Introduction
Conflict is an inherent facet of the political landscape, shaping the dynamics of democratic societies. Rooted in the clash of divergent political ideas and ideological perspectives, democracy finds its essence in contestation among different actors (Sartori Reference Sartori1976; Schattschneider Reference Schattschneider1975). While fundamental to the democratic fabric, conflict adds complexity to governance, particularly in multiparty systems whose decision-making hinges on an intricate balance of cooperation, compromise, and consensus (Bartolini Reference Bartolini2000; Lijphart Reference Lijphart1999). Conflicts can take different forms. In the most original form of conflict, parties are fundamentally competing about different policy measures. However, conflicts may also arise for ideological (i.e., different goals or values) or even personal reasons. For example, the Scottish First Minister recently resigned because he could not agree with his coalition partners on climate policy objectives, among other things.Footnote 1 In Germany, the absence of collaboration between the SPD and Die Linke is also based on the conflict between the former long-time SPD chairman Oskar Lafontaine and the then party leadership under Gerhard Schröder.Footnote 2 There are many more examples, but little is known about how citizens and politicians perceive these different conflicts, especially whether they judge ideological, substantive, and personal conflicts in a distinctive manner and how their perceptions of the role of these conflicts align. In the context of political representation, previous research has primarily focused on policy preference congruence between citizens and elites, largely neglecting the perceptual alignment on core elements of the political system, such as what constitutes legitimate political conflict. Such forms of alignment can be crucial for maintaining a shared understanding of how democracy as a system should function. We contribute to this scholarly debate on the evaluation of conflict by systematically comparing citizens and elites across three key conflicts. Specifically, we focus on individual-level heterogeneity among citizens and politicians. Moreover, we add to the scholarly debate not only by comparing citizens and politicians but also by assessing how elites perceive citizens’ evaluations of conflict.
Existing research on political conflict has primarily examined the causes of political disagreement between political parties or politicians’ antagonistic behavior (e.g., Aistrup Reference Aistrup2004; Skoog and Karlsson Reference Skoog and Karlsson2018, Reference Skoog and Karlsson2022) and how these conflicts affect citizens’ political attitudes and behavior (e.g., Han and Federico Reference Han and Federico2017; Kleinnijenhuis, van Hoof, and Oegema Reference Kleinnijenhuis, van Hoof and Oegema2006; Schuck, Vliegenthart, and De Vreese Reference Schuck, Vliegenthart and De Vreese2016). For example, Skoog and Karlsson (Reference Skoog and Karlsson2018) show that an increase in conflicting interests in society, a harsh economic climate, and intense party competition lead to more antagonistic behavior by politicians. Additionally, research shows that news presented through the lens of conflict (Schuck, Vliegenthart, and De Vreese Reference Schuck, Vliegenthart and De Vreese2016), focusing on clashes between political figures, has been associated with increased political distrust (Kleinnijenhuis, van Hoof, and Oegema Reference Kleinnijenhuis, van Hoof and Oegema2006). These studies, however, overlook what ordinary citizens think about the role of conflict in politics and how their views differ from political elites. This is exactly what we add in this study.
The question of whether citizens and their elected representatives share a common understanding of the role of conflicts in politics is a pertinent one. Representative democracies institutionalize conflict within the party system, which is often organized along ideological lines, resulting in clashes between political parties (Kriesi, Grande, Dolezal et al. Reference Kriesi, Grande, Dolezal, Helbling, Höglinger, Hutter and Wüest2012). The institutionalization of conflict raises critical questions about political representation and whether citizens and their elected representatives share a common understanding of the role and meaning of political conflict. The way conflicts are perceived is essential for the process of policy-making; no public support for policies creates barriers to effective governing (Fung Reference Fung2015). Furthermore, discrepancies between citizens’ and politicians’ understandings may challenge political representation (André and Depauw Reference André and Depauw2017). Constituents may not feel adequately represented when their elected officials engage with or support political conflict that is counterproductive to the function of the political system in their eyes. A misalignment between elites’ and citizens’ perceptions of conflict may not only affect evaluations of politicians or parties but also the perceived legitimacy of democratic institutions themselves (Banda and Kirkland Reference Banda and Kirkland2018; Hibbing and Theiss-Morse Reference Hibbing and Theiss-Morse2002). For example, if citizens interpret conflict as dysfunctional while elites view it as a deliberative necessity, democracy might risk being seen as self-serving or even chaotic. Such perceptual misalignment could reflect (or help explain) broader trends of, for example, declining political trust and populist discontent.
Drawing on previous research on partisan conflict, we argue that citizens and politicians perceive political conflict differently. For citizens, conflict within the political arena may come across as hampering effective governing (Hibbing and Theiss-Morse Reference Hibbing and Theiss-Morse2002). Through mediated coverage of politics, where conflict and game framing tend to gain disproportionate attention or be portrayed sensationally (Aalberg, De Vreese, and Strömbäck Reference Aalberg, de Vreese and Strömbäck2016), citizens might perceive conflict as an unnecessary clash between parties, hindering the smooth functioning of government. In contrast, politicians are directly involved in the political decision-making process and may thus interpret political conflict in a different light. Rather than viewing conflict as inherently difficult, politicians may acknowledge it as an integral aspect of democracy. Political conflict, from their perspective, serves as a mechanism for debate, negotiation, and compromise, all essential elements for the functioning of a vibrant democratic system (Sartori Reference Sartori1976; Schattschneider Reference Schattschneider1975).
Empirically, our study relies on a survey experiment conducted among a representative sample of citizens (n = 8264) and political elites (n = 331) in four countries (French-speaking Belgium, the Netherlands, Portugal, and Switzerland). These countries are diverse in terms of party landscapes and approaches to political conflict resolution, which allows us to understand whether perceptions of political conflict are comparable across multi-party systems. During the survey, both groups were presented with a survey experiment in which they had to evaluate different conflict types. Overall, politicians were more likely to endorse conflict scenarios than citizens, confirming the study’s main hypothesis. We also found differences in the preferences for certain conflicts. While both groups prefer substantive and ideological conflicts over personal conflicts, politicians are more likely to approve of substantive and ideological conflict than citizens and less likely to approve of personal conflicts. Interestingly, politicians seem to project their own perceptions on citizens, expecting them to be more approving of substantive and ideological conflict. When zooming in on the differences between parties and voters across different party families, we find that citizens’ levels of ideological extremity correlated with increased approval for conflict in certain countries, but politicians from radical party families were not consistently more approving of conflict, signaling another mismatch in perceptions.
Theoretical framework
To understand how citizens’ and politicians’ perceptions of conflict align, we first focus on their general perceptions of the role of conflict in politics and their evaluation of different types of conflict. Additionally, we explicitly examine how politicians think citizens evaluate conflict, formulating a separate hypothesis for politicians’ perceptions of citizens’ evaluation of conflict. Finally, since we anticipate that politicians from different party families and supporters of different parties have distinct preferences for conflict, we reflect on the role of party family in the last part of this theory section.
