1 Introduction

The COVID-19 crisis has brought the problem of social integration back into the limelight. When the first wave hit Europe and North America in the spring of 2020, the pandemic initially served as evidence for a sociological theorem that promised grounds for optimism: facing a shared challenge and fighting a common “enemy”—even one that measures only 100 nm—will turn even diverse, conflict-ridden societies into cohesive groups that are capable of collective action.Footnote 1 A year later, during the fourth wave of the pandemic, public and sociological discourse has become noticeably more pessimistic. It is not the cohesion-generating potential of common threats that is now widely debated, but the equally time-honoured thesis that crises begets conflict and conflict begets division.Footnote 2 Societies already divided by divergent interests, incompatible ideals and diverse identities are, according to this narrative, inherently unstable (Putnam 2007; Deenen 2019) and will be stretched to their breaking point, when conflicts over resources, rights and recognition intensify owing to external shocks. From this perspective, the pandemic appears as a catalyst of societal conflicts that are ultimately caused by the fragmented character of modern, liberal societies.Footnote 3

If this analysis is correct, it does not bode well for the fate of liberal democracies: if modern societies have become irrevocably pluralized, then how do they manage to create a form of social integration that makes collective will formation and the maintenance of common institutions possible? How can societies provide public goods and preserve mutually accepted norms, when traditional certainties, religious bonds and shared cultural values seem to evaporate under the pressure of pluralisation? These concerns about the integration of democratic societies are not merely theoretical. In many Western liberal societies, the division of society along political and cultural identities has intensified, whereas trust in democratic institutions declined. In the USA, the results of democratic elections are no longer readily accepted by large sections of the losing party’s supporters, whereas populist parties dedicated to fighting social plurality are on the rise in many European countries. Democratic societies, or so it seems, are increasingly at risk of falling apart owing to their internal differences. Against this background of a glooming crisis of democracy, one might be tempted to seek refuge in the idea of suppressing conflicts emerging from plurality. Another response to this challenge, as we will suggest in this article, is to embrace plurality and—in consequence—conflict, not only as inevitable components of any free society but also as a potent source of social integration. We argue that modern democratic societies have the capacity to create and sustain social integration by nurturing and containing conflict at the same time. Democracies can promote social integration through conflict by employing three discernible mechanisms: by diluting more concentrated and threatening conflicts with more specific and limited forms of conflict, by embedding actors in institutional procedures that structure conflicts and channel their energy and by enabling their members to adapt and appropriate social norms and institutions through conflict.

To develop this argument, we will first sketch out the basic socialising function of conflict before we deal more concretely with the question of how conflict can generate or strengthen social integration (Sect. 2). These integration-promoting or, as we would like to shorten it in the following, productive effects of conflict are, however, always precarious. Not only do all conflicts tend to escalate, most of them are also not solvable. They require institutions and procedures that nurture conflict but at the same time are able to contain it to such an extent that conflict between opponents (understood as disputes between at least two individuals or groups) does not degenerate into a fight between enemies. Combining insights both from sociology and social psychology of conflict as well as democratic theory, we argue that modern democracies in particular have developed a number of normative as well as rational institutional mechanisms that allow them to dilute, embed and nurture conflicts in order to both control and harness their energy (Sect. 3). However, this balance between containing and nurturing conflict is never stable. Institutions tend to inertia over time and thus give preference to stabilisation over change (Mansbridge et al. 2010; Deitelhoff 2012), whereas external shocks such as the pandemic or the emerging climate crisis may ignite conflicts that cannot be accommodated by existing institutions. At the end of this article, we will therefore outline three recent developments, each of which threatens one of the mechanisms safeguarding the productive functions of conflict: polarization, radicalization and depoliticization (Sect. 4).

2 The Link Between Conflict and Social Integration

In order to examine whether conflict is a possible source of social integration, we must first briefly explain how we understand the concept of integration in the context of this article. We want to propose a relatively lean conception that defines social integration as the acceptance of and conformity with a social order by its members. In line with the introduction to this special issue (Grunow et al. this issue), this definition does not presuppose a thick normative consensus, generalised trust or shared identities, but merely a general disposition of actors (acceptance) to observe established rules (conformity) that enables social coordination and cooperation. This disposition to accept social rules can be based on normative or rational reasons and it may relate primarily to first-order rules that regulate behaviour, or to second-order rules that provide the membership conditions of the social order and the institutional framework for setting, interpreting and changing first-order rules (for this distinction see Holm 1995; Fehmel 2014; Schmelzle 2020).

These two variables give rise to four ideal types of sources of integration. Acceptance of the social order may be normatively grounded in (a) belief in the substantive rightness of first-order rules or (b) the legitimacy of second-order rules, or it may arise from a rational choice perspective if (c) the first-order rules currently in force or (d) the continuation of the current second-order institutional framework are considered to be in the best interests of the actors. In diverse, democratic societies, sources that depend on the concrete content of the first-order rules—sources (a) and (c)—are likely to be less important than those that relate to the acceptance of the second-order procedural rules (Schmelzle 2015; Schmelzle and Stollenwerk 2018). This is so because content-dependent reasons can secure social order only if the substance of the rules in question is either in (nearly) everyone’s interest or if substantial moral convictions converge perfectly. Both should be extremely rare in liberal societies, where people are free to develop diverse moral outlooks and ideas of the good (Rawls 1993). The remaining reasons (b) and (d) seem more promising under these conditions. The belief in the legitimacy of democratic procedures is possible from a variety of ideological positions and does not require a strong convergence of substantial moral convictions (Tyler 2006). Similarly, a rational interest in the continued existence of an established social order does not require substantial homogeneity of interests. It is likely to develop whenever there is such a high degree of functional interdependence within that order that the realisation of any life plans is tied to its long-term stability.

