, and Wüst 2010). Distrust stemming from injustice may lead citizens to prefer descriptive representation (Mansbridge 1999). Moreover, integration regimes historically differ. Citizens could deem descriptive representation more feasible in countries that accommodate ethnic/racial group politics, compared to regimes that discourage minority politics (Ersanilli and Koopmans 2011). Next to migration and integration, political institutions and party landscapes affect representation (Norris and Inglehart 2001). Party lists tend to be more diverse in proportional electoral systems, depending on the rules (Bloemraad and Schönwälder 2013). Parties signal to politicians how they should act. In established parties, minoritized politicians do not always advocate for minority rights (Aydemir and Vliegenthart 2016). Legacies and past experiences with minoritized politicians can affect future expectations (Bird 2015).

Focus groups in Germany and the Netherlands

The countries selected are Germany and the Netherlands. The country-comparison is not the primary research goal. By including two countries, I aimed to see if there is something shared in how minoritized citizens view representatives. I treat Germany and the Netherlands as a “conceptual spectrum of cases” (Bloemraad 2013, 40). They vary on key characteristics: integration regimes, electoral systems, party landscapes, and representation legacies. Accordingly, I investigate whether citizens’ views reflect country-specifics.

Immigrants to Germany had historically difficult access to citizenship. Integration policies discouraged group-based politics (Ersanilli and Koopmans 2011). Dutch policies granted easier access to citizenship and accommodated diversity (Ersanilli and Koopmans 2011). Both countries have proportional representation systems. Germany has a mixed member system. Germans cast two votes, one in constituencies with a first-past-the-post system and the other for closed state-party lists. The latter ensures proportionality overall, but a 5 per cent threshold means that comparatively few, established parties make it to Parliament. The Netherlands has one of the world’s most open political systems. It has a single constituency with a low electoral threshold and preference votes. The open system makes it easier for new parties to gain a foothold. In 2021, there were twenty-one political parties, including two anti-racist “minority” parties: DENK and BIJ1. German minoritized politicians entered the Bundestag late and numbers slowly increased (Donovan 2007). The Dutch parliament has exceptionally high minority representation. The percentage of minoritized MPs has matched the percentage in the wider population since 2003 (Mügge, Van der Pas, and Van de Wardt 2019).

To explore who minoritized citizens perceive resemble and represent them, I conducted focus groups. Focus groups reduce researcher control. They give space for participants to introduce their agendas and concerns (Morgan 1996). Participants compare and contrast their views, exposing possible differences in how citizens identify and evaluate representatives. Focus groups also reveal country-context specificities in discussions (Van Bezouw et al. 2019).

I invited participants based on ascribed demographic categories, to contrast this to how participants self-identify. In both countries, the official ethnic/racial category is “immigrant background”. Citizens have immigrant backgrounds if they or their parents migrated to the Netherlands or Germany. In practice, immigrant background is often used as a proxy for ethnicity/race (Mügge, Van der Pas, and Van de Wardt 2019). Germany distinguishes between European and non-European immigrant backgrounds and, until recently, the Netherlands, between Western and non-Western.

The selection includes Turkish, Moroccan, and Surinamese-Dutch, and Turkish, African- and Resettler-Germans (Spätaussiedler). Citizens in these categories face racism based on roots, skin color and/or religion (FRA 2017; Goerres, Mayer, and Spies 2020). The selection covers various immigrant trajectories: mostly labor (Turkish/Moroccan-Dutch; Turkish-Germans) and postcolonial immigration (Surinamese-Dutch). Resettler-Germans, a relatively privileged category, migrated to Germany from former socialist countries and had easy access to citizenship (Donovan 2007). African-German is a category that remains invisible in statistics, consisting of German citizens who self-identify as Black, African, or Afrodiasporic (Aikins et al. 2021).

I follow McCall's (2005) intersectional approach. It departs from pre-existing social categories to explore how privilege and disadvantage operate “within” and “between” groups. In the same immigrant background focus groups, I investigate how individual views differ by social class, gender, age, and so on. I further investigate variation between immigrant background focus groups. For each immigration category, I conducted one women-only group. Theoretical sampling fits with common focus group practice. Establishing homogeneity on key demographics facilitates conversations and permits systematic comparisons between groups (Morgan 1996). Although this selection could make ethnic/racial identifications more salient, it also allowed participants to articulate if and how immigrant background mattered.