Politicians’ and citizens’ perceptions of conflict
Politicians hold differing ideas on policy decisions and governance, which makes conflict an essential element of politics (Bartolini Reference Bartolini2000; Schattschneider Reference Schattschneider1975). Erasing conflicts from politics is not only impossible but also undesirable. Conflict allows consideration of different perspectives and provides legitimacy to political decision-making (Schattschneider Reference Schattschneider1975). Especially in consensus democracies, both conflictual and consensual patterns of behavior of political actors are central (Burkhart and Lehnert Reference Burkhart and Lehnert2008). Political parties need conflict to signal their distance towards other parties and to distinguish their position from that of others, yet political parties also want to influence policy decisions that push them toward cooperation (Laver Reference Laver2005; Meguid Reference Meguid2005; Moury and De Giorgi Reference Moury and De Giorgi2015). Hence, while consensus may often be needed for policy outcomes, this is often preceded by fierce disputes between and within parties, between government and opposition, or within the coalition cabinet (Burkhart and Lehnert Reference Burkhart and Lehnert2008; Vercesi Reference Vercesi2016).
Consensus and conflict are both crucial elements in politics. Yet, politicians’ relationship with conflict can be considered ambiguous. As representatives of part of the electorate, normative perspectives of representation prescribe that representatives should express citizens’ positions and preferences, which in turn makes conflict salient (Mansbridge Reference Mansbridge1999; Wolkenstein and Wratil Reference Wolkenstein and Wratil2021). This is crucial from a vote-seeking perspective (Müller and Strøm Reference Müller and Strøm1999). To gain and maintain electoral support, politicians need to cultivate distinct party brands (Downs Reference Downs1957; Meguid Reference Meguid2005). However, politicians also have to overcome conflicts and seek common ground and compromise to translate preferences (partially) into policies (Müller and Strøm Reference Müller and Strøm1999). Likewise, citizens’ attitudes toward conflict are complex. Some scholars have argued that citizens tend to dislike political conflict and that observing elite conflict can lower levels of political trust (Hibbing and Theiss-Morse Reference Hibbing and Theiss-Morse2002; Lipsitz, Trost, Grossmann et al. Reference Lipsitz, Trost, Grossmann and Sides2005; Ulbig and Funk Reference Ulbig and Funk1999). Hibbing and Theiss-Morse (Reference Hibbing and Theiss-Morse2002) argue that citizens dislike conflict because they think it poses an obstacle to effective governing. However, research also shows that citizens appreciate conflict. For instance, Fortunato (Reference Fortunato2012) finds that citizens have a higher perception of the quality of representation when parties in a coalition fail to reach an agreement to maintain their electoral promises.
Nevertheless, we do expect that, in general, politicians are more likely to approve of conflict compared to citizens. Politicians are likely more aware of, and exposed to, the function conflict serves in a democracy. Importantly, this might depend on the type of conflict. Conflict can take different shapes and forms. For example, it can concern negative rhetoric towards the opponent (Røed, Bäck, and Carroll Reference Røed, Bäck and Carroll2023), ideological distance (McCarty, Poole, and Rosenthal Reference McCarty, Poole and Rosenthal2016), the inability to form policies (Bäck and Carroll Reference Bäck and Carroll2018), and differences in party positions on policy issues (Franchino and Høyland Reference Franchino and Høyland2009). Before turning to the different conflict types, we pose a general hypothesis:
H1: Politicians are, in general, more likely to approve of conflict, regardless of its type, as opposed to citizens.
Types of conflict
While conflict is often treated as a broad, overarching concept, it is crucial to recognize its various manifestations. While politicians and citizens may differ in their overall evaluation of conflict, they could align in the kind of conflicts they value over others. Existing work suggests that to understand the role of conflict in democracy, it is essential to distinguish between conflicts about substantive issues and those concerning personal relationships (Bartholomé, Lecheler, and de Vreese Reference Bartholomé, Lecheler and de Vreese2018; van der Goot, Kruikemeier, de Ridder et al. Reference van der Goot, Kruikemeier, de Ridder and Vliegenthart2024). The core of politics is conflict about substantive matters. Previous research defines conflict as a situation where there is disagreement over objectives or where different interests cannot be reconciled (Skoog and Karlsson Reference Skoog and Karlsson2018). For instance, as Skoog and Karlsson (Reference Skoog and Karlsson2018) highlight, ‘parties may disagree on political objectives and on what constitutes a good society, or the parties may have similar objectives but different views on how these should be pursued’ (184). However, in politics, conflicts can also concern personal relationships. Politicians may express disagreement based on personal performance or evaluation of other politicians (Waddle, Bull, and Böhnke Reference Waddle, Bull and Böhnke2019). They can decide to base their criticism not on the viewpoint of their opponents, but rather on their character and skills (Macagno Reference Macagno2013).
Although the literature has yet to explore how citizens evaluate the role of personal and substantive conflicts in parliament, studies explore how different types of criticism in political campaign communication affect citizens’ perceptions (Brooks and Geer Reference Brooks and Geer2007; Carraro and Castelli Reference Carraro and Castelli2010; Hopmann, Vliegenthart, and Maier Reference Hopmann, Vliegenthart and Maier2018; Min Reference Min2004; Nai and Maier Reference Nai and Maier2021; van der Goot, Hameleers, Vliegenthart et al. Reference van der Goot, Hameleers, Vliegenthart, de Ridder and Kruikemeier2026). These studies differentiate between politicians criticizing each other on their substantive ideas (e.g., policies and electoral programs) versus attacking each other on their personality traits and personal wrongdoings. When asked directly about their perceptions, citizens argue that they do not find criticism per se problematic but prefer criticism on substantive issues over personal attacks (Lipsitz, Trost, Grossmann et al. Reference Lipsitz, Trost, Grossmann and Sides2005). Additionally, research finds that citizens judge personal attacks as less valuable than substantive attacks because they distract from the important issues at stake (Brooks and Geer Reference Brooks and Geer2007). Furthermore, within the negative campaigning literature, extensive research examines the impact of distinct types of conflicts on citizens’ political attitudes and behavior (Brooks and Geer Reference Brooks and Geer2007; Carraro and Castelli Reference Carraro and Castelli2010; Hopmann, Vliegenthart, and Maier Reference Hopmann, Vliegenthart and Maier2018; Min Reference Min2004; Nai and Maier Reference Nai and Maier2021). Findings suggest that personal attacks tend to disengage citizens from politics (Min Reference Min2004) and evoke dislike for the party resorting to character attacks (Carraro and Castelli Reference Carraro and Castelli2010; Hopmann, Vliegenthart, and Maier Reference Hopmann, Vliegenthart and Maier2018; Nai and Maier Reference Nai and Maier2021). Taken together, prior scholarship suggests a general aversion to personal conflict among citizens, while substantive conflicts may be judged less harshly.