Given this definition, the notion that conflicts could be a source of social integration initially has something counterintuitive about it. If we understand social integration as the acceptance of and conformity with a social order, then conflict, understood as situations in which the attitudes and behaviour of two or more social entities towards each other are altered by their perception that their goals are mutually incompatible seems rather to threaten than to secure the status quo.Footnote 4 For one thing, conflicts are an indicator that the current distribution of “resources, rights and recognition” (Offe 1998, p. 119) is not met with universal approval. They show that the corresponding interests, ideologies and identities of the conflicting parties are, at least currently, mutually exclusive. For another, it seems that conflicts are also causes of social disintegration in themselves: They change the emotional and cognitive attitudes between the parties, challenge the validity of shared norms and institutions, and entail a considerable risk of escalating into intergroup violence. Polarisation, anomie or civil war are terms typically associated with conflict, not social integration and cohesion.

We believe that this sceptical perspective towards the integrative power of conflict is misguided for two reasons: first, it ignores the productive potential of conflicts, which has been accentuated in the sociology of conflict by authors such as Simmel (1964), Coser (1956, 1957) or Dahrendorf (1972). In this tradition conflict is seen as a force that, under the right conditions, has the potential to create and stabilise social bonds rather than one that necessarily impedes and threatens them. Its proponents argue that even hostile conflict creates social relations by compelling the actors to relate to one another and thus perceive themselves and others as part of a common social world. Once social relations are established, conflicts reveal the actors’ interests and mutual normative expectations, which in turn enables them to amend the first- and second-order rules of the social order to meet their needs. Conflict thus contributes to social integration by (1) constituting a shared social space and (2) providing the means of shaping it. According to this view, social orders can be stabilised by a series of necessarily provisional agreements about the conditions of co-existence that can and will be regularly contested and updated through conflicts (see also Clasen 2019, pp. 61, 236).

Second, the exclusive focus on disintegrative effects of conflict not only paints an unbalanced picture of the social impact of conflict but also an increasingly irrelevant one: given the inescapable plurality of modern societies, the communitarian complaint about the loss of shared values and increasing social discord seems outdated. Moreover, the longing for a society of shared values and identities also seems normatively problematic, as current levels of diversity could only be reduced through the oppressive use of state power (Rawls 1993, pp. 36–37). That is why conflicts are an unavoidable part of any open society. The integration of societies like ours can therefore only succeed if we learn to use and regulate conflicts in a socially productive way.

Before we tackle this problem directly in the third section of this article, we first take a closer look in this section at the mechanisms that generate the socially integrative effect of conflicts, both at the intra- and the intergroup level. Only then can we examine in the next section the institutional and structural conditions under which they can flourish.

2.1 Intragroup Integration Through Conflict

When considering the integrative effect of conflicts, it is first helpful to distinguish between two levels at which conflicts can potentially generate cohesion: within the groups involved in the conflict and between the conflicting parties. Although in this paper we are primarily concerned with the latter phenomenon, intergroup cohesion, we will first take a look at the problem of intragroup cohesion. This detour is useful for two reasons: first, because we can observe mechanisms that are also relevant for integration at the intergroup level and, second, because intragroup cohesion resulting from conflicts is a central problem for the formation of intergroup cohesion.

When social groups enter into conflict with one another, the confrontation between them usually increases the social integration within the respective groups (Simmel 1964). A number of interlocked mechanisms are responsible for this integrative function of conflicts at the intragroup level: first of all, conflicts help to create groups by causing agents to adopt group identities. For many actors who objectively belong to the same social category—students, the working class, those living near an airport—conflicts establish a subjective awareness of group membership or at least reinforce it (Coser 1957; Dahrendorf 1972). In conflict with the respective antagonist—the professoriate, capital, regulating authorities—those affected assume a we-identity as a member of their group and adopt its goals and values. This transition from “latent” to “manifest” group membership (Dahrendorf 1972) then has a considerable impact on the behaviour of group members—both towards each other and to the outgroup, as social identity theory posits (Tajfel et al. 1971; Tajfel and Turner 1979).

Identification with a group generates a desire in its members to raise the group’s status. Tajfel and Turner explain this pattern by a basic need of individuals to maintain a positive self-image. Through identification with a group, the self-esteem of group members is linked to the (relative) success of the group. This need is reflected in the behavioural dispositions of group members. They develop an ingroup bias, i.e. the tendency to perceive group members favourably and to treat them preferentially at the expense of the outgroup. In doing so, group members do not primarily aim for the absolute improvement of their group’s welfare, but rather for a relative improvement vis-à-vis the outgroup, even if this implies a net loss of resources for the ingroup. In Tajfel’s experiments, participants regularly chose to allocate rewards in such a way that the difference between members of their group and the outgroup was largest rather than choosing an alternative distribution that would have yielded a higher but more equal payoff to members of both groups (Tajfel et al. 1971). Groups that perceive themselves to be in competition with other groups or that see their existence threatened by other groups are more likely to form central authority structures and to sanction deviant behaviour by group members (Sherif 1966; Benard 2012). This suggests that conflicts have in themselves a remarkable potential for generating group cohesion, at least in part irrespective of the underlying reasons for conflict.

This short overview of the effects of conflict on group cohesion at the intragroup level suggests initial pessimism about the integrative potential of conflict at the intergroup level: conflict reinforces identification with the ingroup, group identity leads to status competition between groups, and status competition tends to produce group conflicts—at least under certain conditions. Or, to paraphrase Charles Tilly (1975, p. 42), it seems that conflict makes groups and groups make conflict. Does this not suggest that conflicts rather contribute to the disintegration of societies than to their integration?

2.2 Intergroup Integration Through Conflict

A closer look at the impact of conflict on intra-group cohesion thus seems to suggest a new version of the old argument that conflict is divisive. According to this version, conflict increases social cohesion, but at the wrong level: it strengthens group identities, ingroup bias and internal group structures, which is disintegrative for society as a whole because it leads to segregation and polarisation. Against this argument we want to present three mechanisms that make an integrative effect of conflicts at the level of the whole society plausible; despite or even because of their integrative effect at the intra-group level. The analysis of these mechanisms sets the stage for the discussion in the next section, where we will discuss how democracies deal with conflict.