Data collection took place between 2020 and 2021. In both countries, I collaborated with research assistants who helped recruit participants and moderate focus groups. I am a white woman; the research assistants have various ethnic and gender identities. How researchers identify and are seen affect data collection and interpretation (Chiu and Knight 1999). The whole team together reflected on findings and researcher positionalities. Our impression was that participants saw majority team members as outsiders. Minoritized members were often seen as insiders, though experiences varied (Appendix 4). Some African-German participants indicated that a white researcher's presence may have made them downplay their views. Though other participants made no such statements, we cannot exclude this possibility.

Participant recruitment included distributing flyers, drawing on our networks, social media, and snowballing. To ensure that participants were at ease, we informed everyone about the conversation topic and focus group make-up. Participants filled out an informed consent form⁴ and survey (Appendix 2). We aimed for mixed age, class, and gender groups to facilitate an intersectional analysis. But, we weighed diversity against power differences in the focus groups, as social positions affect who dominates or speaks less. Moderators discussed conversation ground rules and encouraged everyone to contribute. We organized conversations on Zoom due to the Covid-19 pandemic. Meetings lasted between 1.5 and 2.5 hours and were videorecorded.

The topic list was designed to not prime participants with ideas from the literature (Appendix 1). Moderators followed a non-directive style (Morgan 1996) and conversations started with open-ended questions. Moderators asked participants what (not) feeling politically represented meant to them. Next, participants discussed instances when they felt represented or not. The moderators followed up as much as possible. In a second, semi-structured part, moderators zoomed in on descriptive and substantive representation.

We conducted nineteen focus groups with ninety-eight participants. There were four each with Moroccan, Turkish- and Surinamese-Dutch; three each with Turkish- and Resettler-Germans, one with African-Germans. One group in each category was women-only (Appendix 3). We held an extra, in-person focus group with first-generation Moroccan-Dutch women. It took place in a neighborhood center. The sample has some limitations. University- and applied sciences educated and politically engaged participants are overrepresented. Descriptive representation may matter less to privileged citizens. But, education can also increase awareness of exclusion (Geurts, Davids, and Spierings 2021). This could lead participants to prioritize descriptive representation. Most have left/progressive political orientations, which may influence results as right-wing voters may support descriptive representation less (Street and Schönwälder 2021)

Focus groups were transcribed and analyzed in Atlas.ti. I indexed all documents by descriptive characteristics, topics discussed, political actors, theoretical concepts, and political concerns. I then analyzed segments about descriptive representation and representatives in-depth. The analytical codes were based on the research questions. I coded the identifications that participants mention, why participants discuss someone as resembling them, and whether the representatives mentioned shared participants’ immigrant background. I then analyzed how participants evaluated political representatives in relation to their descriptive status. Finally, I systematically compared codes within- and between focus groups and between countries. Pseudonyms are used to guarantee confidentiality.

Descriptive representation according to citizens

Most participants extensively discuss politicians who resemble them before moderators introduce the topic. This sometimes made the semi-structured questions redundant. Some state that they have already reflected on this issue or discussed it elsewhere. Participants describe politicians as resembling them if they share lived experiences of disadvantage. To most, shared experiences of racialization are key, based on shared immigrant background, skin color or religion: “It’s very important that someone looks like me, because race still plays a role” [Sandra, Surinamese-Dutch_FG9]. Moroccan-Dutch, and some Turkish-Dutch/Germans mention Muslim representatives.

Nadira:

Few Arab people have a seat in parliament. So, nobody passes on what the problems are of Muslims here in the Netherlands and for Muslims women, العنصرية [racism] […]

Chourouk:

Doesn’t necessarily need [to be] Moroccan [politicians] […] Just Muslim. Doesn’t matter where he comes from. [Moroccan-Dutch_FG8]

Participants did not want to be seen as “different” or as a “minority” in principle. Yet many view descriptive representation as a self-evident concept because they see political representation as unequal. Their concerns and identities are not equally visible and represented, compared to majority citizens. This dynamic is well illustrated in Surinamese-Dutch conversation_FG11. Ishaan, an interviewee, describes politicians who resemble her:

Ishaan:

Someone who […] belongs to a minority group [Samantha frowns].

Rachelle:

I understand what you are saying [Samantha and Rachelle laugh].

Ishaan:

Who are a minority in being represented [laughs].

Samantha:

Yes, in the minority represented I would say.

Ishaan:

Yes. […] An underrepresented party, let me put it this way.

Most participants saw “politicians who resemble you” as an intuitive concept, stemming from shared experiences of disadvantage. However, descriptive representation was less self-evident for a small subset of participants: mostly Resettler-Germans, and, though few in the sample, interviewees with right-wing views. These interviewees did not perceive themselves as minoritized, nor always saw their identities or interests as shared. Some disagree that politicians with whom they are associated stand for them.

