Extremes and divides in electoral politics: the need to build bridges

With the decline of mainstream parties in European politics and the more general fading of long-term attachments to political parties, contemporary election research has focused on short-term factors and the rise of new electoral dynamics (Giger et al. 2018, p. 375). In particular, since the 2016 Trump election, the Brexit vote and the steady rise of radical and populist parties all across Europe, studies of electoral politics often steered towards important but often in isolation studied issues of (de)democratic politics like polarization (see Tucker et al. 2018), radical social movements (see Givens 2022; Paternotte and Kuhar 2018), new fringe parties and fragmentation (see Ford and Jennings 2020), conspiracy theories (Butter and Knight 2020), and political (dis)trust (see Zmerli and Van der Meer 2017) or in general: extremes and divides that might undermine sustainable democratic politics.

What is generally lacking is the connection of these particular issues revolving around democratic divided and extremes (Rooduijn 2019; Blais et al. 2020) while, as also underscored by Russell Dalton (2020, p. 96) on political science in general, it is very fruitful to build “bridges between related islands of theory and research”. Therefore, in this special issue, we aim to contribute to the field of electoral politics, and more particularly to focus on strands of research on issues and/or politics of extremes and divides that are related. It focuses on extremes on the ideological dimensions, belief polarization and non-mainstream voting. It also focuses on extremes in doing politics and how voters perceive incivilities of political actors in the political arena. As such it contributes to insights in how extremes and divides, like radicalization and fragmentation, reinforce or counteract each other and might help to pinpoint underlying processes. We do so by bringing together insights from different fields of extremes and of divides in one single issue.

Theoretically, this contribution brings together studies on the causes, political manifestations, and democratic implications of these extremes and divides. Methodologically, we keep core elements constant across the empirical contributions in this issue: each takes the same elections as a starting point and all have the Dutch Parliamentary Election Study 2021 at their core. This lays a foundation to bridge different debates and literatures. By having an invariant context—the 2021 Dutch parliamentary elections—and keeping the core data and method constant, we assure commonalities between the studies in this special issue. This in turn, allows for a co-reading of multiple articles in this issue that is particularly meaningful and allows us to provide the first connected insights and above all a springboard for future work into the study of extremes and divides.

Due to the broadness of the research fields under study, it would be impossible to build the bridges between them in the remainder of this introduction. We gladly refer to the different contributions themselves. Rather, this introduction aims to provide three building blocks that allows us to build the bridges. First, we will sketch the context of the 2021 Dutch parliamentary elections, including for instance the run-up to the elections, the main elements of the campaign, and the role of the covid pandemic. This section positions the Dutch case and discusses the ways in which it is particularly relevant for the literature at large. Second, we will provide brief information on the dataset central to each of the articles, highlighting how it can be connected to other data sources as done by several of the contributions. Third, we will provide a brief discussion of each of the seven contributions to this special issue, highlighting the main focus of the article, some results particularly interesting for the overall special issue and providing examples of connections with the other contributions.

The 2021 Dutch parliamentary elections

In 2022, Otjes and Voerman (2022, p. 323) wrote: “the political year 2021 was the most tumultuous period in the Netherlands since 2002”, but actually this was not due to the elections. The 2021 Dutch parliamentary elections illuminated the ongoing process of realignment between political parties and voters in Dutch politics, and can be considered rather typical in a trend. As such they form a relevant case to study extremes and divides, which may deepen in the future and that concurrently develop across Western European democracies. Before we delve into the question of how this particular case relates to other cases, we will first describe how this tumultuous period in Dutch political history unfolded.

Rutte III

On January 10th, 2021, the third Cabinet Rutte (i.e. the Dutch government) resigned before the end of its term, resulting in new elections: the 2021 Dutch parliamentary elections. What happened?