Before we present the individual mechanisms, it is first important to specify what is meant by integration at the intergroup level. As indicated above, we understand social integration as the acceptance of and conformity with a system of norms. Thus, the integrative effect of conflicts does not primarily refer to the relationship between rival groups—for example, between labour and capital, students and professors, people living near airports and airport authorities—but to the effect that conflicts have on the general acceptance of the social order by its (individual) members. Accordingly, it would be at least conceivable that hard-fought collective bargaining conflicts may strain the relationship between employers and employees but have an overall integrative effect in that the process or outcome of the conflict increases the net acceptance of the social order as a whole. The question is then, through what mechanisms can conflicts be allowed to contribute to the acceptance of the social order that frames them?

(1) The first mechanism is conflict dilution through the pluralisation of conflicts. Conflicts are particularly risky for social integration when they lead to a polarisation of society into two opposing blocs that are superimposed onto all other lines of social conflict (as Coser already observed in 1956). Such polarised conflicts are fatal because they turn every social dispute into zero-sum games between parties competing for resources, rights and recognition, in which one side’s loss is the other’s gain. This turns into a problem of social integration when the parties are no longer willing to cooperate with each other and to accept the second-order institutional rules that distribute political power and constrain its use.Footnote 5

How can such totalising conflicts be prevented? The idea is here that in polarised conflicts, the central problem is not too much conflict but too little: the foundations of society are endangered precisely because two cohesive conflict parties have formed along a single axis, competing with each other on all issues of social distribution (Coser 1956; Schattschneider 1960; Lipset and Rokkan 1967). If this front line gets diluted by additional conflicts limited in scope and cross-cutting to the former, the danger to the social order as a whole is defused as well. First, because new cleavages bring new cross-cutting alliances with their own internal cohesion with them and increase the chance of minorities to participate; second, because the actors no longer attach such importance to any individual conflict and social identity that they would be prepared to sacrifice the social order as a whole to it (see also Goodin 1975).

(2) The second mechanism—conflict embedding—promotes the productivity of conflict, not by manipulating the number and patterns of conflicts and conflicting groups but by embedding them into an institutional framework. This approach has a normative and functionalist dimension. In the normative dimension, conflicts are embedded in a common group identity by being accepted and possibly even encouraged rather than tabooed and repressed. By providing institutional frameworks for engaging in conflicts, societies recognise, at least implicitly, that the expression of different interests, ideals and identities is a legitimate part of the role of citizens. From this perspective, political activism, protests and public debates are not seen as signs of political disintegration but as necessary contributions to the common democratic project and as valuable as such (Deitelhoff 2013; Volk 2021, p. 455). This openness to dissent and debate gives critical citizens a reason to accept the authority of second-order social institutions, even if they disagree with many first-order rules.

If citizens understand democratic institutions as a legitimate arena for the joint struggle over issues of public order and the common good, then this is an important contribution to the integration of society. But second-order beliefs in the legitimacy of political institutions will not suffice to guarantee social integration in the long run if they become incompatible with the self-interest of the actors involved. The normative embedding of conflicts in the self-conception of democratic societies must therefore be supported by mechanisms that allow democratic institutions to structure and specify conflicts, channel their energy and to bind conflicting groups to the social order through their self-interest (Przeworski 2019, ch. 9). This is the functionalist side of institutional embedding, which will be explored in the next section as well.

(3) The third mechanism refers to the adaptation and appropriation of social norms through conflict. The integrative function of conflict here derives from its ability to adjust social norms to the needs of members of society, either by reforming or by transforming them. Lewis Coser in particular has pointed out this function (Coser 1956, 1957). We distinguish between norm adaptation and norm appropriation. Norm adaptation refers to the, often reluctant, adjustment of norms from above in order to discourage (further) political change and to contain conflict—think of Bismarck’s social security legislation, for example. We speak of norm appropriation when actors without legislative authority change the meaning of existing norms or establish new norms through their behaviour from below. This mechanism highlights the function of conflicts to give rise to political alternatives and social innovations in social orders (Deitelhoff and Thiel 2014, p. 441).

These three mechanisms seek to describe how integration through conflict can work in any social order: Through the dilution of destructive conflict via the fragmentation of conflict axes; through the embedding of conflict in an institutional framework that enables, legitimises and regulates conflict; and through the adaptation and appropriation of social norms that occurs when conflicts are anticipated or allowed to play out. In the following section, we examine more specifically how modern democracies in particular can take advantage of these integrative functions of conflict to enable social integration without shared pre-political values and identities.

3 Nurturing and Containing Conflicts: A Democratic Dilemma

Theorists of radical democracy such as Chantal Mouffe (2000, 2018)Footnote 6 or authors in the republican tradition, such as Helmut Dubiel (2008), argue that the focus on conflict and the management of conflict is precisely the feature that distinguishes democracy from other political systems. According to these approaches, it is not consensus on common norms and institutions that lies at the heart of democracy but the freedom, equal for all, to contest such a consensus. As Dubiel once put it: “Democratic societies are not maintained by conflicting groups sacrificing their particular interests and opinions to an imaginary consensus. Rather, the normative capital that integrates them arises precisely in the chain of conflicts that are fought out according to rules” (Dubiel 2008, p. 666, our translation). Understood this way, democracies are structural entities that have developed precisely such rules or institutional arrangements in order to be able to draw integration from conflict, namely by linking their own continued existence to the freedom of citizens to contest this. However, the question then arises as to what institutional arrangements democracies can and should use to support the dilution, embedding but also harnessing of conflict, without risking the escalation of conflict? How do they solve the dilemma to nurture conflict while containing it at the same time, to reap its productive effects?Footnote 7

3.1 Conflict Dilution in Democracies

The mechanism of conflict dilution is based on the idea that the disintegrative force of each conflict is likely to decrease if the axes of conflict diversify, thereby preventing bipolar confrontation and generating internal cohesion within multiple diverse and overlapping conflict coalitions at the same time. This strategy thus has a negative and a positive integration mechanism: integration is achieved, on the one hand, by preventing the polarisation of society along one dominant axis or cleavage—such as that between labour and capital or along ethnic or sectarian divides—and, on the other hand, by the experience of sharing interests and group identities with different allies in different conflicts. The success of both mechanisms, however, depends on the emergence and persistence of cross-cutting cleavages and the corresponding diverse coalitions. Democracies traditionally engage in conflict dilution by directly creating forums for the expression of conflicts, such as in labour relations, or indirectly through electoral law,Footnote 8 through the expansion of the welfare state (Esping-Anderson 1996; Nullmeier 2000), and individual rights to protest and assembly that enable citizens from various backgrounds to engage in politics.