Additionally, scholars differentiate conflicts based on the depth of the disagreement among politicians (Druckman, Peterson, and Slothuus Reference Druckman, Peterson and Slothuus2013; Skytte Reference Skytte2021). For one, politicians can find themselves in a conflict where they (strongly) disagree about an isolated issue but still agree on the core moral values surrounding the issue. By contrast, politicians can have more fundamental disagreements in which they do not share core moral values and thus hold clashing worldviews (De Ridder Reference De Ridder, Hannon and Edenberg2021; Ranalli and Lagewaard Reference Ranalli and Lagewaard2022). As discussed before, conflicts can be about broader objectives or how to pursue these objectives with concrete measures (Skoog and Karlsson Reference Skoog and Karlsson2018). Following this view, we consider conflicts arising from different policy positions or clashing ideological perspectives as distinct phenomena. An example of a substantive conflict is a disagreement over how to reduce climate change impacts, such as whether to increase taxes on flights or prohibit short flights. Within this type of conflict, politicians agree on the core moral value that is necessary to reduce the impact of climate change but differ on how to achieve this. In an ideological conflict, politicians hold clashing moral values, for instance, whether the life of an unborn child should be protected or whether women should have the choice of abortion.
Previous research on the impact of ideological polarization on citizens produced mixed findings. Ideological polarization refers to divergences in political beliefs and opinions between political actors in relation to fundamental ideological principles. While Banda and Kirkland (Reference Banda and Kirkland2018) find that ideological conflict causes distrust in politics, van Elsas and Fiselier (Reference van Elsas and Fiselier2023) report that it increases support for democracy. On the one hand, ideological conflict may create dissatisfaction among the public. Strong disagreement risks being perceived as fostering legislative gridlock and thereby limiting effective governance (Hibbing and Theiss-Morse Reference Hibbing and Theiss-Morse2002; Uslaner Reference Uslaner2015). On the other hand, ideological conflict may be appreciated by citizens as it functions as a signal of the quality of political representation (van Elsas and Fiselier Reference van Elsas and Fiselier2023). Citizens often match their own beliefs to those of parties based on their ideological beliefs (Levendusky Reference Levendusky2010). If politicians stick to their ideological beliefs and refuse to make any concessions regarding their core principles, citizens may potentially value it. As a result, citizens may feel their vote matters, thereby increasing their satisfaction with the functioning of representative democracy (Kittilson and Anderson Reference Kittilson and Anderson2011). Against this backdrop, we hypothesize that personal conflict will be the least valued by citizens, while conflict arising from ideological differences will be the most valued.
H2: Citizens are more likely to approve of ideological conflict as opposed to substantive and personal conflict (a) and are more likely to approve of substantive conflict compared to personal conflict (b).
While little research has assessed politicians’ perspectives towards different kinds of conflicts, it seems likely to assume that politicians’ evaluations follow the same pattern as for citizens, and we expect that politicians are less in favor of personal conflicts compared to substantive conflicts. First, from an office and policy-seeking perspective, personal conflicts may be seen as pointless strategic conflicts, staged to garner media attention. Due to the news media’s preference for individual politicians, their personal lives, and conflict (Vliegenthart, Boomgaarden, and Boumans Reference Vliegenthart, Boomgaarden, Boumans, Brants and Voltmer2011), the non-substantive aspects of political conflict often gain attention in political coverage (Bartholomé, Lecheler, and de Vreese Reference Bartholomé, Lecheler and de Vreese2018). A recent interview study with politicians indeed discloses that politicians prefer to stay away from personal conflicts, as they add little substance to the debate (van der Goot, van der Meer, Hameleers et al. Reference van der Goot, van der Meer, Hameleers and Vliegenthart2025). At the same time, from a vote-seeking perspective, personal conflicts can also be beneficial for politicians. Simply put, if it helps them to attract media attention, they will also be more visible to citizens. However, being critical of the opponents’ character is a risky strategy that can backfire (Nai and Maier Reference Nai and Maier2021; Walter and van der Eijk Reference Walter and van der Eijk2019). If citizens find the criticism unfounded, politicians risk damaging their own reputation instead of that of their opponents. Hence, if politicians expect citizens to dislike personal conflict, there is also less benefit in them.
Second, in addition to anticipating what citizens may (dis)like, politicians may also more inherently be against personal conflicts. These types of conflicts often take the form of uncivil behavior that limits the possibility of respectful democratic debate (Stryker, Conway, and Danielson Reference Stryker, Conway and Danielson2016). By being uncivil or refusing to collaborate with an opponent due to their personality, political actors essentially deny their opponent’s legitimacy and right to participate in politics (Massaro and Stryker Reference Massaro and Stryker2012). This goes against the core ideals of deliberative democracy: that politicians should engage in thoughtful and respectful debate, that everyone has the right to participate, and that politicians should – at least in multiparty systems – search for compromise and common ground (Habermas Reference Habermas2006; Polletta and Gardner Reference Polletta and Gardner2018). Hence, politicians may value conflicts about personal relationships to a lesser extent than conflicts about different issue positions.
In exploring differences in perceptions of ideological and substantive conflict, we anticipate politicians to favor ideological conflicts, like citizens. This may be because politicians are less likely to make concessions on issues that are core to their party’s ideology (Lefevere, Tresch, and Walgrave Reference Lefevere, Tresch and Walgrave2015). According to Gutmann and Thompson (Reference Gutmann and Thompson2010), the ‘spirit of permanent campaigning’ in politics has caused an ‘uncompromising mindset’ for politicians (1128). Politicians can mobilize citizens by highlighting their uncompromised commitment to core values and differentiating themselves clearly from the positions of the opponents (Gutmann and Thompson Reference Gutmann and Thompson2010). The ideological alternatives that parties formulate help citizens to align their preferences to those of parties (Sniderman and Bullock Reference Sniderman, Bullock, Saris and Sniderman2004). Ergo, politicians owe it to their political supporters to adhere to these ideological core values, as a perceived deviation may cause distrust. While, for a democracy to function well, concessions and compromise are essential, there are also limits to what parties can compromise on, and politicians are also expected to remain loyal to the ideological core of the political party (Andeweg and Irwin Reference Andeweg and Irwin2005). At the same time, we also witness politicians voting against their party line (e.g., Slapin, Kirkland, Lazzaro et al. Reference Slapin, Kirkland, Lazzaro, Leslie and O’grady2018). While politicians are expected to remain loyal to their party, they also have their own policy stances, which they may prioritize, and this could be electorally rewarding (Bundi, Varone, Loewen et al. Reference Bundi, Varone, Loewen and Breunig2023).
H3: Politicians are more likely to approve of ideological conflict as opposed to substantive conflict and personal conflict (a) and are more likely to approve of substantive conflict compared to personal conflict (b).
