The problem with polarization research in an age of extremism

1 Introduction

“I have been gravely disappointed with the white moderate. I have almost reached the regrettable conclusion that the Negro's great stumbling block in his stride toward freedom is not the White Citizen's Counciler or the Ku Klux Klanner, but the white moderate, who is more devoted to “order” than to justice; who prefers a negative peace which is the absence of tension to a positive peace which is the presence of justice”

—Martin Luther King, Jr., Letter from Birmingham Jail (1963)

Over the past decade, political polarization has become one of the most studied topics in the behavioral and social sciences. A search for “political polarization” in the Web of Science database (conducted on 23 October 2025) yielded 13,387 results, and more than 60 per cent of which were published since 2020. During the same period, authoritarian and anti-democratic movements have gained strength across diverse contexts. A Friedrich-Ebert-Stiftung (2023) study reported that support for a dictatorship led by a “strong leader” in Germany increased, and endorsement of social-Darwinist statements such as “some lives are worthless” more than doubled compared with previous years. Similarly, the Leipzig Authoritarianism Study (Decker et al., 2024) reported renewed increases in xenophobia, chauvinism, and the downplaying of Nazism. In Germany, national reports indicate that extremist offenses have increased year after year (Reuters, 2024). Similar patterns are also observable in the United States and Brazil where not only national surveys reveal an increased openness to military coups and one-party rule (Pew Research Center, 2024a,b) but also extremists have attempted to overturn electoral outcomes by storming their respective congresses (European Parliamentary Research Service, 2023).

Authoritarian, conservative, and reactionary networks have been among the principal drivers of the surge in distrust, conspiracism, and anti-scientific sentiment (Backhaus et al., 2023; Benegal et al., 2022; Cologna et al., 2025; Dakin et al., 2023; Pavlović et al., 2022; Philipp-Müller et al., 2022; Van Bavel et al., 2022; van der Linden et al., 2020; Vlasceanu et al., 2024; Voelkel et al., 2024). Coordinated reactionary movements couple conspiratorial and antiscientific narratives as a central element of their political strategy, which, in the recent decades, has been successful in undermining pluralist opponents, galvanizing reactionary followers, shifting blame away from authoritarian sources, and weakening democratic values and institutions (Ren et al., 2022). Through national and international political campaigns, these networks have attempted to replace societal trust and factual knowledge with alternative belief systems that frame them both as victims of ideological attacks and as opponents of a corrupt establishment against the “will of the people” (Golder, 2016; Mudde, 2007; Rothmund et al., 2020; Van Hauwaert et al., 2018, 2020). By portraying democracy and science as corrupt, elitist, or ideologically captured establishments, these actors erode public trust in civic society, empirical inquiry, and in the institutions that sustain both (van der Meer, 2026). Moreover, in framing pluralist democracy and scientific expertise as nefarious, they reframe their quest for political power and dominance as a project of “renewal” and “social progress”.¹ This rather paradoxical—and populist (Golder, 2016; Mudde, 2007)—framing has been a potent tool spreading and legitimizing hostility toward pluralist democracy and its knowledge-based institutions. Opposition to democracy, science, and institutional expertise is therefore not an accidental feature of anti-democratic populism and authoritarianism, but its defining component (Capraro et al., 2026; Levitsky and Ziblatt, 2019; Sandel, 2022; Snyder, 2021; Stanley, 2020).

These developments illustrate that contemporary polarization does not occur in a vacuum. Rather, it is embedded within a wider ideological shift toward authoritarianism marked by the normalization of violence, systematic dis/misinformation campaigns, and hostility toward scientific expertise (Azevedo and Jost, 2021; Roozenbeek and van der Linden, 2024; van der Linden, 2023; Van Dijck et al., 2018; Vilanova et al., 2025). Against this backdrop, the widespread assumption of the field that polarization is inherently detrimental warrants re-examination. Notably, most empirical studies conceptualize polarization either in ideological terms—typically as increasing divergence or clustering of citizens' attitudes along an ideological spectrum (Philipp-Müller et al., 2022)—or in affective terms, defined as intensified animus toward political outgroups coupled with ingroup favoritism (Iyengar et al., 2012). While analytically tractable, this approach is nonetheless conceptually naïve and methodologically too coarse for the current political moment. Polarization can reflect (i) agonistic democratic conflict—a constitutive feature of pluralist democracy in which adversaries contest values and policies while accepting democratic norms, rules, and outcomes (Mouffe, 2000)—or (ii) democratically corrosive conflict, in which political identity is asymmetrically organized and strategically exploited into distrustful camps and wielded to justify rule-breaking, delegitimizing of elections and pluralist institutions, and erosion of accountability (Somer and McCoy, 2019). Rather than reflecting a symmetric breakdown of civility, this form of dissension constitutes a well-documented, elite-driven pathway to democratic backsliding (Bermeo, 2016).