Of course, one could argue that such strategies have become increasingly irrelevant in modern, diverse democracies. Given the “fact of pluralism” (Rawls 1993, p. 36), it would seem natural that a multitude of conflict axes would arise by themselves as a result of an open society (see Pappi 2002; Fehmel 2014). However, as the increasing levels of political polarisation in the USA, the UK and some European states demonstrate (Gidron et al. 2020), this is clearly not an irreversible process. In these societies, political identities—e.g. as Republican and Democrat, Leaver or Remainer—have become so salient, distinct and encompassing that they threaten to mute all other societal cleavages and group identities or bring them into alignment altogether. In both cases, the division of society along political identities goes along with an increase in affective polarisation, that is, “sympathy towards partisan in-groups and antagonism towards partisan out-groups” (Wagner 2021, p. 1). Affective polarisation threatens to divide society into opposing political camps that are integrated less by a common ideology or shared interests than by an “us-vs.-them” mentality that is fuelled by the desire to defeat their opponents (Mason 2018). Members of the outgroup are then no longer perceived as persons with different values and interests with whom one has to come to terms but as morally deficient characters whose positions have to be fought.

Comparative polarisation research has recently identified three types of conditions that seem to favour the affective polarisation of societies: (1) economic inequality and periods of high unemployment, (2) the politisation of cultural conflicts by political elites, and (3) structural factors such as a majoritarian electoral system that favours the emergence of a two-party system (Gidron et al. 2020).

(1) Gidron et al. find that economic inequality and high unemployment both correlate significantly with an increase in affective polarisation (2020, p. 65). They do not, however, offer a clear theoretical explanation for this. It seems to us that the status-based model for explaining intergroup conflict offered by social identity theory could be helpful here. Members of groups that perceive their status as threatened—a category that can include middle- and high-status groups as well—have three strategies at their disposal to maintain a positive image of themselves: changing their individual group membership (social mobility), changing the standard or group of comparison (social creativity) or changing the social relations that determine group status, i.e. first- and second-order rules that distribute power and resources between groups (social competition). Whereas social mobility and social creativity both evade group conflict, social competition tends to increase both ingroup favouritism and hostility towards the outgroup (Tajfel and Turner 1979). Now, mass unemployment and pronounced inequality are conditions that make social status differences particularly visible and thus generate status stress (Groh-Samberg et al. 2014, 2018) and render individual social mobility and creativity particularly difficult. Mass unemployment and inequality are therefore more likely to be perceived as a collective threat rather than individual bad luck and therefore more likely to increase ingroup bias and resentment towards the political outgroup, i.e. the main characteristics of affective polarisation. This transformation of distributional conflicts into identity conflicts is not something that is unique to underprivileged groups. Perceptions of scarcity can also cause the dominant class to question the claim of disadvantaged groups such as migrants or ethnic minorities to membership in the distributional order (Fehmel 2014). Economic shocks can thus trigger fundamental conflicts of identity in which the self-image and membership conditions of the social order as a whole are negotiated.

(2) Another factor associated with high levels of affective polarisation is the increase in disagreement on cultural issues. Problems of migration and diversity, the influence of international organisations and gender and race relations, are examples of such questions that currently shape the public discourse in many western democracies. These kinds of cultural conflicts between a “green/alternative/libertarian (GAL)” or “cosmopolitan” and a “traditional/authoritarian/nationalist (TAN)” or “communitarian” camp (Hooghe et al. 2002; de Wilde et al. 2019) form a relatively new social cleavage that cuts across the previously dominant economic antagonism between capital and labour. In countries where the distance in this dimension between political parties has increased, hostility and resentment across party lines—i.e. the level of affective polarisation—has increased as well (Gidron et al. 2020). The GAL vs. TAN cleavage thus seems to have a stronger polarising effect than the capital versus labour antagonism. From a theoretical perspective, we suggest two explanations for this. First, the structure of the conflicts differs. Economic conflicts are typically in principle divisible (Hirschman 1994; Offe 1998) and non-zero-sum conflicts. They can usually be solved by compromises of more-or-less, e.g. an increase in the minimum wage by one Euro instead of two, or by mutually beneficial positive-sum solutions. Ideological and especially identity conflicts, on the other hand, tend to be non-divisible and often follow a zero-sum logic, i.e. gradual compromises are not possible, at least on individual issues, and the success of one side implies—or is even constituted by—the defeat of the other (Offe 1998). Second, as we have seen above in the case of conflicts over membership rules, the normative valence of the conflicts differs for the conflicting parties immensely: ideological and identity conflicts, in which the basic values of the parties are at stake, are perceived by them much more directly as threatening than as distributional issues. This does not mean, of course, that economic inequality does not contribute to political polarisation. But as we discussed above, distributional issues are more likely to cause conflict when they are associated with cultural or political identities. Such a linkage between distributional patterns and group identities makes inequality appear to be an act of intentional discrimination and facilitates political mobilisation, which is a prerequisite for conflict (Stewart 2000).

(3) Finally, Gidron et al. also explore the hypothesis that affective polarisation is more pronounced in countries with majoritarian electoral systems that favour the formation of two-party systems. Behind this assumption lies the idea, central to the conflict dilution strategy, that a wider range of policy options structurally prevents the formation of “‘us-versus-them’ zero sum politics” (Gidron et al. 2020, p. 45) and that the need for coalitions in multiparty systems promotes a more compromising and “kinder” political culture (Lijphart 1999, p. 301). The results here are less straightforward (see also McCoy and Somer 2019). They show, however, that in political systems with proportional representation, satisfaction with one’s political camp is higher and dislike of one’s political opponent is lower than in majoritarian systems. Both dimensions of this finding provide evidence that the risk of destructive conflict is lower in proportional systems. Stronger ties to one’s own political camp can be seen as an indicator of successful representation, whereas reduced hostility towards political competitors also reduces the likelihood of escalating bilateral conflicts. In a diverse and interdependent political ecosystem, in which opponents are also always potential coalition partners, political emotions are diluted and kept in check both by experiences of past cooperation and by expectations of future cooperation.