Alexander:

I get the idea that they built a cage for me as a person with a migration background and they try to position me exactly in this cage. […] I don’t want to be represented by some people who only speak my language [Russian]. That’s not an [important] issue at all, being represented based on a group [membership]. [Resettler-German_FG18]

Alexander’s statement illustrates that how groups are politically constructed (i.e. Saward 2010) does not always correspond to how citizens want to be identified in politics. Later, when asked about politicians who resemble him, Alexander pushes back against the question: “I have a feeling I’m talking to [moderators] now. I get the impression that […] they want to help us. With all the questions, how can they represent us well in politics.” Other participants agree. Olga: “you just get lumped into a statistic […]: “people with an immigrant background. […] I don’t want to feel like a minority”. Yet, some hold that Resettler-Germans are disadvantaged and face common obstacles, like having foreign diploma’s recognized in Germany. They identify with Resettler-German politicians. Thus, views among Resettler-Germans differ. For some, group identification, stemming from perceived disadvantage, mattered in politics (cf. Goerres, Mayer, and Spies 2020). For others this was not the case. Relative privilege in legal and ethnic/racial visible minority status may make it easier for these participants to “pass” and identify as part of the dominant majority, rendering group-based descriptive representation irrelevant.

Intersectional- and country-variation

Participants who identify as a minority refer to minoritized politicians with various immigrant backgrounds as descriptive representatives. Michelle [Surinamese-Dutch_FG10] mentions Moroccan-Dutch politicians: “You may not look like me, but you […] know what it’s like to be colored in the Netherlands.” Participants qualify a political actor as more or less descriptive depending on the politicized experiences they share along intersecting axes of ethnicity/race, religion, gender, age, and social class: “If someone walked in [parliament] who is not Muslim but grew up in the same economic environment as me, that would already help.” [Emir, Moroccan-Dutch_FG6] Pelin [Turkish-German_FG14] explains who represents her:

Aminata Touré [Senegalese-German politician, Greens] comes to mind […] even if she is a Black woman and I am a Woman of Color – I’m not Black – I think we already have a big overlap of experiences and I think the content or the opinion that she represents also partly comes from her experience as a Black woman in Germany. […] She’s very young, I think in her late 20’s […].

To determine whether Touré resembles her, Pelin mentions that shared experiences are key, relating to intersections of race, gender, and age, rather than immigrant background. Who participants mention as a descriptive politician also depended on the political issue. Pelin later mentions Helin Evrim Sommer, a Kurdish-German politician [Die Linke]: “She also deals with […] the situation of Kurds in Turkey. As a Kurd I feel seen there.” Likewise, Turkish-German women mention Angela Merkel when discussing gender inequality. Later, they relate Turkish-German descriptive representatives to social class and ethnic/racial inequalities. Men participants do not discuss majority men as descriptive representatives, indicating that (intersecting) social markers matter when they result in shared experiences of disadvantage. Findings are consistent in the German and Dutch groups. How states historically accommodated diversity – the Netherlands more than Germany – was not visible in how minoritized citizens identify or discuss descriptive representation.

Most participants found “descriptive representation” intuitive. They discuss descriptive representatives as those with whom they share experiences, stemming from minorization. Such experiences coincided with, but also departed from ascribed “immigrant background” demographic labels. Among participants with a less visible minority background like Resettler-Germans, the idea of descriptive representation was more controversial. Not everyone identified as part of a minoritized group. Some found descriptive representation irrelevant, or a misrepresenting construct imposed on them by policymakers or researchers. Thus, how participants discuss descriptive representation seemed linked to their political self-identification, and related visible minority status. The intersectional analysis reveals that participants see descriptive representation as a gradual concept. Politicians resemble them more or less, depending on politicians’ proximity to their intersectional experiences of disadvantage. This varies depending on the political issue at stake. Findings are similar in both countries.

Citizens’ evaluations of (descriptive) representatives

Participants evaluate representatives in two main ways. They see if representatives (1) speak and act on their political concerns, and/or (2) positively change dominant societal images of their group. These conditions resemble substantive and symbolic representation (Lombardo and Meier 2014; Pitkin 1967). For most, substantive representation weighs more heavily than symbolic representation. Sara [Moroccan-Dutch_FG5] says: “It matters most what the person stands for. But […] it is important that minorities can recognize themselves in politics.” Similarly, Angela [Surinamese-Dutch_FG12] states: “My point of view trumps [a politician’s] color, but it will count […] representation matters.” How do participants evaluate representatives? And how does a descriptive status – understood as shared experiences of disadvantage⁵ – matter? I discuss substantive and symbolic representation in turn, considering intersectional- and country variation.