This framing constitutes a tacit belief that social harmony and dialogue are the ideal state of democratic life, and political antagonism is a failure of civility. In the current moment of rising extremism, not all polarization is equivalent nor undesirable. Historical experience underscores this asymmetry. Normative appeals to civility and harmony, when applied indiscriminately, can obscure the moral distinction between domination and resistance—condemning confrontational opposition even in contexts where democratic survival or basic human dignity is at stake. Under conditions of systematic persecution, conflict is not a pathology of democracy but one of the few remaining mechanisms through which injustice can be resisted. Moreover, opposition between democratic and authoritarian values cannot be treated as symmetrical disagreement. When one side rejects pluralism, electoral legitimacy, or epistemic accountability, then “reconciliation” and even “civility” function less as neutral civic virtues than as asymmetric constraints, and they risk unilaterally disarming pro-democracy forces. Indeed, democratically corrosive conflict primarily disciplines democratic defenders while leaving anti-democratic actors free to escalate, delegitimize, and capture institutions. Under such asymmetric conditions, appeals to reduce polarization risk equating resistance to authoritarianism with intolerance.

The assumption that polarization must be reduced carries profound epistemic and political consequences for social and political psychology and political science. It shapes research designs, variable operationalizations, and the normative interpretations of findings. Interventions aimed at reducing affective polarization, for example, often promote empathy and perspective-taking across ideological divides without considering whether the underlying attitudes sustain or undermine the very conditions that allow for democratic discourse and evidence-based reasoning (e.g., Gillissen et al., 2025; Santos et al., 2022; Tausch et al., 2024; Voelkel et al., 2024; Warner et al., 2020). When authoritarian and anti-scientific movements are gaining strength, such naïve depolarization efforts risk demobilizing both scientific communities and democratic publics at precisely the moment when active defense of evidence and reasoned inquiry is most needed. Safeguarding the norms that make civil society and empirical inquiry possible requires an explicit commitment to democratic rules, pluralism, and epistemic standards (e.g., electoral legitimacy, institutional accountability, and evidence-constrained claims-making). We therefore propose a distinction between democratic polarization—agonistic contestation among actors who accept multicultural pluralism, democratic institutions and election outcomes, civil and human rights, and epistemic accountability—and anti-democratic polarization, in which conflict is strategically mobilized to delegitimize opponents, erode institutional constraints, and normalize dehumanization, scapegoating, misinformation, anti-scientific, and conspiratorial narratives as a route to political power. To clarify this distinction, we introduce a minimal framework specifying when political conflict remains within democratic bounds and when it begins to undermine them. This clarification is necessary because much of the polarization literature treats all forms of conflict as analytically comparable, regardless of their implications for democratic functioning.

1.1 Conceptualizing democratic conflict

Under contemporary conditions of authoritarian resurgence, disinformation, and institutional erosion, the meaning of democratic conflict cannot be inferred from intensity, incivility, or ideological distance alone. Research treating polarization as presumptively pathological may misclassify democratic resistance as equivalent to anti-democratic escalation. Indeed, democratic politics is not reducible to civility, consensus, or the absence of conflict. In pluralistic societies, conflict is often unavoidable and sometimes necessary. Political disagreement commonly concerns collective priorities, the allocation of public resources, and the design of institutions that structure social life. But democratic disagreement does not extend without limit. It unfolds within a democratic framework that must secure equal standing, basic rights, and meaningful inclusion for all members of the polity. Polarization researchers must contextualize and interpret political conflict in light of the political reality, including stakeholders' commitments to pluralism, equal citizenship, collective welfare, equitable social and economic opportunities, and the maintenance of democratic institutions.

This point is especially important because the boundary between “ordinary” policy disagreement and anti-democratic politics cannot be drawn by separating social and economic arrangements from questions of rights, recognition, and membership. Social and economic arrangements are not external to democratic inclusion: they help determine who has security, health, mobility, voice, and genuine access to collective life. When such arrangements systematically expose some groups to deprivation, precarity, premature death, or political marginalization while insulating dominant groups and concentrated economic power, they cease to be ordinary policy disagreements in any normatively innocent sense. They become mechanisms through which hierarchy is organized and reproduced. For this reason, conflicts over rights, recognition, and membership cannot be cleanly separated from conflicts over economic life, since the latter shape whether democratic citizenship is substantive or merely formal. When such inequalities are re-described as routine political disagreement, domination is recast as policy and anti-democratic projects can pass as normal democratic contestation.