In sum, these findings suggest, first, that status stress triggered by economic conditions has a negative impact on the quality of conflict. Resentment and aggression against political opponents increase when the economic situation of one’s own group appears hopeless or under threat. Thus, despite the focus on identity and value conflicts in current debates over an alleged crisis of democracy, democracies are well advised to pay more attention to the political costs of mass unemployment and inequality. Second, there are likely political costs in the sense of increasing affective polarisation emerging from publicly fought cultural conflicts. This observation is not meant to imply that public debates about important moral issues such as migration or racial justice should be stifled for the sake of social integration. But it is worth pointing out, in our view, that exploiting the potential for partisan outrage over these issues for the purpose of political mobilisation carries real risks. Finally, it is apparent that the institutional framework of democracy is of crucial importance for the representativeness and quality of the democratic process. This includes measures that affect both the demand and the supply side of democratic politics. On the demand side, political education programmes, a diverse media infrastructure and legal and social safeguards for a vibrant civil society are needed in order to help citizens to effectively channel and articulate their preferences into the political process. On the political supply side, it is necessary to establish the conditions for citizens to be able to choose from a diverse set of political choices that reflects their interests and values. The structural framework of voting and campaign finance laws is of particular importance here, as it has considerable influence over the range of viable options in a political system (Müller 2021, pp. 172–192; Schmelzle 2021).

3.2 Conflict Embedding in Democracies

The basic idea of the strategy of embedding conflicts is that conflicts lose their disintegrative potential if they take place within a given institutional framework and are recognised as a legitimate or even desirable option for action within that framework. This argument has a normative and a functional dimension that corresponds with the normative and rational choice reasons for accepting an institutional framework discussed in Sect. 2. In pluralistic societies, institutions develop normative integrative powers when they provide roles and identities with which social actors can identify despite conflicting values and interests, thus binding them to the social order. Under these conditions, institutions develop a functional integrating power if it is in the interest of the conflict parties to settle their conflicts within the framework of the social order. But how can democratic institutions achieve social integration in these two dimensions?

(1) Normative Embedding of Conflicts: one of the most famous—and from a research ethics perspective most notorious—experiments in the history of social psychology is Muzafer Sherif’s Robbers Cave Experiment (1961). The experiment is best known for its contribution to understanding the causes of intergroup conflict.Footnote 9 Equally significant, however, was a second hypothesis related to the conditions of intergroup integration in the aftermath of conflict. Sherif believed that the best way to reduce conflict and increase intergroup integration would be for the groups to cooperate towards a superordinate goal, i.e. a shared overarching project in which the groups would have to cooperate to realise the project’s aim. This latter strategy of diffusing group conflict is known as social re-categorisation in contemporary social psychological research (Gaertner et al. 1993). Its goal is not to dissolve group identities and intergroup conflict altogether, but to embed them in a broader shared identity associated with a common project.

Varieties of this strategy of social re-categorisation are also used in dealing with conflicts in modern democracies. In the case of ethnic divisions, this might be a multicultural national identity (Kymlicka 1995), or in the case of ideological differences, a shared republican ethos—sometimes referred to as “Verfassungspatriotismus” in the German debate—based on rights to social and political participation (Offe 1998). The idea in both cases is to establish overarching shared identities that do not resolve group differences but embed them in a common normative framework. One important method democratic societies use to achieve this goal is simply to provide institutions for resolving conflicts, such as collective bargaining and democratic elections (Luhmann 1983, pp. 100–107). By enabling their members to legally engage in, for example, labour disputes or political organisation and opposition, societies recognise, at least implicitly, that conflicts about interests, ideals and identities are permissible and that their protagonists are entitled to articulate them. This symbolic recognition of the permissibility of conflicts already contributes significantly to their de-escalation, because it implies that the articulation of dissent no longer has to be construed as an assault on the existing social order in toto, which has to be prevented as such. Already, Robert Dahl has argued with regard to the historical development of democracies that the institutional inclusion of conflict, for example, through the institutionalisation of opposition, has been a central factor in the democratisation of Western societies (along with the right to representation and the right to vote) (Dahl 1966, p. xi). The legitimisation of conflicts is thus the prerequisite for conflicts to become specific and thus manageable within the existing social order. Only by recognising conflicts as permissible does it become possible to delimit their scope and to decouple them from general questions of the social distribution of power.

(2) Functional embedding of conflict: it would be naïve, however, to trust that a normative belief in the legitimacy of conflict and of second-order social rules alone is sufficient to make social integration sustainable. The normative reasons for accepting democratic institutions must be supported or, if necessary, replaced by rational choice reasons that bind even those conflicting parties to the democratic order who are not sufficiently convinced of its legitimacy. It is conflicts that are managed, as Dubiel puts it (Dubiel 1997, p. 425), that can be integrative in this sense. Drawing on a helpful distinction by Adam Przeworski, one can differentiate three dimensions of the functional embedding of conflicts. Political institutions (a) structure, (b) absorb and (c) regulate conflict (2019).

(a) With an analogy by Douglass North, institutions can be described as the “rules of the game of a society” (North 1990, p. 3). These rules not only determine which actions are possible, desirable and permissible in social conflicts but they also influence what kind of actors can participate in them and what kind of goods are at stake. Institutions regulate, for example, how groups must be constituted in order to count as political parties, what legislative, judicial, and/or executive positions are distributed through elections, and what rules apply in the process. We have already seen that the electoral rules by which legislative mandates are allocated have a considerable effect on the electoral choice of citizens and the political culture at large. The rules of collective bargaining law are another example of the encompassing effect of institutions on the participants, processes and scopes of conflicts. By regulating access to this institutionalised space (e.g. by defining what constitutes a tariff partner), defining legitimate objectives of labour disputes and sequencing the conflict process (e.g. by formalising the right to strike/the obligation to keep the peace), the inherent conflict between labour and capital is not resolved, but structured and contained. Moreover, the privilege of trade unions to engage in industrial action shapes the organisational structure of the labour movement, which in turn determines how the conflict between capital and labour is addressed politically (Blau and Kahn 1999). This logic of containment by regulating actors, access and procedures can be witnessed throughout the institutional set up of democracies, albeit in very different shapes. We can observe it in the regulation of competition of political parties for shares of the electorate (Heidenreich 2020), in the transfer of conflicts into legal disputes and settlements (Luhmann 1983, pp. 55–136) or in the public sphere, where civil rights and criminal law ensure access to and the civility of public disputes (Deitelhoff 2013). By shaping the composition, options and aspirations of social actors, institutions create a thick net of mutual expectations and interdependencies that have a structurally integrating effect.