Politicians’ perceptions of citizens
What should we expect that politicians think about citizens’ preference for conflict? While direct evidence is lacking, the literature presents arguments that result in opposite expectations. On the one hand, even if politicians themselves are more favorably predisposed towards conflict (see H1 above), it seems reasonable to assume that they should still report that citizens have an overall lower tolerance for it relative to them. First, politicians have a strong motivation to positively portray their constituents, and if a preference for conflict is perceived as a negative valence trait (Hibbing and Theiss-Morse Reference Hibbing and Theiss-Morse2002), then, all else being equal, politicians should be reluctant to assign this preference to citizens. Moreover, politicians – who are acutely aware of declining trust in them and in elected institutions more broadly (e.g., Loat and MacMillan Reference Loat and MacMillan2015) – may genuinely believe that citizens are averse to conflict because it is inherent to the functioning of the same partisan politics that they dislike.
On the other hand, when given the opportunity, politicians show a strong preference for selecting extreme, conflict-prone candidates (Broockman, Carnes, Crowder-Meyer et al. Reference Broockman, Carnes, Crowder-Meyer and Skovron2021), suggesting that they see them as electorally appealing and that citizens appreciate such conduct. Furthermore, recent work shows that politicians are likely to project their views on their assessments of the public’s policy preferences (Pereira Reference Pereira2021; Sevenans, Walgrave, Jansen et al. Reference Sevenans, Walgrave, Jansen, Soontjens, Bailer, Brack, Breunig, Helfer, Loewen, Pilet, Sheffer, Varone and Vliegenthart2023), something that they may also be doing when considering the public’s preference for political conflict. This could lead them to erroneously ascribe stronger support for conflict among citizens than there exists. Together, there is no clear consensus in the literature regarding how politicians perceive citizens’ overall support for conflict. However, because we expect politicians themselves to be far less approving of personal conflict than substantive or ideological conflict, it means that even if they engage in projection, they should in any case judge citizens’ approval of personal conflict to be low – potentially lower than it actually is. Hence, we pose the following hypothesis.
H4: Politicians are more likely to assess citizens’ judgment of the conflict as negative in a situation of personal conflict as opposed to substantive and ideological conflict.
The role of conflict for (partisans from) radical party families
The appreciation of conflict may depend on citizens’ party preferences or, in the case of politicians, party affiliation. We expect citizens who vote for a radical right or radical left party, and politicians from these radical parties, to be more likely to accept engagement in conflicts. Radical right parties are here understood as parties that adopt nativism, authoritarianism, and populism as key elements, while radical left parties are considered those that challenge the dominant socioeconomic structure while advocating for an anticapitalistic system (Rooduijn, Pirro, Halikiopoulou et al. Reference Rooduijn, Pirro, Halikiopoulou, Froio, Van Kessel, De Lange, Mudde and Taggart2024).
Our central argument is that, by definition, citizens and politicians of radical parties hold more extreme positions than the supporters and representatives from centrist parties and often are ‘challenger’ parties that are outsiders. Thus, those who hold more radical views tend to deviate more from the views of others, making conflict inevitable when engaging with other parties. Especially for populist parties, which define mainstream parties as the elitist outgroup and hold antagonistic views on the people and elites (Mudde Reference Mudde2004), conflict is almost a raison d’être. If politicians from a radical party join a governing coalition, they probably have to make larger concessions on their policy preferences than more centrist parties, as political compromises are frequently required in consensus, non-majoritarian democracies. If radical parties belong to the parliamentary opposition, they have even stronger incentives to trigger conflicts with governing parties. Whether a radical party is in power or not, the mere fact that its electoral pledges and policy preferences are more extreme than the policy positions advocated by centrist or non-radical parties constitutes, thus, a source of conflict (Kriesi and Schulte-Cloos Reference Kriesi and Schulte-Cloos2020). As citizens from radical parties expect their representatives to defend these extreme positions, even if this is done in a confrontational way, their approval of all types of conflict is higher.
H5: (a) Citizens who support a radical party family (left or right) and (b) politicians who belong to a radical party family (left or right) are more likely to approve all types of conflict than other citizens.
Furthermore, we expect citizens who vote for radical parties and politicians from radical parties to be more likely to appreciate personal and ideological conflict than substantive conflicts. At least three factors might explain why (partisans of) radical parties are expected to support more personal or ideological conflicts. First, since politicians from radical party families are more vote- and office-seeking than policy-seeking and aimed at short-term gains (Gómez and Ramiro Reference Gomez and Ramiro2022; Valentim Reference Valentim2021), they are more likely to adapt to media logic (Esser and Strömbäck Reference Esser and Strömbäck2014; Landerer Reference Landerer2013; Mazzoleni and Schulz Reference Mazzoleni and Schulz1999; Strömbäck and van Aelst Reference Strömbäck and Van Aelst2013) for more attention and visibility. Since ideological and personal conflicts, rather than substantive conflicts, relate more closely to news value criteria such as proximity, conflict, drama, and personalization (Mazzoleni and Schulz Reference Mazzoleni and Schulz1999, 250), these forms of conflict might be valued more. Second, radical parties frequently adopt a populist communication style (Jagers and Walgrave Reference Jagers and Walgrave2007) and delegitimizing rhetoric against opponents (Levitsky and Ziblatt Reference Levitsky and Ziblatt2018), in which they emphasize proximity and sovereignty of the (pure) people and adopt an anti-establishment stance against (corrupt) elites. Populist strategies seem to be more aligned with ideological and personal conflicts, as they tend to contain more negativity towards opponents with character attacks (Nai Reference Nai2021) and aim to reinforce the ideological predispositions of their in-group citizens (i.e., us) to blame some out-groups (i.e., them; Otjes and Louwerse Reference Otjes and Louwerse2014). In addition, a populist communication style is frequently embodied by a charismatic leader (Mudde Reference Mudde2007) who becomes a strong identification factor for citizens and often a target of critical voices and personal conflicts (McDonnell Reference McDonnell2017; Michel, Garzia, Ferreira da Silva et al. Reference Michel, Garzia, Ferreira da Silva and De Angelis2020).
Third, the electoral rise of radical parties coincides with political polarization. Affective polarization can occur at both the elite and citizen levels, which is characterized by more personal dislike of out-group partisans (Iyengar, Lelkes, Levendusky et al. Reference Iyengar, Lelkes, Levendusky, Malhotra and Westwood2019; Torcal and Harteveld Reference Torcal and Harteveld2024). This increase in disliking and distrusting those from the other party is related to increased political elites’ support for extremist politicians (Iyengar, Lelkes, Levendusky et al. Reference Iyengar, Lelkes, Levendusky, Malhotra and Westwood2019, 142). There is no strong and definitive empirical evidence yet on the relationships between affective polarization, ideological conflicts, and policy disagreements (Bougher Reference Bougher2017; Rogowski and Sutherland Reference Rogowski and Sutherland2016). However, it is plausible that affective polarization exacerbates hostility and conflicts between the supporters of radical parties and their out-partisans, and that both politicians and citizens of radical parties (right- or left-wing) approve of personalized conflicts involving their most extreme representatives. This logic resonates with scholarship on political tribalism (Klein Reference Klein2020; Levendusky Reference Levendusky2010), which demonstrates how sociocultural sorting along dimensions such as geography, religion, education, and identity fosters sharper in-group/out-group boundaries. As partisan identities increasingly overlap with social and cultural divisions, conflicts are no longer understood as policy disagreements but as struggles between fundamentally opposed ‘tribes’. This alignment of political and sociocultural cleavages magnifies affective polarization and makes both elites and citizens in radical parties more likely to approve of ideological and personal conflicts.