Substantive representation

Across focus groups, participants positively evaluate politicians who speak or act on their interests: substantive representation. Participants refer to values, ideals, and political concerns that affect them. These often relate to political identifications. To evaluate whether politicians represent their interests, participants look at three successive aspects, resembling political imagination, judgement, and action (Arendt 1978).

First, participants consider whether politicians can imagine their lifeworld and related concerns, political imagination. Good representatives “can put themselves in our position” and “think from my perspective” [Ishaan, Surinamese-Dutch_FG11]. There is a general consensus that descriptive politicians can better imagine participants’ reality because of their experiential knowledge: “I think that more politicians should be like me because I think they can better understand what my interests are, so they can be more empathetic and represent [me] better” [Romana, Surinamese-Dutch_FG10]. Others also think that experiential knowledge is crucial to assess the political imagination politicians have.

Nelson:

[…] my first impulse for me to feel represented, that people who understand the reality of my life, who understand how I feel, what I’ve experienced and can understand my perspective, that these people represent me […].

Faiza:

That they also speak from a perspective that is similar to mine, that is, not only the interests. I mean a white person may be able to represent my interests but may not be able to bring up the hundred percent empathy and knowledge of what Black people feel or need in certain situations. [African-German_FG19]

Like Faiza, several participants refer to a politician imagining feelings and needs in ad-hoc moments. This recalls Mansbridge’s (1999) argument for descriptive representation in contexts of “uncrystallized interests”.

Participants often discuss non-descriptive politicians as out of touch with their lived reality. Nonetheless, interviewees mention examples of non-descriptive MPs who acquire political imagination indirectly. They make up for a lack of experience through empathy, listening, and staying connected to minoritized communities. In Kaan’s [Turkish-German_FG14] words: “Someone who has experienced and lived through certain things can of course represent it better, but you can still, I think, adopt [aneignen] certain things […].” Attendees in other focus groups shared this view. Tom [Resettler-German_FG16] communicated that “it’s very, very important for me to notice that [politicians consider] what are different living environments and situations like? Then the person doesn’t have to embody all the factors themselves”. Participants describe several majority or men MPs as “critical actors” (Celis and Erzeel 2015).

Second, participants evaluate whether politicians draw on their political imagination in forming political judgements. Most hope that descriptive politicians will judge political issues similarly to themselves, but critically examine whether this is the case.

Özlem:

A person who has been racially profiled himself, or who has experienced it firsthand, definitely takes a more critical position on the whole issue […] So, from there I think that you can assume certain positions or a certain wealth of experience behind visual representation.

Marla:

I agree with that. But it’s not crucial, actually, because I think that it also demands a certain reflection. Even if you’ve been through it yourself, it doesn’t mean at the same time that you have empathy. Then again, of course, there are people who don’t go through it themselves, but who engage with people and listen to other people’s perspectives and then take those into account in their decision-making processes.

Özlem:

[…] having experienced it firsthand doesn’t necessarily mean that you’re going to be able to classify it properly, I think. Yes, that is definitely the case when you look at – I don’t know – people from Turkey in the AfD [Alternative for Germany, far right party]. [Turkish-German_FG13]

In Özlem’ reasoning, descriptive politicians experience racial profiling firsthand. They are better able to imagine and independently judge this issue. However, this is not always true, as in the AfD example. Additionally, non-descriptive politicians sometimes judge political issues in line with citizens’ wishes.

In multiple groups, participants argue that politicians should not let their judgment be determined solely by one specific group. In line with Arendt's (1978) theory, politicians should consider the concerns of multiple groups and connect them to arrive at a more general judgment.

Imane:

You [a representative] start looking, trying to abstract, to see if there's not in that self-interest for my group, if there's something of a general interest in there. And trying to look for that and make that clear as well. […] that those [interests] ultimately serve the common good. [Moroccan-Dutch_FG7]

If participants see politicians as able to imagine and judge, a third consideration is whether they can speak or act for them. A general concern was that descriptive politicians have little leeway. They cannot always represent minority concerns within majority-dominated parties.

Merve:

“Cem Özdemir or Canan Bayram [Turkish-German politicians, Greens] […] I can’t say […] what they are fighting for […] in the end it is still the political agenda [of their party] that is fulfilled.” [Turkish-German_FG13]

Benny:

Representation is always substandard and often it’s kind of window dressing […] if you look at the people of color that are represented in the political parties. [Surinamese-Dutch_FG9]

Participants are disappointed with descriptive politicians they considered assimilated or tokenized. Others, however, blame the system. They remark that descriptive politicians face contradictory expectations from their party and their community. Again others think that descriptive politicians can make a difference in established parties. Though opinions differ, many argue that non-descriptive politicians can and should advocate for them; some think that non-descriptive politicians face less backlash when doing so.