(b) Institutions not only structure conflicts but they also absorb and channel some of their energy. Conflicts in themselves tend to expand quantitatively and qualitatively beyond the actual cause by encompassing all interaction with the other party (generalisation) and transforming it from opponent to enemy (totalisation). Preventing this is a second central function of democratic procedures for conflict management. They counteract the generalisation and totalisation of conflicts by forcing the concretisation of the respective demands, excluding issues outside the scope of the procedure, ritualising and professionalising the settlement of the conflict, and finally reaching a binding decision that formally ends the specific conflict (for the time being). This last step, especially if enforced by a third party external to the conflict (Luhmann 1983, p. 103), gives the actors a face-saving opportunity to re-enter into regular social relations with the conflict opponent.

Moreover, the institutional pressure to publicly articulate concrete demands and positions contributes to the integration of society by rationalising conflicts. Jon Elster, for example, argues that the public character of political decision-making forces the conflicting parties to formulate their concerns in a universalistic rhetoric of the common good, which excludes factional interest and is aimed at the approval of society as a whole (1997). Most policy proposals and collective bargaining demands alike are therefore not justified with regard to the economic interests or sheer preferences of the respective clientele, but clothed in the language of equity, social justice and economic common sense. Elster argues that this necessity for hypocrisy leads not only to more inclusive rhetoric but ultimately to more inclusive politics: the need to justify policies in terms of the common good excludes the most blatant cases of factional self-interest from public discourse and forces actors to adhere to certain standards of impartiality (Elster 1998, p. 104). Even where substantive agreement is not possible, for example in polarised politics where political rhetoric is primarily aimed at mobilising one’s own supporters, public procedures for regulating conflicts can still contribute to the integration of society by making political cleavages visible (Dormal 2017, p. 238). In this way, conflicts help to simplify matters of political concern by accentuating the differences between political alternatives on the one hand and bundling arguments into coherent positions on the other. This can contribute to citizens’ willingness to accept the rules of second-order conflict regulation, even if they reject the results of these procedures.

(c) However, these “civilising” effects of institutional procedures for conflict regulation can only take effect if they are complied with by all conflicting parties. As we are currently witnessing in the USA, democratic elections, for example, lose their pacifying effect if their results are not recognised by the losing party. This raises the question of how second-order institutional rules regulating the distribution of social decision-making powers must be designed so that all parties to the conflict are willing to abide by them.

If members of society are to settle their conflicts within institutionally provided arenas and to accept their outcomes, then participation in these procedures must not generate prohibitive costs. Otherwise, official procedures such as elections or legal action will lose their attractiveness and societal tensions are discharged through other, usually more violent channels. Providing institutional safeguards for conflict involves setting clearly defined conflict rules, ensuring physical security in conflict, and providing a robust safety net of social and civil liberties that allows members of all social groups to participate in conflict without existential economic and political risks. The last point is of particular importance: if parties to a conflict fear that they will be deprived of basic rights or even face social or physical extinction in the case of a defeat; there is virtually no chance that they will accept the other side’s victory (see Hartzell and Hoddie 2003).

Additionally, whether institutional procedures for conflict regulation are actually used depends on whether the actors are convinced that these procedures will endure in the long run. Such long-term time horizons are important because they increase actors’ willingness to invest in institutions (Przeworski 2019, pp. 162–163). In addition to time and resources, the main issue here is to adapt the structures and capabilities of the conflict parties to the institutions. Transforming a social movement or a paramilitary organisation into a political party would be a key example of the case of democratic elections. Such transitions entail considerable costs and internal conflicts for the groups, which are only worthwhile if the institutions remain in place for the foreseeable future.Footnote 10 Once such investments are made, however, they in turn generate path dependencies that strengthen actors’ commitment to these procedures.

Finally, for actors to stick to institutions, there must be a general chance to prevail within them, i.e. transitions of power must be possible and have far-reaching, but not unlimited, social consequences. Only when both conditions are met is it reasonable for the current minority to engage with the institutional framework and to conduct its conflicts within it (Przeworski 2019, pp. 156–159). Fair decision-making procedures and iterative decision cycles ensure the possibility of changing majorities. Together with a long-term time horizon, they are necessary conditions for giving the current minority a realistic chance to replace the majority in power. Without this chance, acceptance of democratic rules would not be reasonable from a self-interested perspective. However, the possibility of formal power changes is only relevant for actors if significant social consequences are associated with them. This is ensured, first, by the reversibility of decisions, which increases the value of future power, making a temporary minority position more acceptable. Second, the formal powers of institutions must be broad enough to make participation in official procedures worthwhile but not unlimited, because then the risk of economic and social exclusion already discussed above becomes too high.

3.3 Adaptation and Appropriation of Social Norms Through Conflict

Although the two mechanisms of dilution and embedding that we have discussed so far are particularly suitable for containing conflict, the last mechanism that we now turn to is focused on enabling and nurturing conflict. Adaptation and appropriation of social norms through conflict ensures that norms remain responsive to changing needs and social circumstances. This mechanism, again, has both a functional and a normative dimension. A democracy’s capacity to revise existing norms and practices contributes to its functional integration by constantly adapting its first-order rules to the needs of the actors. Conflict thereby counteracts the inertia and status quo bias that most institutions tend to develop over time (Mansbridge et al. 2010, p. 81). The possibility of repeatedly contesting existing social norms and ideals also contributes to the normative integration of society. It affirms the role of citizens as equal authors of the social rules that they are subject to and thus gives them a reason to accept the legitimacy of the second-order rules of the social order, even if they do not agree with the current decision.