After the 2017 Dutch parliamentary elections, it took 225 days to form a coalition government, the longest Dutch postwar formation process until then. The formation resulted in the Rutte III government, a coalition of the conservative liberal VVD, the Christian Democratic CDA, the liberal democratic D66 and the Christian CU (for all party names, abbreviations and families see Table 1). The government faced several major and pressing policy issues, such as the pension policy in the Netherlands and the climate and environmental policy (i.e. carbon and nitrogen emissions needed to be reduced, partly because of court judgements) (Van Holsteyn and Irwin 2022, p. 1494). Major challenges that came up during the government period included the evacuation of people who had worked in service of the Netherlands in Afghanistan after the Taliban takeover and dealing with the covid pandemic (more below).

Table 1 Results of the Dutch parliamentary elections 2021 and 2017
Full size table

To compound on these policy problems, a crucial policy crisis that unfolded during the Rutte III government was the so-called social benefits scandal (Van Holsteyn and Irwin 2022). This scandal had primarily found their roots under the previous Rutte governments, but was disclosed in a series of reports and government responses during the Rutte III government. At the core, this scandal led tens of thousands of parents with young children into (severe) financial trouble—including homelessness, debt restructuring and displacement of children (see POK 2020)—as the direct consequence of government action and inaction. The scandal was the outcome of a toxic recipe with multiple ingredients: harsh laws that were set in place under pressure of parliament and media, general distrust on the side of the Tax Administration, ruthless fraud hunt downs, governmental unresponsiveness to signals of distress, and structural discrimination of people with a migration background. When the scandal finally came to the light during the Rutte III government period, the “inadequate reactions by civil servants and politicians to the unfair procedures that had been followed in combatting fraud with child benefits payments” (ibid) became politically salient. In December 2020, a parliamentary interrogation committee on the role of the political leaders in this ongoing scandal reported their findings, in a report called Ongekend Onrecht (Unheard of Injustice; POK 2020). Parliament was scheduled to debate the report with the government by mid-January 2021. It became clear that at least the full opposition seriously considered supporting a motion of no confidence against the Cabinet Rutte III. Immediately before that parliamentary debate the Cabinet resigned. A majority of the Dutch population considered the resignation just, and of all parties, Prime Minister Rutte’s VVD was considered mainly responsible (Van der Meer and Lubbers 2021).

The 2021 campaign

Despite the scandal and the resulting fall of the coalition government that instigated new elections, the social benefits scandal ranked only third in the written news topics during the campaign periodFootnote 1 (Takken 2021). First and second came the covid pandemic (including economic consequences) and education and youth, also much in relationship to the pandemic, with the Netherlands being in lockdown during the elections. Only about 15% of the voters indicated before the elections that the scandal and the resignation would influence their vote (Van der Meer and Lubbers 2021, p. 4).

Despite the lockdown, the 2021 campaign was a remarkably normal Dutch election campaign in various ways. Traditional and social media were the dominant campaign modes, in contrast to canvassing. This aligns with the relatively low level of campaign resources in the Netherlands in comparison to other established democracies (Van Holsteyn and Irwin 2022, pp. 1496–1497). At the same time, media were even more important this time as public events and canvassing were forbidden as part of the policies to contain the covid pandemic. In this light, Van Holsteyn and Irwin (2022) report, based on the DPES data (Jacobs et al. 2022) that “about 25% of all voters watched at least one televised debate, 45% parts of at least one, [while] about 30% did not watch any televised debate” (Van Holsteyn and Irwin 2022, p. 1497).

Good governance and new leadership were the main echoes of the social benefits scandal in the campaign. Sigrid Kaag (an acclaimed diplomat turned Minister in Rutte III) became the main, progressive challenger of Rutte for liberal democratic D66 (Otjes and Voerman 2022). Another potential rival, the social democratic leader Asscher, had stepped down as political leader of his party, acknowledging his partial responsibility in the benefits scandal under Rutte II. In opposition to Prime Minister Rutte, who remained party leader for the VVD, Kaag was able to present herself as a new leader promising a new way of politics. In the final days of the campaign, she drew votes from other progressive parties, most notably the social democrats and the Greens (Otjes and Voerman 2022).