In evaluating politicians’ actions, most refer to visible political work like statements, speeches, and sometimes voting in parliament. Some find it hard to assess if and how (descriptive) politicians acted on their concerns in policymaking: “as people who […] don’t sit in the Bundestag, we don’t understand what kind of restrictions you as an individual [politician] have” [Pelin, Turkish-German_FG14]. Many appreciate politicians who show what they stand for, even if they cannot act on their views. This aligns with Arendt's (1978) view that, given constraining circumstances, politicians should not abandon their initial judgements.

Intersectional and country-variation

As discussed earlier, few participants saw descriptive representation as irrelevant. These participants also disagreed that the politicians they are associated with represent them better. The other participants differ in how they evaluate the ability of descriptive versus non-descriptive representatives to imagine, judge, and act for them.

African-German, Surinamese-Dutch and Muslim participants across focus groups discuss descriptive representation more. Many see descriptive representatives as better able to imagine their lifeworld compared to non-descriptive politicians. Selena [Surinamese-Dutch_FG12] expresses this view, when discussing the abolition of Dutch slavery.

[…] not descendants of enslaved people, but plantation owners themselves, were going to negotiate how they were going to abolish slavery, with all the consequences. So, it really matters whether you can put yourself in another person's shoes.

This statement resonates with Mansbridge’s (1999) argument for descriptive representation in contexts of intergroup mistrust. Muslim participants likewise link descriptive representatives who imagine their condition to a hostile anti-Muslim political climate: “Since those tensions have risen […] I think it's important that I recognize myself in the person who is going to represent me” [Omar, Moroccan-Dutch_FG6].

In contrast, some Turkish-Dutch/German and Resettler-German participants noted that politicians with similar immigrant backgrounds do not share their judgements due to political differences in their community. Derviş [Turkish-Dutch_FG3] discusses Turkish-Dutch politicians who he considers extreme. He states that “someone who comes from the same country, or the same culture does not share the same ideology by definition”.

In both countries, participants mostly discuss (descriptive) representatives in their national parliaments. They also refer to prominent politicians, like ministers and mayors. In fewer cases, Germans mention candidates in their local constituencies. They did not evaluate the (descriptive) politicians elected directly in their constituencies or those on party lists differently. Both Dutch and Germans critically discuss representation legacies. But Dutch draw on generational experiences to illustrate that descriptive representatives do not necessarily share their judgements. First-generations voted for descriptive politicians, assuming they shared their views. In the end, some were disappointed.

Hatice:

I know several politicians who were elected solely because of their Turkish name [Pınar and Sarah nod], and afterwards some of the Turks regretted it [Cemre nods] because they hadn't properly researched what they stood for. […] maybe they thought [they] would stand up for Turkey, but that very person proposed a bill against Turkey every time. […] I really try to avoid that now. I mean now we're twenty years down the road [laughs]. I don't look at Turkish names anymore. [Turkish-Dutch_FG4]

Both first- and second-generation Dutch participants discuss how they critically scrutinize whether minoritized politicians share their judgements. Similar conversations about voting for Turkish names are absent in the German groups. Germans also mention relatively fewer minoritized politicians and discuss their presence and actions in less detail. Possibly, because of historically differential access to voting rights, and because minoritized politicians entered the Bundestag relatively late.

Turkish-Dutch participants reflect that, despite historically high numbers of minoritized politicians, they do not feel substantively represented. When Aysun (FG1) indicates feeling underrepresented, Günay disagrees. He argues that there are quite a few Turkish-Dutch politicians. In his view, feeling underrepresented: “might have to do with the fact that people don't share your ideals […] I think a lot of people […] don't count them, because they are too far from their own political views”. Senna adds: “I think it's the feeling of, okay they [Turkish-Dutch politicians] are trying to fit in.” Many Dutch participants see descriptive politicians within “minority parties” as more autonomous. Dutch participants distinguish between feeling represented by politicians affiliated with established parties versus minority parties.

Orlando:

If you look at an Aboutaleb and a Marcouch [Labor Party, Rotterdam and Arnhem majors, Moroccan backgrounds] not. But if I look at Farid Azarkan and Tunahan Kuzu [minority party DENK politicians, Moroccan and Turkish backgrounds], yes. […] Aboutaleb has sold his soul so to speak and is not doing what he really should be doing. […]

Michelle:

Is that what Sam's saying that he's going to kind line up with the rest of the politics? […]

Sam:

I think he sold his soul.