The integrative potential of these mechanisms stems from the ability of conflict to reshape social norms to reflect material, political and ideological changes in society. They do this by, first, performing an epistemic function. When conflicts arise, they often reveal hitherto latent, i.e. unreflected, rules and normative expectations (Coser 1956, p. 152). Conflicts, however, do not only point to what already exists, but they also enable the conflicting parties to change the status quo: normative expectations and rules become thematic in conflict and can be revised, discarded or confirmed accordingly. In conflict, the normative makeup of societies becomes visible and thus subject to change. It is precisely this dynamic function of conflict that Coser credits with contributing to social integration, as it makes members of society aware of their role as joint creators of their shared social space. At the same time, it gives democracies the ability to adapt to ever-changing realities and find innovative ways to deal with them (Coser 1956, p. 153). By constantly challenging the status quo and devising political alternatives through conflict, democracies ensure that positions of power do not become entrenched but remain elastic (Dahrendorf 1972, p. 261; Sunstein 2003).Footnote 11

But how can democratic societies reap these dynamic effects of conflict? One important strategy is to promote an active civil society that engages in political conflicts beyond the confines of formal institutional settings. Voluntary associations such as citizen groups, social movements or NGOs have long been praised for their contribution to democracy, most famously by Alexis de Tocqueville, who saw in them the root of the spirit of cooperation and trust in the American democratic system. Similarly, Robert Putnam argued that democratic institutions in Northern Italy were stronger and more effective because they could rely on a dense net of civil associations that fostered trust and cooperation, which regions in Southern Italy lacked (Putnam 1993).

Voluntary associations and social movements not only enable trust between citizens or build social capital in a broader sense, but they also serve as a social seismograph in democracies: Specifically, social movements form around themes and issues that have not previously been part of political struggles and debates: they detect new challenges that democracies need to deal with. Often, these challenges involve—and are articulated by—previously marginalised voices and groups whose inclusion and fair representation has to be fought for by civil society actors (Della Porta 2020). Social movements thus de-reify the status quo, which is a necessary condition for the revision of existing first- and second-order rules and the development of social and political alternatives in both dimensions (Celikates 2009; Bluhm and Malowitz 2012). Well-known examples of profound social change initiated by social movements include the civil rights movement in the USA and the ecological movement in Germany, which made environmental protection a subject of German public discourse and, finally, legislation.

However, as social movement research has demonstrated, civil society actors can only accomplish these effects under the right legal and political conditions. These include the rights to organise themselves, raise financial support, and participate in public debates as well as the opportunity to access and be heard within the political decision-making process (Tilly and Tarrow 2007). Although all of this seems to be the case in most democracies, it should by no means be taken for granted. In the German context, social movements only gained positive valence in the 1960s and 1970s. Until then, movements and protests were seen as a potential danger to democracy that needed to be strictly contained (Roth 2012, p. 37). With the advent of numerous new social movements in the late 1960s and 1970s this perception began to change. More and more citizens became part of movements that mobilised well into the middle classes. The new social movements were also often not directed at major political transformation of society but focused on specific reforms. As a consequence, street protest became an accepted expression of political participation (Rucht 2001, p. 36).

However, this “normalisation” of protest and movements that has been observed in western democracies since the 1970s and 1980s has faced challenges lately. Several governments, non-democratic but also democratic ones, have begun to tighten the regulations on organisations’ financial bases, their political organisation or their general access to public debates and decision-making bodies and forums (Buyse 2018; Bethke and Wolff 2020). This makes it increasingly difficult for civil society to ignite and spur relevant conflict.

4 Current Trends of Weakening Conflict

Democracies cannot but embrace conflict. Facing the irrevocable plurality of modern societies, democracies can neither suppress conflict nor can they avoid it at any reasonable cost. Instead, they need to find ways to utilise these conflicts for social integration. But how do democracies achieve this and what kind of productive functions does conflict have for social integration in the first place? In our discussion of the sociology of conflict, we distilled two basic productive functions of conflict: conflict contributes to social integration by constituting a shared social space and by providing the actors with the means of shaping social order according to their needs and preferences. We highlighted three analytically distinct mechanisms that help societies to turn conflicts productive: dilution, embedding, and adaptation/appropriation. They are all designed to contain conflicts so as to prevent them from escalating and to nurture them to allow citizens to relate to each other and to appropriate the norms and regulations at hand. We have discussed how democracies use these strategies by institutionalising forums for conflict, structuring the process of conflict, by defining the rights of participation, safeguarding against negative consequences and reprisals and by fostering unconventional forms of participation to support the detection and expression of conflict.

Despite all these precautions and regulations, the balance between containment and nurturing conflict remains fragile in democracies. Specifically, in times of crisis in which uncertainty looms large, managing conflict becomes a daunting task in democracies. In the remainder of this conclusion we discuss three trends in western democracies related to crisis that threaten the productivity of conflicts: the increase in (affective) polarisation that counters conflict dilution by re-aligning conflicts on one predominant axis, the radicalisation of politics threatening the embedding of conflict in institutions, and, finally, the depoliticisation trend that works against adaptation and appropriation by foreclosing conflicts.

4.1 Polarisation

As we discussed above, recent years have witnessed an increase in political polarisation in many democratic societies. From a conflict-theoretical perspective, this is not necessarily bad news: should not the clearer expression of political differences be welcomed, as it renders societal conflicts visible and addressable instead of allowing them to smoulder in hiding (Mouffe 2000)? Could it not be argued that pointed political disagreement is preferable to apathy, which was considered a major pathological condition of politics in western democracies only about a decade ago (Mair 2013; Mouffe 2018)?