The issue of leadership could become prominent in terms of the campaign, because the main issue—the covid pandemic—was an issue with major consensus and few policy alternatives (Van Holsteyn and Irwin 2022, p. 1498), which has been linked to classic ‘rally-around-the-flag’ dynamics (Van der Meer and Lubbers 2021; Van der Meer et al. 2023). Whereas most parties were generally supportive of the government's measures against the pandemic, the populist radical right PVV was somewhat more skeptical about measures regarding the pandemic. But particularly Forum for Democracy (FVD) opposed the lockdown measures (Van Holsteyn and Irwin 2022, p. 1498) in such an extreme way that conspiracy theories and extremism became part of their discourse. This lead to a split and the emergence of the new far right party JA21 as well as some media coverage during the campaign finale, with the FVD, albeit small in the polls, getting even more attention than the left-wing opposition (Takken 2021).Footnote 2 Regarding covid, analyses on voters indeed reflected the above described strong divide. Among FVD voters over 90% was dissatisfied with the covid policies, making it virtually a necessary condition for voting for this party. Among PVV voters, dissatisfaction was slightly above 50%; whereas among voters for all other parties the level of dissatisfaction was below the 50% mark (Sipma 2021a; see also Kessenich and Van der Brug 2024 in this issue).

Despite institutional racism being part of why the Cabinet resigned and despite ongoing debates on Black Pete (a stereotypical blackface companion of the Dutch equivalent of Santa Claus) and the visible presence of the Black Lives Matter Movement, racism and diversity did not become major campaign issues (Takken 2021). However, interlinked with these debates and movements, a new intersectional anti-racist, anti-capitalist, feminist party (BIJ1) was considered to have a serious possibility to win a seat in parliament on this extreme left agenda, next to the more traditional and ethically conservative ethnic-minority interest party DENK, which already held three seats in parliament, mainly drawing from the Moroccan-Dutch and Turkish-Dutch Muslim electorates.

A final noteworthy observation about the campaign is that climate and nitrogen policies played far less a role than was expected by some before. While both D66 and the Greens pushed a green agenda and the newcomer BBB pushed a pro-farmers agenda (i.e. objecting the proposed policies), the campaign dynamics did not lead to clear saliency of this issue.

The 2021 Dutch parliamentary elections result

As the elections took place during a lockdown, different measures were taken to ensure legitimate and fair elections, while protecting the vulnerable and allowing them to vote. Polling stations were open for 3 days instead of one, social distancing was applied, and elderly could vote by mail if they wanted (Otjes and Voermans 2022, p. 323; for details see also Van Holsteyn and Irwin 2022), preventing a lower turnout than initially feared (Sipma 2021b). With a turnout of 78.8%, it was slightly lower than the turnout during the previous parliamentary elections in 2017 (81.9%, see Table 1) and falls within the broader range of turnouts for parliamentary elections since the turn of the century (ranging from 74.6 to 81.9%).

The campaign dynamics described in 2.2 resulted into the continuation of the historically high level of fragmentation. Nevertheless, the election results had a relatively low level of system-level volatility (Otjes and Voermans 2022), with the former coalition parties maintaining their majority and eventually forming the same coalition once again. Van der Meer and Lubbers (2021) compared the election with a rollercoaster: a lot happens between the start and the end, but you end where you started.