Orlando:

Yes exactly. [Surinamese-Dutch_FG10]

Participants positively evaluate politicians who substantively represent them. They see if representatives politically imagine their lived realities, judge, and act. Most think that descriptive politicians are better able to meet these criteria. But in practice, descriptive politicians may interpret their experiences differently. They may not share judgements and could be constrained to act in established political parties. Participants often see non-descriptive politicians as far removed from their reality, though there are some positive examples. Majority politicians can take on a responsive and empathetic role. Expectations differ. Surinamese-Dutch, African-German, and Muslim citizens emphasize descriptive representation more, narrating past and present injustice. Turkish- and Resettler-origin citizens refer to community ideological divisions to explain skepticism toward descriptive representation. High numbers of minority politicians in the Netherlands do not eliminate perceived underrepresentation. But many Dutch participants – who see a range of parties in the open political landscape – hold that descriptive politicians in minority parties have more autonomy to voice their concerns.

Symbolic representation

Although participants primarily look at interest representation, symbolic representation mattered too. Both German and Dutch participants talk about representatives in and outside of parliaments who positively change how their group is perceived. Participants assess the symbolic impact representatives have on majority citizens and on their community, which I discuss in turn.

Across groups and countries, participants positively evaluate mostly descriptive political actors who change negative images about their group, like unintegrated, unqualified, or oppressed, among the dominant majority. Turkish-Germans discuss Özlem Türeci and Uğur Şahin, the co-founders of BioNTech that created the first Covid-19 vaccine.

The BioNTech founders, I think they are the people who truly represent us. It's not these negative headlines: The German-Turk, who was again involved in some fights, but it is the German-Turk, who has done something for Germany […] That's what we are. We grew up here, we live here, but we do something for Germany, and we are a part of Germany. [Ayla, Turkish-German_FG13]

Participants note how Türeci and Şahin made their group “stand for” something positive, an integral part of, and contributing to, Germany. The presence of descriptive representatives alone changes negative stereotypes associated with their group among the dominant majority: “I think that if there are more women in the House of Representatives, that there will be a shift in the conceptions of certain people. She can show that the prejudices are not true” [Nazlı, Turkish-German_FG15]. This fits with Mansbridge’s (1999) argument that symbolic representation particularly matters because it changes perceptions of the dominant group. Participants mention few similar examples of non-descriptive representatives. Canan [Turkish-German_FG15] recounts a white Austrian politician:

This politician gave a speech wearing a headscarf, [she] covered herself in front of everyone and showed solidarity and proved that I can also empathize with this population […] and also stand up for them, even though […] I think she herself is a native Austrian and may never have experienced racism […]

The Austrian politician draws attention to sexism and racism against Muslim women and reconstructs the headscarf – often symbolized as non-belonging – as integral part of Austria. According to Canan, this requires empathy and solidarity. Compared to descriptive MPs, non-descriptive representatives still need political imagination and judgement to symbolically represent minoritized citizens.

Participants further evaluate politicians by looking at the symbolic impact they have on their own community. Participants solely associate this quality with descriptive representatives. Naomi [Surinamese-Dutch_FG9] explains:

I think that for the future generation, Sandra [fellow young participant] and me, that age category, that we see it as a kind of motivation to push higher and perhaps bring that change. […] I think that someone who looks like you is very important for that, because it gives the feeling of yes, I can do it and that person is really there for me.

Germans and Dutch discuss prominent descriptive politicians, who inspire them to get involved in politics. Disengagement from politics is a collective worry. Participants connect disengagement to underrepresentation. In Faiza’s [African-German_FG19] words: “If there were figures who come from the working class and have a migrant background […] they would pick up many more people who disengage because they do not feel represented by old white men.” Many see descriptive representatives as role models who fuel community participation.

The symbolic impact that descriptive political actors make depends on their identity performance. Ayla [Turkish-German_FG13] mentions former chancellor Angela Merkel:

[…] if there is a woman in the Bundestag who doesn't dress so masculine, but like Angela Merkel a bit colorful, whether it's a light green or a pink blazer, where she shows: “Hey guys, we have color here in the Bundestag, I'm a woman and that's important to me.”

Many participants appreciate descriptive representatives who visibly deviate from the white, masculine norm in politics. Visibility normalizes the idea that minoritized citizens are able to rule (Mansbridge 1999). Some participants question the symbolic impact of perceived “assimilated” descriptive representatives.