We believe that these questions deserve a nuanced answer, which we can only hint at here: it is first crucial to distinguish between issue polarisation, which is concerned with substantive policy differences, and affective polarisation, which refers to sympathy and antipathy between political groups (Iyengar et al. 2012; Mason 2018). Although increases in issue polarisation do indeed promise integrative potential from a conflict-theoretical viewpoint, the rise of affective polarisation is concerning for three reasons, even from this perspective. First, affective polarisation usually generates identity conflicts that are often not amenable to compromise and productive solutions but instead adhere to a zero sum logic that is aimed at the defeat of the opponent as such—an outcome that cannot be expected to have an integrative effect. Second, this fosters an expressive style of politics that transforms all factual issues into questions of identity, as these determine the makeup of political camps—and not the other way around. This style of politics is attractive to political elites because it relieves them of their accountability for the substantive outcomes of their policies (Krupenkin and Iyengar 2018). Third, it is evident that polarisation of this variety undermines the legitimacy and, ultimately, the stability of democratic politics. If democratic procedures no longer serve to negotiate the common good, but are rather seen as a means of subjugation and suppression, then the minority can no longer be expected to accept the political decisions of the majority as binding. The rules of democratic institutions that regulate political competition lose their binding force when conflicts are totalised to such an extent that the equal status of the minority is at stake. The more institutions get discredited as fair arenas of conflict, the less likely citizens will make use of or accept the outcomes reached within them (Schmelzle 2021).

4.2 Radicalisation of Populist Politics

Polarisation is an important cause of the second of the threats to democracy we have diagnosed: the radicalisation of political dissent. By radical dissent, we mean here specifically efforts to achieve political change outside democratic procedures through direct action and violence. Radical political groups are part of the reality of all free societies. However, political radicalism threatens to become a mass phenomenon whenever relevant segments of the population come to believe that they have no chance of introducing their essential concerns effectively into the political process through the established channels, be it through voting, petitions, peaceful protests or legal strikes (Klandermans 2014).

On the one hand, as in the case of highly polarised societies, this conviction may stem from a perception that the political process is deeply unfair and corrupt. Such doubts regarding the fairness of the political process are currently often fuelled by populist rhetoric—think of the allegations of election fraud that initiated the 6 January insurrection in the USA—that questions the legitimacy of political opponents as such, characterising them as inherently corrupt or even as enemies of the people. On the other hand, radicalisation tendencies can feed on frustrations about the (supposed) low effectiveness of democratic systems, as is the case for radical representatives of the climate movement, for example. Thus, both forms of radicalisation formulate to some extent an always exaggerated and often inaccurate critique of the procedural or instrumental legitimacy of democratic institutions. They would be well advised not to reject this criticism wholesale but to accept it where it has a genuine core (Schäfer and Zürn 2021).

4.3 Depolitisation and Consensualising Politics

Another trend that has gained increasing attention during the COVID-19 pandemic but that developed long before the pandemic hit western democracies is the turn to experts and expert bodies in decision-making visible throughout liberal democracies. Political decisions are increasingly removed from social conflict and declared to be an administrative matter or a matter for experts, i.e. depoliticised. Political actors react to crises and conflicts with arguments of an alleged practical—or moral—necessity (the so-called TINA principle: “there is no alternative”) and justify its decisions, which may involve negative effects for some parts of society, as an inevitable reaction to (external) circumstances (Suhay and Druckman 2015; Séville 2017). Framed as such, politics is experienced as a constraint, but not as an object of joint struggle about the conditions of living together. It becomes a depoliticised administration in the face of overwhelming threats, but no longer a matter for citizens who decide who they are and how they want to live. But this is not just a matter of political elites who turn to experts to release themselves from responsibility. At the same time, and in particular in times of crises such as the current pandemic, the public also supports a stronger role for experts in political decision-making. In the latest Science barometer of 2020 (focusing on the pandemic situation) more than 80% of the German population favoured more inclusion and a greater impact of science in politics (Wissenschaft im Dialog 2020). This implies that scientists and science are still a major source of trust in modern societies deemed to be capable (competent and ethically committed) of solving political problems. But at the same time, it favours non-competitive policy-making. This is not very surprising because crises are characterised by heightened uncertainty and related insecurity about the future that makes citizens less inclined to engage in conflict, which rather increases uncertainty (Bösch et al. 2020). Depolitisation works as an effective showstopper to conflict because it de-legitimises conflict as being destructive on the one hand and superfluous on the other (experts know better).

These trends are being fuelled by the enduring crises experience of the last two decades. Global crises phenomena transported and amplified by the media, such as the economic and financial crisis of the 2000s and the refugee movements in the 2010s, which was exaggerated into an existential crisis, have created a feeling of elementary uncertainty among many citizens as to whether institutional politics and political elites have the competencies to counter the unleashed forces of globalisation or, in a negative connotation, whether they even have the will to do so.

Increasing inequality also becomes relevant in this context. Although globalisation has benefited some and created new opportunities, for others it has led to the loss of opportunities, social decline and increased competition for welfare services (Manow 2019). Those who perceive themselves as disconnected no longer see opportunities in globalisation but only elementary existential risks. In conjunction with the declining effectiveness of national institutions, the group of ‘losers’ is showing a sharply declining willingness to engage in open conflict, combined with calls for more authority (Ipsos 2017). Globalisation has not only accelerated the individualisation of life plans and values it has also reduced the congruence of political decision-making spaces and social action contexts. The transnationalisation of social contexts of action reduces the effectiveness of political decisions at the national level. At the same time, the more decisions migrate to the transnational level in order to restore their effectiveness, the less incentive there is to perceive the national political institutions of decision-making as an arena for political debate (Mair 2013; Schäfer and Zürn 2021). Alienation and mistrust are the likely consequences (as early as Kirchheimer 1957).

Finally, the inherent logic of institutional systems to get more resistant vis-a-vis dissent plays out particularly strongly in times of crisis (Mansbridge et al. 2010, p. 68). When institutional systems are under pressure, their representatives try to find work-arounds to avoid conflict given its inherent risk of escalation. The same holds true for citizens. The more they experience a sense of crisis, the less they feel inclined to engage in conflict with its uncertain process and outcomes. Thus, not only institutional inertia but also the continuing threats arising from major political crises put the democratic strategy of promoting social integration through conflict under pressure. It is therefore all the more important to tolerate or—even better—protest against expressions of political opinion that seem provocative or even foolish to us. They are a price that we should be happy to pay for a lively democratic culture.