Accordingly, not all politically salient disagreements are democratically equivalent. Some conflicts occur within democratic bounds, even when they are intense and adversarial. Others target the very conditions that make democratic life possible. When political actors seek to deny full standing to opponents, restrict who counts as a legitimate member of the polity, erode rights protections, weaponize institutions, or degrade the epistemic conditions of public reasoning, the issue is no longer conflict within democracy alone. It is conflict over democracy's own normative and institutional basis. Under such conditions, domination can be redescribed as policy, exclusion can be redescribed as disagreement, and anti-democratic projects can pass as routine democratic contestation.

To tentatively point out this distinction, Table 1 provides a heuristic framework for evaluating when political conflict remains consistent with democratic functioning and when it does not. It is not intended as a rigid classificatory scheme, nor does it aim to resolve ambiguities inherent to socio-political realities. Rather, it identifies a set of dimensions along which political conflict may be evaluated. The purpose of Table 1 is to make explicit why polarization cannot be evaluated solely in terms of distance, hostility, or civility, as it is often the case in the published literature. What should matter most, in our view, is the nature and kind of conflict being mobilized, by whom, under what institutional conditions, and toward what ends.

Table 1. Indicators for evaluating whether political conflict is agonistic or democratically corrosive.

This framework indicates that polarization cannot be adequately evaluated solely in terms of attitudinal distance, hostility, or affect. By distinguishing between agonistic and corrosive forms of conflict, it helps account for why comparable levels of polarization may have different implications across contexts. This distinction, however, introduces challenges for classification and measurement. Simpler models may offer greater clarity, parsimony, and cross-context comparability, but these advantages often involve the omission of distinctions that are normatively and politically relevant. As a result, analytically distinct forms of conflict may be treated as equivalent. These limitations are not only methodological but also epistemic, as they relate to how analyses incorporate asymmetries of power, institutional conditions, and the processes through which political meanings and attitudes are formed. Engaging with these dimensions complicates measurement and interpretation, yet their exclusion does not eliminate their influence. Instead, it may lead to accounts that are descriptively tractable but conceptually incomplete, in which forms of domination are interpreted as ordinary disagreement and anti-democratic mobilization is analyzed as equivalent to democratic contestation.

2 Discussion

The central problem is not polarization itself, but the tendency of parts of polarization research to treat conflict as presumptively pathological while insufficiently distinguishing democratically productive contestation from democratically corrosive mobilization under conditions of structural, institutional, and epistemic asymmetries. Thus, it is not that efforts toward depolarization are inherently misguided, nor that all research on polarization treats polarization as uniformly undesirable. As noted in the review by Kreiss and McGregor (2024), a substantial portion of the polarization literature does adopt this assumption. Our goal is to challenge the framing of polarization as the primary problem of democracy, suggesting instead that polarization may be understood as a necessary byproduct of the struggle to realize democratic principles, proposing that the status of polarization as a social good is normatively conditional. Efforts to reduce conflict presuppose that antagonism reflects symmetric disagreement among actors who share commitments to pluralism, democratic rules, and epistemic accountability. When this assumption does not hold—when conflict is corrosive and thus reflects domination, exclusion, or the strategic erosion of democratic norms—calls for empathy, reconciliation, or civility risk reproducing the very asymmetries they seek to resolve. In such contexts, urging those targeted by discrimination or authoritarian mobilization to empathize with their aggressors mirrors a familiar moral inversion: it relocates responsibility for harmony onto the vulnerable rather than the powerful. What appears as neutrality thereby becomes complicity in the depoliticization of injustice. This is why the evaluative task for polarization research cannot be limited to measuring disagreement or animus, but must also assess whether the conflict in question is organized around inclusion, accountability, and collective welfare, or around domination, exclusion, and democratic erosion.