H6 Footnote 3:(a) Citizens who support a radical party family (left or right) and (b) politicians who belong to a radical party family (left or right) are more likely to approve of ideological and personal conflict compared to substantive conflict.
Method
Sample
Our study relies on (1) a survey experiment among a representative sample of citizens in four countries (Francophone Belgium, the Netherlands, Portugal, and Switzerland; n = 8264) and (2) a survey experiment with political elites in the same countries (n = 331). This study has been preregistered on Open Science Framework (OSF).Footnote 4 The citizen and elite surveys are part of the broader POLPOP II project, which collected data in thirteen countries (Australia, Canada, Czech Republic, Denmark, Belgium, Germany, Israel, Luxembourg, the Netherlands, Norway, Portugal, Sweden, and Switzerland), aiming to study how political elites perceive public opinion and the congruence between politicians’ and citizens’ views.Footnote 5 The survey included several general questions such as left-right placement and more specific modules on, amongst others, political conflict, media bias, and the state of democracy. Not all modules have been asked in all countries. The questions about political conflict have only been asked in four countries, so not all countries are included in our sample selection. Nonetheless, we ensured a mixture between countries that have a distinct party landscape and different ways to deal with political conflict, make political decisions, and forge agreements. Our sample includes three countries that have a clear tradition focusing on consensus and consociationalism (Lijphart Reference Lijphart1999), namely Francophone Belgium, the Netherlands, and Switzerland. Consociationalism refers to a model of governance to manage a deeply divided society via implementing elements of power sharing among different segments or groups in society. This creates distinct cases in which conflicts are handled through different institutional logics. In contrast, Portugal has a more bipolar party system, where two main political blocs (a left and right one) usually face each other. Thus, by focusing on these four different countries, we capture significant variation in how conflicts unfold within parliamentary settings. Also, our countries differ in terms of polarization amongst citizens and elites. For instance, the Netherlands is one of the least affectively polarized party systems in Europe, while Portugal has a relatively high level of affective polarization, and Switzerland is situated between these two countries (Reiljan Reference Reiljan2020). All countries, especially Belgium, the Netherlands, and Switzerland, are characterized by strong right-wing populist challenger parties (Rooduijn, Pirro, Halikiopoulou et al. Reference Rooduijn, Pirro, Halikiopoulou, Froio, Van Kessel, De Lange, Mudde and Taggart2023), who are known for their aggressive and fierce attacks (Nai Reference Nai2021). By including these four countries, with their varied political contexts, we aim to deepen our understanding of how conflicts are perceived in multi-party systems, especially in environments where both consensus and adversarial dynamics exist.
First, the citizen survey was fielded in collaboration with Dynata between 7 February 2022 and 4 May 2022, although some variation exists between the participating countries (see overview in Table A1 in the Online Supplementary Materials). The sample aimed to represent the population of the four countries (in terms of age, gender, education, and geographical location). All data was estimated without using weights. The panel was administered online. All common quality check controls were carried out to limit undesirable response behavior, such as speeding and straightlining. Respondents were rewarded for their participation. In total, 11,108 citizens were initially invited for all four countries, of which 8264 citizens were included in the final sample. A lot of respondents were excluded as they did not meet the necessary conditions (such as informed consent, showing undesirable response behavior, and being younger than 18 years old; n = 2369). Several citizens were also excluded because they did not answer all the questions about conflict (n = 463) or due to missing data on the control variables (n = 12). The final sample sizes for the individual countries are as follows: 2119 for Belgium, 1865 for the Netherlands, 2114 for Portugal, and 2166 for Switzerland.
The elite survey aimed to question all Members of Parliament (MPs) in the four included countries; for Francophone Belgium, both regional and national/federal MPs were included. Ethical clearance for the elite survey was obtained in each country by its respective university.Footnote 6 All MPs were initially contacted via a formal e-mail sent by the principal investigator on the project, and after one week the MPs were contacted by a mix of phone calls, e-mails, and other contacts (e.g., personal connections). Reminder e-mails were sent out. In total, 842 MPs were eligible for participation, of which eventually 359 MPs participated, resulting in an average response rate of 42%. The interviews with the politicians were conducted either face-to-face or online (via Zoom) by a researcher or research assistants who were thoroughly trained in interviewing (see Online Supplementary Files A, Table A2, for the timing of the fieldwork). The setup was similar online and for the face-to-face interviews, as politicians participated in a 30-minute interview with closed questions. In some situations, this was followed by a list of semi-structured interviews. Still, there were differences in the response rate across countries (see Online Supplementary Files A, Table A2). We excluded 24 politicians because they did not answer our conflict questions; additionally, we have missing data for four politicians on their party family. Due to anonymity reasons, we cannot trace every politician back to their party. The final sample therefore holds 331 MPs: 143 for Belgium, 35 for the Netherlands, 61 for Portugal, and 92 for Switzerland. More details on the representativity of the sample, including in terms of gender, age, seniority, and ideology, can be found in the Online Supplementary Files A in Tables A3 and A4.
Procedure
In the survey, citizens and politicians were presented with three different hypothetical conflict scenarios, combining within- and between-subject designs. Insofar as we are interested in how politicians and citizens perceive conflict, we focus on a common way in which disagreement becomes visible in parliament, that is, parties disagreeing about a policy proposal and voting against it (see also Fortunato Reference Fortunato2012). In the context of consensus democracies, this type of conflict scenario is a common aspect of parliamentary debates, which makes politicians more likely to support it. However, at the same time, voting against a bill can also gridlock the legislative process, and the reasons for doing so are therefore of crucial importance. We asked representatives and citizens to imagine a situation in which a hypothetical party (i.e., ‘party A’) introduced a bill containing new measures regarding a policy issue (i.e., immigration, environment, health care, or civil rights). We only referred to the issue and did not include any details about specific bills. Subsequently, we provided three reasons for why another hypothetical party (i.e., ‘party B’) does not support the bill, including (1) it does not offer the best solution to the problem according to party B (i.e., substantive conflict); (2) it goes against the ideological core (values) of party B (i.e., ideological conflict); and (3) the leader of party A has a personal problem with the leader of party B (i.e., personal conflict). For each reason, we asked participants to indicate the extent to which they approve or disapprove of the reasons on a 7-point scale (1 = completely disapprove; 7 = completely approve). The different scenarios (options 1/2/3) reflect a within-subject design. All participants were presented with the three options. The four policy issues were randomized among participants. Participants only saw one of the four policy issues. This is, thus, a between-subject design. For politicians, an additional question was posed after the conflict scenarios. They were asked to indicate to what extent they thought citizens judged the three different situations of conflict positively or negatively. Randomization was performed per country, and we used block randomization. Each participant was randomly assigned to one of four predetermined blocks, which presented all three conflict scenarios with one policy issue.