An overview of the results per party is given in Table 1. As already stressed, the low system-level volatility is one of the main patterns. However, in terms of divides and extremes more results are notable. First, whereas junior partners in coalition governments are often electorally punished in the next elections (Klüver and Spoon 2020), D66 gained a total of 6 seats compared to the previous election, implying that the aforementioned campaign for new leadership by D66’s party leader Kaag seems to have paid off (Van Holsteyn and Irwin 2022, p. 1499). Second, the elections resulted in the largest number of small parties since turn of millennium and also the two or three largest parties together continued to be on an all-time postwar low (Van der Meer and Lubbers 2021, p. 6). In fact, the elections resulted in the largest number of parties since introduction of universal suffrage as a further sign of a fragmentized electorate (Van Holsteyn and Irwin 2022, p. 1499), as four political parties made their electoral inroads: BBB, BIJ1, JA21 and Volt. In combination with other parties that made their way into Dutch parliament during the past decade, it reflects the emergence of new politics of extremes and divides.

Various political parties within the Dutch system are positioned at the political flanks. BIJ1 won their first national seat running on an extreme left-wing and identity-based platform, Volt is a progressive European federalist party and positions itself on the flanks on the libertarian-authoritarian dimension. Moreover, FVD was the major winner and biggest party in the 2019 regional elections, but having lost much of its support as a result of the many controversies and infighting in the party. JA21 is a split of FVD which presented itself as the reasonable alternative, positioning itself as less extreme, but therewith contributing to further fragmentation within the far-right electorate being the third party in this cluster. Other smaller parties won seat(s) running on a platform that was more centered on a specific interest, resulting in various divides being represented by different parties in parliament, including the already mentioned BBB (agrarian interests), BIJ1, DENK (ethnic-minority interests), PvdD (animal rights). This argument could also be made for the oldest party in parliament, the orthodox Christian SGP. Third, left-wing parties (PvdA, GL, SP, PvdD, BIJ1) hit a postwar low. They obtained merely 33 seats, 9 (6% points) fewer than their previous record low of 42 seats that they had obtained twice in the previous two decades: in 2002 (the elections after the electoral rise and subsequent assassination of radical-right populist newcomer Pim Fortuyn) and the previous elections of 2017 (when the social democrats lost three quarters of their seats after their participation in the Rutte II coalition government).

In brief, on the one hand the elections showed stability. On the other hand, however, the elections lay bare a population and party system in which extremes are accommodated. Extreme politics are present at multiple flanks and on multiple axes. Moreover, social divides have translated to broad parties being partly replaced by special-interest parties, and together with a politics of extremes this resulted in the decline of the left, paired by the rise of the radical right.

Extremes and divides in a consensus culture and open system

Extremes and divides are a recurring theme in the discussion of (the run up to) the 2021 Dutch parliamentary elections. Elements such as (a lack of) political responsiveness, trust, conspiracy thinking, extreme parties, institutional racism, volatility, media coverage, social group segregation and other processes are (likely to be) interrelated, also at the level of voters and in the micro dynamics of elections. Considering them together might shed new light on existing debates and unanswered questions and lead to new and innovative ideas.

The 2021 Dutch parliamentary elections provide a unique setting that can be particularly fruitful for uncovering these connections and sparking new ideas. On the one hand, Dutch political culture has long been, and in comparative perspective still is, one of consensus. But, as stressed by Otjes and Voermans (2022, p. 323) when referring to the elections and ensuing events resulting from it, the politics of 2021 “stands in stark contrast to the consensual political culture in the Netherlands”. In other words, to some extent the Dutch case is a least likely case for finding strong extremes and divides in politics, so what found for this case is very much likely to be relevant to other cases too (Seawright and Gerring 2008, p. 213).

Second, whereas extremes and divides at the political arena are relatively less likely to occur due to the political consensual culture, extremes and divides at the voter-level relatively easily make way into parliament. The Dutch political system is one of the most open ones among European or Western democracies (Jacobs and Spierings 2016), which makes it easier for parties to gain seats, for groups to gain (segregated) representation, and for individual politicians to be elected. In other words, if micro-level demand processes are taking place that steer politics towards divided and the extremes, the supply side of politics easily accommodates this and make such processes visible in macro-level politics. As such one could say the Dutch institutional setting provides a most-likely setting (Seawright and gerring 2008, p. 214) for actually being able to observe micro-level processes.