This dynamic also reflects a broader disciplinary discomfort with explicit engagement in politics, ideology, and power. Across parts of the behavioral and social sciences, phenomena closely tied to political elite strategy, organized countermobilization, propaganda infrastructures, race and ethnicity research, or structural and power injustices are sometimes treated as normatively contaminated objects—too political, professionally risky, or epistemically improper for direct analytic confrontation. While such caution may be individually understandable, its collective effect can be epistemologically consequential. Moreover, many defend that neutrality in social and behavioral sciences is a methodologically virtuous goal to which we scientists should aspire. We defend, however, that nothing could be farther from the truth: the cumulative effect of this discomfort, of this depoliticization or neutrality of science is—and has been—epistemically distorting and detrimental to the quality of science produced. Indeed, it is the very observation of these negative consequences that motivated the writing of this piece. When the political production of meanings, narratives, or identities is bracketed in the name of neutrality, social and behavioral science risks rendering unintelligible the very forces that generate the attitudes and conflicts it seeks to explain. In this sense, neutrality does not merely describe a stance toward evidence; it can function as an analytic refusal to model inequities, power, and human suffering. Where elite-level political production is constitutive of the phenomenon under study, such refusal threatens the interpretability of analysis and interpretation itself and, by extension, the capacity of research to engage with the phenomena being studied and illuminate democratic erosion. The challenge, therefore, is not that social science becomes too political, but that in avoiding politics it risks becoming descriptively inadequate to political reality.

From this perspective, the task of polarization research is not simply to reduce antagonism but rather to determine when conflict constitutes democratic erosion and when it represents democratic resistance – an effort which we have tentatively done here. This distinction does not deny that polarization—even among democratic actors—can in some contexts erode institutional trust, increase political violence, or impair collective problem-solving. Rather, the present argument is that normative evaluations of polarization must be conditioned on actors' relationship to democratic rules and epistemic accountability. Accordingly, instead of focusing primarily on depolarizing society, research could investigate how the denial of conflict contributes to authoritarian consolidation. This calls for a shift in focus, from measuring attitudinal divergence or negative affect toward political outgroups to analyzing how different audiences respond to messages that promote consensus. Researchers could, for instance, trace how appeals to “unity” and “civility” circulate across political discourse and media narratives, and assess how such appeals influence the willingness of citizens to recognize or challenge the political actors advancing them. This dynamic parallels scholarship on respectability politics (Harris, 2014; Jefferson, 2023), which shows how moralized expectations of composure, civility, or propriety are often imposed asymmetrically on marginalized groups, thereby disciplining dissent and obscuring underlying structures of domination. Bringing these literatures into dialogue may clarify how consensus-oriented norms can function less as neutral democratic virtues than as mechanisms of political regulation. Similarly, experimental and longitudinal studies could test how exposure to harmony-oriented framings shapes perceptions of legitimacy and dissent, while cross-cultural research might investigate how these effects vary across contexts with distinct social norms and historical cleavages. By treating consensus-seeking itself as an empirical object, polarization research could clarify when and how conflict functions as a source of democratic resilience rather than erosion. Crucially, this requires classifying polarization by actors' relationship to democratic rules and epistemic norms—for example, willingness to accept electoral losses, support for institutional checks, endorsement of political violence, and tolerance for misinformation as “strategy”—so that interventions do not inadvertently equate democratic resistance with authoritarian delegitimization (Bermeo, 2016).

History shows that authoritarian movements flourish under conditions of depoliticization, portraying their dominance as the necessary route to social renewal. In this context, the scientific enterprise must resist becoming a depoliticizing force. At a time of escalating attacks on evidence, equality, and the scientific method itself, research should strengthen the capacity of society to recognize and confront genuine threats to democratic and epistemic integrity. Yet much of the current literature on political polarization moves in the opposite direction, framing conflict as a phenomenon to be avoided regardless of its context. What endangers democracy—and the autonomy of science—is not polarization, but the insistence that confrontation with authoritarianism can or should be transcended through consensus. The simultaneous rise of polarization research and authoritarian pressures should therefore prompt critical reflection on the assumptions guiding our analytical frameworks and on how they might be reoriented to meet the challenges of the present.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. AI was used for proofread.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Cengiz Erisen, Yeditepe University, Türkiye

Reviewed by: Yelda Yılmaz, Istanbul Sabahattin Zaim University, Türkiye

Yasemin Tosun, Sabanci University, Türkiye

Footnotes

Quotation marks highlight the rhetorical inversion through which populist and radical-right movements appropriate the language of progress to legitimize projects of democratic erosion. In this context, appeals to “renewal” and “social progress” often signify not emancipatory transformation but reactionary restoration, exclusionary nationalism, and the consolidation of hierarchical power relations under the symbolic cover of reform (Levitsky and Ziblatt, 2019; Mudde, 2007; Stanley, 2020). Hence, “renewal” and “social progress” operate as legitimizing frames and dog-whistles for a political agenda and an ideological project that are inherently reactionary, exclusionary, anti-scientific, anti-pluralist, and anti-democratic. Similar to the will of the people, these are known to be rhetorical vehicles that invert the conventional moral valence of progressive, restorative, and social justice terminology.