Dependent variables
Approval of conflict. We asked both politicians (M = 4.13, SD = 2.22) and citizens (M = 3.75, SD = 1.76) to indicate on a 7-point scale if they approved of the separate conflict scenario (1 being ‘completely disapprove’, 4 being ‘neither approve nor disapprove’, and 7 ‘completely approve’).
Perceived judgments of conflict by citizens . We also asked politicians to indicate whether citizens judged the three conflict situations negatively or positively (1 being ‘very negatively’, 4 being ‘neither negatively, nor positively’, and 7 being ‘very positively’; M = 3.72, SD = 2.00).
Independent variables
(Partisans from) the radical party family. We asked citizens which party they would vote for if elections were held today. The preferred party of citizens was grouped under party families (Conservatives, Greens, Liberals, Radical Left, Radical Right, Social Democrats, Independents, Special Issue Party, Other) based on the classification of the Chapel Hill Expert Survey (CHES; Rovny, Bakker, Hooghe et al. Reference Rovny, Bakker, Hooghe, Jolly, Marks, Polk, Rovny, Steenbergen and Vachudova2025) and the Political Party Database Project (PPDB; Scarrow, Webb, and Poguntke Reference Scarrow, Webb and Poguntke2022). We constructed a dummy variable for whether the participant voted for a radical left or right party (1) or not (0; M = 0.25, SD = 0.44). There are several missing variables on this question for the citizens (N = 1895) because citizens either did not respond to the question of who they would vote for (item nonresponse), did not vote, or voted blank or invalid; moreover, some of the parties could also not be reliably assigned to a party family.Footnote 7 We constructed a similar measure for the politicians.Footnote 8 To construct this measure, we relied on their party affiliation and used the same classification as for the citizens to group them under a party family (M = 0.11, SD = 0.32).
Ideological extremity. Given the absence of information on party support among a significant portion of citizens, we adopted an alternative operationalization to measure the degree of citizens’ radical political preferences. Respondents were asked to position themselves on a left-to-right scale, with 0 denoting ‘left-wing’ and 10 indicating ‘right-wing’. We standardized this variable and used the absolute values for the analysis. Consequently, individuals with higher scores on this variable are situated at a higher distance from the mean, signifying a greater degree of ideological extremity, either leaning towards the radical left or right ends of the political spectrum (M = 0.74, SD = 0.67). This same question was posed to politicians, allowing for the construction of a comparable variable to assess their ideological positioning (M = 0.93, SD = 0.60).
Control variables
Since the main aim of this paper is to compare citizens’ and politicians’ perceptions of conflict, we include control variables for which we have comparable measures across the two samples. Preferably, we would have liked to have included more control variables, but due to pseudonymization of the politician’s data,Footnote 9 several background variables are missing, such as age, which limits the number of control variables we can consider. Nonetheless, we do include some key control variables for which we have theoretical expectations that they may influence the results. First, we control for the type of policy issue the conflict revolves around. We choose issues that are more complex (health, environment) and less complex (civil rights, immigration) and that differ based on whether they are valence (civil rights, health) or positional issues (immigration, environment). Valence issues refer to issues on which there tends to be consensus about the goal (i.e., it is desirable to protect civil rights and have a good health care system), while positional issues include issues on which political parties tend to be strongly divided (Green Reference Green2007). It is important to control for these different issues, as perceptions of the appropriateness of conflict may matter depending on the issue at hand. Second, we control for whether people would vote for a government party if elections were held today (M = 0.34; SE = 0.47) or whether politicians belong to a government party (M = 0.68; SE = 0.46). Given that the primary function of opposition parties is to criticize the government (Russmann Reference Russmann2017) and that coalition partners tend to adhere closely to their agreements (Vercesi Reference Vercesi2016), we expect government parties to be less likely to approve of conflict. Third, we account for the ideological leanings of citizens and politicians, determining whether they align more with the left or right (measured on a 10-point scale, with higher values denoting more right-wing leaning); citizens (M = 5.06; SE = 2.40) and politicians (M = 3.93; SE = 2.42). Finally, we aimed to control for gender (citizens: M = 0.51; SE = 0.50; politicians: M = 0.42; SE = 0.49), as women are expected to value conflict less than men (Wolak Reference Wolak2020), and female politicians are, for instance, found to use less confrontational language in parliaments (Hargrave and Langengen Reference Hargrave and Langengen2021). However, since we have quite some missing data on this variable for politicians (n = 23), we do not control for gender in the main analyses and instead conduct some robustness checks.
Analyses
First, we have transformed our data so that each row in our dataset represents a combination of respondent x conflict situations (=times three, because of the substantive, ideological, and personal conflict situations). As a result, our data has a nested structure: each conflict scenario is nested within respondents (citizens and politicians), and respondents are nested within countries. We deal with this nested structure by including fixed effects in our models at the country level, making sure that the hierarchical nature of the data is addressed. Because we control in our model for the four different policy domains and since respondents saw only one policy domain, we cannot include fixed effects at the participant level. However, to test the robustness of our findings, we also conduct the analyses separately for each country, without fixed effects on the country level, and then we do include clustered standard errors for the participants.Footnote 10 Importantly, to test H5a and H6a, we conducted separate analyses, one with ‘(voted for) radical party (or not)’ as an independent variable and one for ‘ideological extremity’. We specified in the pre-registration that we will consider the effect of ideological extremity. However, since we can construct this measure in different ways, we want to explore if results remain similar when considering different operationalizations. Hence, we also consider if someone voted for a radical party family. We do the same for politicians (H5b; H6b) by considering the effect of whether they belong to a radical party family or not and to what extent they are ideologically extreme.
Results
First, we investigated whether politicians and citizens differ in their approval of the conflict scenarios. Overall, we found that politicians are more likely to approve of the different conflict scenarios than citizens (b = 0.39, SE = 0.05, p < 0.001, 95% CI [0.28, 0.49]), controlling for the different conflict types, the policy domains, left-right placement, and being a member or voting for a government party (see Online Supplementary Information File B, Table B1). These findings hold across the countries when tested in a separate country model and when controlling for gender (see Online Supplementary Information File B, Tables B2 and B3). This confirms H1; when we do not differentiate between types of conflicts, politicians are, in general, more supportive of conflict. Second, we investigated how politicians and citizens evaluated the different conflict scenarios. Figure 1 plots the marginal effects of the different scenarios on approval of the conflict, separated by citizens and politicians (see also Online Supplementary Information File B, Table B4). As Figure 1 shows, both citizens and politicians are more likely to approve of ideological as opposed to personal conflict. However, we also expected that both citizens and politicians were more likely to approve of ideological conflict compared to substantive conflict, but we found the opposite. In contrast to our expectations (H2a and H3a), citizens and politicians are more likely to approve of substantive than ideological conflict. A possible explanation for his counterintuitive finding may be that politicians and citizens value conflicts around substantive issues more as they offer an opportunity to address problems. In contrast, ideological conflicts are more abstract and may be perceived as causing gridlock.