In brief, we argue that the contributions in this special issue are important studies to understand Dutch politics. Moreover, zooming in on the Dutch case is particularly useful to identify micro-level processes on the theme of the special issue that are likely to be relevant to other established West-European democracies, which could be further assessed in future work replicating the results from this issue on other cases and settings. As such, this issue thus contributes to generating new theories and hypotheses on the interrelatedness of different aspects of political extremes and divides.

The Dutch Parliamentary Election Study 2021

One common dataset: DPES21

Above we contextualized the contributions in this study by providing background information regarding this election and the Dutch party system, and we discussed how the setting of these elections provide an interesting case to study the interrelatedness of the politics of extremes and divides for the larger literature. However, to actually provide studies shedding more light on these issues, rich data on these elections is needed, and that is where the Dutch Parliamentary Election Study 2021 (DPES21) comes in (Jacobs et al. 2022; documentation: Sipma et al. 2021a, b).

All contributions in this special issue use the DPES21 survey data.Footnote 3 This data collection is part of one of the longest running data projects, as DPES has been collected among representative samples of citizens eligible to vote since the 1971 Dutch parliamentary elections. While keeping core items comparable over time, the survey developed over the years, including topical issues and keeping pace with societal and technological developments.

Divides and extremes in DPES21

To the end of including topical issues and scientifically innovative concepts, the DPES21 organization launched a 'junior scholars' call. Via this call, junior scholars from the Netherlands or connected to a Dutch academic institute could apply for including items in the survey on the topic of their research, with a criterion that it should fit the goals of DPES (understanding the elections results and dynamics and politics more generally) and deepen or broaden the dataset. Nine of twelve applications were (partly) honored, that led to the inclusion of several important new themes.

Evidently, this special issue cannot and does not tackle each of these themes or all possible interconnections between them. However, the junior scholars call has been a way to involve scholars who are at different stages of their career, but all at the forefront of their field, putting forward new ideas and measurements, such as perceived incivility among political actors and affective and factual polarization. As such the DPES21 includes a unique set of items related to extremes and divides that can be studied, also in combination with each other, and offers promising avenues for future research. This issue represents a first step in this process.

Connecting DPES21

Creating future opportunities also informed the various ways in which DPES21 connects to other data projects. The data collection was organized and designed to allow merging and matching with several other data sources. First, DPES21 is part of the Comparative Study of Electoral Systems (CSES) dataset (Module 5). This entails that an internationally standardized set of survey items has been included in DPES21 allowing for cross-national analyses.

Second, for the first time in the DPES history, a separate sampling of respondents who (have at least one parent who) migrated from Africa, Asia or Latin AmericaFootnote 4 to the Netherlands. This first Dutch Ethnic Minority Elections Study (DEMES21; Lubbers et al. 2022)Footnote 5 used by and large the same questionnaire as the DPES21, adding some migration and marginalized-minority specific questions, dropping some,Footnote 6 and slightly rephrasing some for clarity (cf. Sipma et al. 2021a, b).Footnote 7 The DEMES21 data allows more detailed comparisons between the ethnic-majority and ethnic-minority citizens [e.g. Spierings and Vermeulen 2024 (this issue)] as well as in-depth analyses of ethnic-minority and/or migrant-background voters [e.g. Lubbers et al. 2024 (this issue)].

Third, as of 2017 the DPES data are partly collected within the LISS panel data infrastructure. Consequently, a considerable number of respondents has participated in both the 2017 and 2021 DPES, allowing for panel analyses [e.g. Kessenich and Van der Brug 2024; Vermeer et al. 2024 (both this issue)]. Moreover, the LISS panel itself included many more waves and modules allowing for multiple waves panel analyses. This includes additional elections study waves in between 2017 and 2022 that have been collected and merged (Van der Meer and Fabry 2023). In brief, the DPES21 is part of panel structures that allow for analyses of what drives individual-level change and more stringent tests of causality.