Nazlı:

Cem Özdemir [Turkish-German minister, Greens] is portrayed a bit as a role model for perfectly successful integration and this spreads a definition of integration in society […] cultural aspects […] for example Turkish origin, are then actually devalued and not included in integration at all. [Turkish-German, FG_15]

This statement indicates that what a descriptive representative symbolizes can, in some cases, result in the inclusion of only a subgroup that conforms more to the dominant (white) norm. However, most participants argue that descriptive representatives have symbolic impact, even if they downplay minoritized identities.

Intersectional and country-variation

The importance participants attribute to symbolic representation differs within- and between groups. Mostly Resettler-Germans, and some Turkish-Dutch/Germans who see their group as relatively privileged mention role models less. Some relate this to a less-visible minority status: “As far as my migration background is concerned, I can choose when I want to talk about it and when I don't, because it doesn't show, and white people are represented enough” [Tom, Resettler-German_FG16]. Others refer to a group that they consider more disadvantaged: “I think honestly, underrepresentation for Black Dutch is much worse, I don't think there's even one [Black politician] in the House of Representatives. It's not too bad with us” [Günay, Turkish-Dutch_FG1].

Conversations about role models reflect country-specificities. The 2021 Dutch elections saw some notable minoritized politicians on the ballot including the first Black woman party leader, Sylvana Simons [minority party BIJ1] and the first woman politician wearing a headscarf, Kauthar Bouchallikht [GreenLeft]. Considering this opportunity, some Dutch women who identify as Muslim or Black hesitate between voting for a role model or for another politician who better represents their interests. Michelle [Surinamese-Dutch_FG10] asks:

I don't know what matters most: whether the person resembles you or simply follows your values […] it is nice, the image that you give young girls or young boys that it is possible to achieve a certain position in politics. But is that really decisive? So, I find it very difficult to know what the most important thing in this political representation is.

Despite this dilemma, most interviewees evaluate politicians based on interest representation. Dutch participants also mention preference voting as a way to counter underrepresentation, provided that descriptive MPs represent them substantively.

Sarah:

It does matter to me that I vote for a woman, because women are still underrepresented in politics. […] As long as she represents my views. [Turkish-Dutch_FG4]

Other participants in this group share Sarah’s reasoning. Hatice: “Yes what Sarah says. […] Sometimes I look for a woman too.” For Arzu, this idea is new: “Never thought about it. Now that I hear it, it is something to consider […] it's nice to give a woman opportunities.” In Germany, similar conversations are absent, possibly because there is no preference voting on party lists.

Though nearly all participants prioritize substantive representation, most value representatives who positively change the social meaning of their group. For participants, the importance of symbolic representation lies in the structural, political consequences it has on emancipation and legitimacy (Mansbridge 1999). In the long run, symbolic representation affects substantive representation. Mostly descriptive, and sometimes non-descriptive politicians who demonstrate political imagination and judgement, contest negative stereotypes that majority citizens hold. But participants see only descriptive politicians as role models, particularly if their identity performance deviates from the white, masculine norm in politics. The intersectional analysis shows that role models notably matter to intersectionally minoritized participants, like Women of Color, Black, and Muslim women. Comparatively privileged participants valued role models less. Strategic considerations reflect country-specific political systems.

Conclusion

Political theorists consider the presence of minoritized politicians to be crucial for the representation and inclusion of minoritized citizens. Drawing on nineteen focus groups (N = 98), this article centers constituents’ insights on representatives “like them”. Who do citizens consider resembles them in politics? How do they evaluate (descriptive) representatives?

Most citizens found descriptive representation an intuitive concept. Politicians resemble them when they share lived experiences, stemming from disadvantage based on race, gender, class, and other intersecting positions. Identification with politicians depended on the political issue. Turkish-Dutch women identified with politicians of color when discussing racism and mention women politicians in relation to sexism. Citizens see descriptive representation as a gradual concept. Representatives are more or less “descriptive”. Descriptiveness depended on which social categories making up intersectional identities are activated in relation to particular political contexts or issues.

Citizens positively evaluate representatives who (1) understand their lived realities (political imagination), (2) adequately interpret this knowledge in forming a political stance (political judgement), and (3) speak or act for them. Although secondary to most, many value symbolic representation by political actors who (4) challenge negative meanings about their group among the dominant majority, and/or (5) empower their community (as role models). In theory, most think that descriptive representatives better represent their interests. Politicians who share their lived experiences can better imagine their lifeworld. In practice, citizens see whether descriptive politicians share their views and can act for them within established parties. Citizens mention instances where non/less-descriptive politicians represent them through empathy, knowledge, and connections to minoritized communities. But, they see only descriptive representatives as role models.