Figure 1. Approval of the conflict scenarios by citizens and politicians and politicians’ perceptions of citizens’ approval.
We do, however, confirm hypotheses H2b and H3b. Both citizens and politicians are more likely to approve of substantive compared to personal conflict. Most results hold across the different countries (see Online Supplementary Information File B, Table B6, and Table B7). Only in the Netherlands do we not find a significant difference in the evaluation of substantive and ideological conflict for the politicians. However, this could be due to the small sample size (N = 35), which may limit statistical power. We also explored to what extent the approval of the different conflict types was similar for politicians and citizens. We conducted a linear regression with fixed effects, in which we included an interaction between the type of conflict (reference category: ideological conflict) and citizens versus politicians (see Online Supplementary File B, Table B5). Figure 2 presents the marginal effects with 95% confidence intervals.Footnote 11 We find that politicians are more approving of substantive and ideological conflict than citizens. However, in the case of personal conflict, they are less approving (b = −2.20, SE = 0.13, p < 0.001, 95% CI [−2.45, −1.95]). This finding holds across the four countries (see Online Supplementary File B, Table B5). We find that politicians value personal conflict less than citizens.
Figure 2. Interaction between approval of the conflict scenarios x citizens versus politicians.
Furthermore, we assessed how politicians thought that citizens would judge the different conflict types. Figure 1 shows that politicians are more likely to assess citizens’ judgment of the conflict as positive in the situation of an ideological and substantive conflict as opposed to a personal conflict, which offers support for H4. We find the same result across the four countries (see Supplementary Information File C, Table C1). Interestingly, politicians tend to assess the approval of ideological and substantive conflicts higher than citizens indicated and tend to assess the approval of personal conflicts lower (see Figure 1). Politicians’ perceptions of citizens’ approval seem to mimic politicians’ approval, and they thus seem to project their perceptions on citizens.
Turning to our hypotheses on party family, we explored whether citizens who support a radical party family (left or right) are more likely to approve of conflict than other citizens. Table 1 presents the results. We do not find any significant differences between citizens who support a radical party compared to citizens who support another party family. Hence, we cannot confirm H5a. We also do not find any interaction effects between the different types of conflicts and party family (H6a; see Table 1). However, importantly, we have a lot of missing values on this variable; almost a fourth of our sample did not fill out the party they voted for (n = 1895).Footnote 12 We also explored ideological extremity (as a proxy for radical party support) and found that the more ideologically extreme citizens are, the more likely they are to approve of all types of conflict. We find this effect in the Netherlands, Portugal, and Wallonia but not in Switzerland when controlling for gender (see Supplementary Information File D, Tables D1 and D3). Hence, with a different measure of preference for ideological parties, we can partially confirm H5a. Regarding H6a, we find an interaction effect for personal conflict and ideological extremity, but not for ideological conflict and ideological extremity. This means that the more ideologically extreme, the more likely you are to approve of the personal conflict scenario, which partly confirms H6a.
Table 1. Citizens’ approval of conflict
Note: The sample of citizens used for the analyses is smaller than the total sample of citizens. There are 1895 missing values for belonging to a radical party family and only 12 for ideological extremity. Standard errors in parentheses.
*p < 0.05, **p < 0.01, ***p < 0.001.
For politicians, we also do not find that politicians who belong to a radical party family (left or right) are more likely to approve of all types of conflict than other politicians, in contrast to H5b. However, in Portugal and Switzerland, we do find a significant difference. Politicians in Portugal (b = 0.88, SE = 0.40, p < 0.05) and Switzerland (b = 0.65, SE = 0.31, p < 0.05) from a radical party family are more likely to approve of the conflict scenarios.
Furthermore, results show that the effect of belonging to a radical party family on conflict approval differs for different types of conflict (H6b). As Table 2 shows, politicians from more radical party families are more likely to approve of ideological and personal conflict than politicians from other party families. We also considered the ideological extremity of politicians. We do not find that the more ideologically extreme, the more likely to approve of conflict, again refuting H5b. Except for the Netherlands, where we do find that the more ideologically extreme politicians are, the more they approve of conflict (b = 0.38, SE = 0.17, p < 0.05). Furthermore, we find an interaction effect with ideological conflict (H6b). While more ideologically extreme politicians are not more likely to approve of personal conflict, they are more likely to approve of ideological conflict, again partly confirming H6b. These findings hold when controlling for the gender of the politicians (see Supplementary Information File D, Table D4).
Table 2. Politicians’ approval of conflict
Note: Standard errors in parentheses.
*p < 0.05, ***p < 0.001.
Finally, while we did not discuss hypotheses about the type of policy, it is interesting that we did not find any differences in the approval of conflict across policy domains for the politicians (see Table 2), but we did find differences for citizens (see Table 1). Citizens are less likely to approve of conflict in the case of a health care or environmental policy issue compared to immigration. This may confirm the idea that for politicians, conflict is part of the game in general, regardless of policy domains, while citizens find it more worrisome when it comes to certain policies.
Discussion
This article examines the perceptions of citizens and politicians towards different types of conflict and the harmony between these perceptions. Overall, our findings show that there seems to be a misalignment between the perceptions of politicians and citizens on the role of conflict in politics. In general, politicians are more positive about conflict than citizens. When considering the types of conflict, we find that politicians are especially more approving of substantive and ideological conflict than citizens. Surprisingly, politicians are more critical of personal conflict than citizens. We also find this misalignment when considering politicians’ estimation of citizens’ perceptions: politicians’ assessment of citizens’ approval of conflict seems to align more with politicians’ own perceptions than with those of citizens.
Politicians’ more positive attitude towards substantive and ideological conflict compared to citizens’ is in line with expectations. Politicians potentially appreciate conflict more because they are accustomed to the important role that conflict serves within the political system, and refusing compromise helps politicians stay close to their own priorities and ideological beliefs. It also helps them keep their promises to citizens and be responsive to their needs (Mansbridge Reference Mansbridge2003). Instead, for citizens, ideological and substantive conflict may in some cases also be perceived as an obstacle to effective governing (Hibbing and Theiss-Morse Reference Hibbing and Theiss-Morse2002), especially when it concerns certain issues such as healthcare, causing them to disapprove of conflict more.