More specifically, and fourth, the Stichting Kiezersonderzoek Nederland (Dutch Foundation for Election Research; see SKON 2024) has presided data collections on other nation-wide elections and voting booth activities within the LISS panel, including the Dutch Local Election Studies 2016, 2018, and 2022; the Regional and Water Board Election Study 2020 and the Dutch National Referendum Studies 2016 and 2018. These surveys can be linked to create panels and for instance compare first and second order elections with each other.

Fifth, as mentioned, DPES21 is part of a tradition since 1971 and the separate DPES datasets are continuously merged in the longitudinal file. Currently the 1971–2006 files are merged and publicly available (Aarts and Todosijevic 2009) and the updated 1971–2021 dataset is being worked on and planned to become available in the coming year. These datasets enable diachronic comparative analyses to describe and understand long-term trends on an internationally unique time span of five decades.

Sixth and finally, next to these options to merge the DPES21 data in ways that have been considered when designing the data collections, other options are available due to the detailed information in DPES. For instance, the detailed information on which news sources respondents consume is linkable to the unique automated content analyses of newspaper from the ODISSEI Media Content Analysis Lab [Vermeer et al. 2024 (this issue)], and individual-level register data available through Statistics Netherlands (CBS 2024).

Turning out attention to the core dataset in this special issue again (DPES21), more detailed information on it, including the questionnaire, is provided in the publicly available data documentation (Sipma et al. 2021a, b). In this introduction, we set out to show how the DPES21 forms the stable core of this issue and how it is connected to other datasets. Each of the contributions to the special issue highlights how it uses DPES, what part of it, and, when applicable, combined with what other data. Below, when we raise new questions based on the contributions in this issue, the possibilities for other data connections might be relevant again, as they are for other future work.

Contributions to this special issue and the field

This special issue features seven studies, each with its own unique contributions to a special field within comparative politics, public opinion and voting studies. Below, we do not intend to exhaustively summarize each study; we focus on the elements most central to the special issue at large and the connection between dimensions of extremes and divided within the study (i.e. some focused teasers) as well as the ideas they sparked for further studies (i.e. a research agenda for the field at large). We do so by clustering the contributions in terms of the focus of their explanandum, be it either voting behavior or attitudes related to the fundaments for our democratic politics.

Voting behavior

Four studies contribute to understanding voting behavior, of which the first two do so in a general way (i.e. not focusing on specific, extreme parties). Van der Meer and Damstra (2024) zoom in on associative issue ownership (AIO). They study the extent to which this concept manifests within a fragmented multiparty system, and the extent to which it shapes voting behavior as a two-stage process (explaining consideration set composition and party choice). Among other things, they show that AIO influences both voting and peoples’ choice set of parties. The latter contributes to our understanding of inter-party-bloc volatility and thus party system instability. Moreover, this raises important questions for future research on party switching between party blocs. For instance, what causes people to do so, in particular towards a bloc including parties with very extreme parties, or those that are presented as such in media. Which brings us to the study of Vermeer et al. (2024), who focus on media coverage and impact thereof on voting. By combining survey data on voting and media consumption with the actual media content, they show that political parties benefit from coverage in newspapers as well as online news outlets, and this particularly holds for new parties, feeding into fragmentation. Moreover, negative online coverage seems to benefit parties too, which suggests that standing out by taking extreme positions or rude behavior (see Van Elsas and Fiselier below) might be beneficial. One of the questions this result raises for future research on the theme of this special issue, is whether there is a turning point from which extremity and rudeness are not beneficial anymore and what the characteristics are of voters who are mobilized by such negative coverage.