Muslim, Black, and postcolonial citizens more emphasize the importance of descriptive representation to achieve substantive- and symbolic representation. Intersectionally minoritized participants (Women of Color) particularly valued role models. Among citizens with less visible and politicized backgrounds (mostly Resettler-Germans), descriptive representation was less present. Group differences seemed more salient than country-contexts. Despite high minority representation in the Netherlands, both Dutch and Germans often perceive their representation as insufficient. Many question whether descriptive representatives can deliver change within established politics.

This finding complicates the link between descriptive and substantive representation (see Geurts et al. 2024). Many participants want to see politicians who are outspoken on issues that established politics neglects, like racism, and value politicians who visibly deviate from the male, white norm. But many citizens believe that minoritized politicians are not given the space to make alternative voices heard within established politics. Outspokenness that minoritized citizens desire from politicians may be at odds with interests of established political parties to cater to established majority electorates (Dancygier 2018). Additionally, minoritized politicians, like citizens, likely differ in how they self-identify, and in whether and how they want to visibly communicate a (minoritized) identification in politics, especially since visibility can trigger vulnerability and backlash (Schotel 2023).

Rather than as a binary (being represented or not), expanding Pitkin’s (1967) with Arendt’s (1978) concepts clarifies how citizens feel represented on some aspects and not on others. Citizens experience partial representation by politicians who understand their lifeworld, even if they cannot act. Additionally, political imagination as a quality moves beyond the descriptive/non-descriptive distinction. Descriptive representatives have political imagination through their direct, experiential knowledge. But non- or less descriptive politicians can obtain this knowledge too. This requires empathy, responsiveness, and connections with minoritized communities. Imagination does not presuppose that constituents have fixed or objective interests (Saward 2010). Rather, representatives draw on real-world knowledge to imagine groups and interests. As minoritized citizens assess representation in multifaceted and context-specific ways, researchers need a broader range of questions to capture perceived political representation (cf. De Mulder 2023; Street and Schönwälder 2021). Building on single-country studies (Akachar 2018; Bird 2015), I found similarities between Germany and the Netherlands. This suggests that there is something shared in how people think about representatives.

Future measurements of descriptive representation should consider the agency of individuals in defining who resembles them in politics. Researchers could ask respondents to describe degrees of shared experiences with representatives with varying (intersectional) profiles. Studies could also involve minoritized citizens with differing identities and views, for instance, through participatory action designs, where participants co-define the research problem and identity categories (see Amara-Hammou 2023). Focus groups are useful to integrate into such designs. They shift power to participants, allowing them to challenge taken-for-granted concepts and categories. More research is needed into the barriers descriptive politicians face within established parties that impede them from advancing ethnic/racial justice. Studies asking which “critical actors” represent minoritized groups could broaden their scope to representatives who are less tied to established structures, like minority parties or grassroots groups. From citizens’ point of view, what politicians “acting in their interests” means could be further unpacked. Do citizens mostly value representatives’ visible actions, as in this study? If so, how can representatives articulate tradeoffs and political work behind the scenes? The importance of symbolic representation to minoritized citizens suggests that this dimension should be more central to future research. For instance, by further examining how minoritized representatives claim to represent groups (Saward 2010), and how citizens receive claims, depending on politicians’ identity performance.

Together, findings from this study suggest that increasing the number of minoritized representatives in positions of power is necessary but not sufficient. To achieve equal political representation, constituents’ perspectives need to be adequately imagined and articulated in politics.

Acknowledgements

Many thanks to all the participants in this study for sharing their perspectives. I would also like to thank the research assistants who contributed to the data collection process: Hannah Abdullahi, Roman Akyüz, Sati Cakar, Aya El-Yamany, Mariam Hothout, Sina Rauch, and Daniel Schneiß. I am grateful to Liza Mügge, Evelyn Ersanilli, Ton Nijhuis, Anna Korteweg, Michael Hunklinger, Daphne van der Pas, Floris Vermeulen, the special issue editors, and the three anonymous reviewers for helpful comments on earlier versions of this text.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Nederlandse Organisatie voor Wetenschappelijk Onderzoek [452-17-008].

Notes

1 Arendt’s work (1978) is unfinished. Hence, interpretations differ.

2 Arendt (1978, 3) mentions that the Eichmann trial partly motivated her to study mental activities. Eichmann’s inability to think critically caused his evil actions.

3 Arendt (1978, 260–261) takes war as an example. Although one can judge that peace is better, war can happen. Individuals should not abandon their initial judgement (war is bad). But, fully acting on one’s ideals could be foolish.

4 The University of Amsterdam ethics committee approved this project, 2020-AISSR-12519.

5 Although participants discuss descriptive representation as gradual and issue dependent, I use “descriptive” and “non-descriptive” for matters of readability.