When we zoom in on personal conflict, our findings seem to be somewhat paradoxical. Politicians demonstrate greater aversion to personalistic conflict than citizens do but expect citizens to share their level of aversion – again projecting their own preferences onto the citizenry. At the same time, however, we witness politics becoming increasingly personalized, with individual political actors becoming more central to politics at the expense of political parties (Rahat and Kenig Reference Rahat and Kenig2018). This is reflected, for instance, in political news coverage (Vliegenthart, Boomgaarden, and Boumans Reference Vliegenthart, Boomgaarden, Boumans, Brants and Voltmer2011) or politicians’ online communication (Metz, Kruikemeier, and Lecheler Reference Metz, Kruikemeier and Lecheler2020). Consequently, competition and conflict in politics increasingly revolve around candidates instead of parties or ideologies (Rahat and Kenig Reference Rahat and Kenig2018), as research on the prominence of personal attacks in parliamentary debates suggests (Waddle, Bull, and Böhnke Reference Waddle, Bull and Böhnke2019). Furthermore, studies show that attacks, including character assassinations, are more likely to attain media attention (Haselmayer, Meyer, and Wagner Reference Haselmayer, Meyer and Wagner2019). Hence, it seems surprising that politicians are so negative toward personal conflict, especially considering that it may help them attract media coverage. Perhaps their answer reflects a social desirability bias. Personal attacks are seen as a form of uncivil behavior that adds little substance to the political debate (Brooks and Geer Reference Brooks and Geer2007) and is found to be detrimental to democracy by reducing political trust and increasing affective polarization (Skytte Reference Skytte2021). While politicians may see the strategic value of personal attacks in parliament, they simultaneously want to object to them because they bring about negative effects and citizens dislike them. In that sense, it is interesting that we found that citizens like personal conflict more than politicians do.
We also considered the alignment more specifically between supporters and politicians from radical party families. While we find that citizens who are more ideologically extreme are more likely to approve of conflict in general, more ideologically extreme politicians or politicians affiliated with radical parties are not. This is surprising since these parties seem to be inherently conflictual, and research consistently finds that radical parties are more likely to seek out conflict or attack opponents during election campaigns (e.g., Nai Reference Nai2021; van der Goot, Kruikemeier, de Ridder et al. Reference van der Goot, Kruikemeier, de Ridder and Vliegenthart2024). Yet, we do find that elected politicians affiliated with radical parties are more likely to approve of certain types of conflict, this being ideological and personal conflict. Similarly, more ideologically extreme citizens also value personal conflict more. Hence, when it comes to personal conflicts, the preferences of radical politicians and their voters seem to align more.
Overall, the misalignment between the conflict preferences of citizens and politicians may be worrisome. Particularly considering that politicians’ assessment of what citizens think should be the role of conflict in parliament was a lot more positive than citizens’ actual perceptions on ideological and substantive conflicts. This aligns with recent research that investigates politicians’ assessment of public opinion and shows that politicians tend to assume that citizens hold the same preferences as they do and thus project their preferences on others (Pereira Reference Pereira2021; Sevenans, Walgrave, Jansen et al. Reference Sevenans, Walgrave, Jansen, Soontjens, Bailer, Brack, Breunig, Helfer, Loewen, Pilet, Sheffer, Varone and Vliegenthart2023). This projection can have important consequences for political representation. If politics becomes increasingly about political strife, both within parliament, news coverage, and in the online communication of politicians, because politicians fail to recognize that citizens dislike conflict, it could contribute to citizens’ declining trust in politics and satisfaction with democracy. Especially, perhaps, when politics becomes more polarized on identity grounds, this may fuel citizens’ distrust (Skytte Reference Skytte2021).
Our study does not come without limitations. First, our sample of politicians is not equal across the countries; for instance, in the Netherlands, only 38 politicians participated, and our sample of politicians deviated slightly from the population in terms of ideology. Right-wing politicians were somewhat underrepresented in our data, and especially the number of radical right-wing and radical left-wing politicians was not substantial (N = 17 and N = 21, respectively). While our results do not show differences between MPs belonging to a radical party family or not, it could be that this is due to the sample size. Thus, we encourage future research to investigate the perceptions of conflict for radical parties and their alignment with their voter base.
Second, the findings regarding the relative importance of substantive conflict compared to ideological conflict may be the result of how substantive conflict is phrased in the experiment. To operationalize substantive conflict, we mention that a party opposed a bill because, according to them, it did not offer the best solution to the problem. In so doing, we aimed to present respondents with a conflict situation in which the policy positions of a political party do not align with those of another party. However, it also may have signaled to respondents that politicians disagree due to technocratic reasons and that the bill would create a suboptimal situation if it got passed. This underscores the necessity of the party voting against it and thereby also increases citizens’ and politicians’ approval of this decision. Unfortunately, we are unable to assess how citizens and politicians perceive the conflict types. Nonetheless, since the core of our study is also to assess whether citizens’ and politicians’ preferences overlap, we do believe our results provide sufficient insight into this.
Third, we investigated the perceptions of hypothetical conflict scenarios, which makes it challenging to generalize our findings to real-world conflicts. Although our scenarios are based on the types of conflict covered in the news and present in politicians’ online communication (van der Goot, Kruikemeier, de Ridder et al. Reference van der Goot, Kruikemeier, de Ridder and Vliegenthart2024), we omit information about the involved parties and their standpoints. Following motivated reasoning theory (Kunda Reference Kunda1990), reactions are likely to be stronger when the conflict involves a favored party or aligns with people’s beliefs. Hence, we believe that our study provides a conservative test of conflict evaluations, and we encourage future research to investigate perceptions towards real conflict.
Overall, given that we found differences in the approval of conflict types and some preliminary evidence that conflicts over different policies are evaluated differently, it would be worthwhile to further explore perceptions of conflict in politics and how they relate to citizens’ satisfaction with democracy. Moreover, while our analysis centers on party-family differences, we acknowledge that individual dispositions held by politicians, such as pragmatism, responsiveness, or negotiation style, may also condition politicians’ willingness to engage in distinct types of conflict, a promising avenue for future research.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S1475676526100966.
Data availability statement
The data will be made publicly available after the ending of the POLPOPII project, a transnational collaboration initiated by Stefaan Walgrave. The underlying code for the analyses can be found on OSF.
Acknowledgements
An earlier version of this paper was presented at the Nordic Political Science Congress (Bergen, Norway, 25–28 June 2024). We are grateful to all participants for their feedback, especially Sona Golder, Raimondas Ibenskas, Oliver Kannenberg, and Zeynep Somer-Topcu.
Funding statement
This work was supported by the European Research Council [grant number: 101018105]. Open access funding provided by University of Amsterdam.
Competing interests
No potential conflict of interest was reported by the author(s).
Ethical standards
Ethical approval was obtained from the University of Amsterdam for the Dutch data collection (2022-PCJ-1477), from the University of Geneva for the Swiss data (CUREG-2021-10-10), from the University of Lisbon for the Portuguese data (07/2022), and from the University of Bruxelles for the Wallonia data (R2022/004).
Open access