The other two studies explaining voting behavior zoom in on extreme or special interest parties. Kessenich and Van der Brug (2024) assess who votes for populist radical right and far right parties, in particular switching between and to those parties. A core finding here is that voters switch more between these parties than from other blocs to these parties, but that having more radical and far right parties also increases their voter base (switchers towards the bloc). Illustrative here is that Forum for Democracy at first presented itself as the reasonable version of Wilders’ PVV, but turned more and more extreme, espousing conspiracy theories of the most manifest kind, antisemitism, crude homophobia and sexism and blatant racism. Following Kessenich and Van der Brug’s results, we could raise the question whether switching from non-radical right parties to FVD helps us understand how people are being draw into extremism, and ultimately may get lost in conspiracy theories and rejecting facts as a result of such a switch (see also Rekker and Harteveld below). Lubbers et al. (2024) delve into the question of who among ethnic-minority migrant-background voters have a higher propensity to vote for one of the different ethnic-minority interest parties that were on the ballot. Based on their results we can conclude that political fragmentation takes place along ethnic and religious dividing lines. Moreover, their study provides indications that perceived group discrimination might feed into voting for these special interest parties, which cannot be seen independently from other developments. Considering the politicization of migration and the context of the social benefits scandal, we suggest that the popularity of populist radical right (see Kessenich and Van der Burg above) and harsh uncivil campaigning (Van Elsas and Fiselier below) may have pushed citizens with a migration background towards a new type of parties that is regarded as issue owners on issues related to ethnic identity and discrimination.

Fundaments for our democratic politics

The final three studies help to understand the support of the democratic system and its responsiveness. Van Elsas and Fiselier (2024) assess how perceived ideological fragmentation (more choice) and perceived incivility (conflict) in the campaigns shape peoples’ judgement on the functioning of the democratic system. They show that the first leads to more positive judgement, but the latter is more likely to feed into negative judgement. As such they show that different divides can have contradictory effects, including positive ones (see also Spierings and Vermeulen below) and that studying them simultaneously is worth the effort of bridging fields. A further issue these results raise is what happens with the judgement of the functioning of democracy if different types of divides overlap or even map onto demographic divides (i.e. segregation).

Spierings and Vermeulen (2024) also go into the evaluation of the democratic system and its actors together with one’s likelihood to vote, but particularly among ethnic-minority migrant-background voters. Besides showing differences between ethnic groups, they show that experienced discrimination and underrepresentation is, if anything, likely to lead to more negative evaluation of the political system’s functioning, while it spurs electoral participation. Considering this outcome, together with the Lubbers et al.’s (above) finding that perceived group discrimination might drive ethnic-minority migrant-background voters to prefer ethnic-minority interest parties, makes it interesting to raise the question how this translates to other countries and systems. The openness of the Dutch democratic system for new and small parties and in terms of preference voting is rather exceptional. In more closed systems, ethnic-minority migrant-background voters who perceived more discrimination might not only judge the system less positively, but also turn away from participating as they have no fitting option, which is likely to create a further feedback loop with the evaluation of the system. Comparative studies on this dynamic are welcomed.

Finally, Rekker and Harteveld (2024) zoom in on a more fundamental prerequisite for holding meaningful discussions in a democratic system: factual belief polarization. In brief, they assess whether one’s ideological position on an issue is linked to whether they assess the facts to be ‘in favor’ of their argument, and how this correlation varies by other factors. One of the arguably more surprising, albeit not unequivocal, patterns they discern is that among higher-educated and politically interest citizens perceptions of the facts and attitudes are more in line. They provide several methodological and substantive explanations for this, but it also raises wider questions on the causes of factual beliefs and the role media play or whether ‘facts’ from parties that are considered issue owner on that topic are accepted more (see Van der Meer and Damstra above).

In brief, reading one of the studies in this special issue will certainly provide new insights on what drives or withholds politics in becoming more extreme and divided, and what the effects of such processes are. It does so for the contemporary Dutch case particularly, and with significance for democratic politics in other countries, as argued above. However, we would suggest reading multiple of the contributions. It is combining the insights from them and letting the mind create connections and formulate new expectation what especially drives innovation needed to understand extremes and divides in democratic politics.