1 Introduction

The idea of organising a special issue on the topic of social integration arose in the context of the founding of the interdisciplinary Research Institute Social Cohesion (RISC). In this research collaboration, funded by the German Federal Ministry of Education and Research, scientists from 11 research institutions in Germany collaborate in research and transfer activities centred on the issue of social cohesion. We—the editors of this special issue—also serve as Principal Investigators in this institute, and from the beginning we acknowledged the need for a publicly visible and scientifically sound social science contribution to the highly politicised debates regarding the supposedly precarious state of social cohesion in Germany, Europe and the Western world. We are indebted to RISC and its members for valuable intellectual exchange and financial support during the process of realising this special issue.

As an editorial team, our shared interest has been twofold. First, we aimed to utilise the sociological concept of social integration as a tool to reframe the various normative and political discussions around social cohesion within a well-established research tradition that lends itself to empirical–analytical investigation. To this end, and based on Uwe Schimank’s groundwork, we have developed the concept of social integration as outlined in the introduction to this special issue. We have shared and discussed this framework with the special issue contributors early in the process and received valuable feedback in a workshop held in June 2021. In this final chapter, we review how the papers relate to, use and critically reflect upon our concept of social integration.

Our second shared interest was to take stock of the state of the art in social integration research both nationally and internationally, and especially, to seek answers to questions we consider crucial for the field, but which remain unanswered. We identified scholarly discord or neglect regarding the questions (1) how social integration is generated and what its central mechanisms are; (2) whether social integration is considered a functionally necessary precondition of societies or a normatively desirable state; and (3) what the “dark”, unintended, negative effects of any socially integrated society are. In this conclusion, we compile the answers that the contributions to this special issue provide to these questions.

Various articles also raised new topics about social integration. One of them—relating social integration to societal macro structures—seems to be of special importance, so we want to highlight it here as deserving further attention. To be sure, our review will not be exhaustive, and our reading may be incomplete, but we believe that since our first tentative ideas, the field has made important progress.

2 Social Integration as Balanced Equilibrium of Four Ingredients

We begin by relating the contributions to this special issue to the theoretical concept of social integration proposed in the introduction (Grunow et al. 2023, this issue). To be sure, we do not use our concept as a yardstick to determine which contributions have gaps or ambiguities. Such an assessment would be inappropriate because, depending on the specific topic, only a small selection of the analytical tools we propose may be useful, and often the tools used might have to be adapted to the issue at hand. Therefore, our concern here is to examine where our concept can learn from the contributions compiled here and assess its relevance. To this end, we first consider which of our proposed ingredients of social integration the authors applied in which analytical context and in what combination. We then turn to three features of our concept, which characterise its analytical potential: a conflict–theoretical approach, a multi-level view of social integration and an understanding of social integration as a balanced state between dis- and over-integration.

2.1 Ingredients of Social Integration

The contributions to this special issue exhibit a wide spectrum of combinations of the four ingredients of social integration (consensus and trust as co-orientations, conformity and cooperation as co-interactions). Five contributions consider all four ingredients. The measurement instruments for social cohesion or social integration reviewed by Jan Delhey and his colleagues are in accordance with our proposal that this social phenomenon must be viewed as multidimensional and encompassing both attitudinal (co-orientation) and behavioural (co-interaction) elements. The terminology, the number of dimensions and their exact content differ from index to index (see Delhey et al. 2023, this issue, Table 2), but especially trust and cooperation often figure as key ingredients. The contribution by Olaf Groh-Samberg and colleagues also includes aspects of all four ingredients. The authors propose and apply a concept of social milieus that operationalises shared co-orientations through the Schwartz-value scheme (Schwartz 2015), whereas co-interactions are structured by socio-economic status (as indicated by education and household income). Thomas Schwinn’s theoretical reflections, although focused on other aspects of social integration, contain several references from which we infer that he, too, includes all four ingredients. Matthias Koenig’s review of recent scholarship on religious diversity and social integration in Western Europe applies a unifying framework that centres on the theoretical notion of “boundary-making”, i.e. practices of cognitive classification and social distancing. Boundary making encompasses an emphasis on consensus/dissent and trust/distrust as well as the performance of conformity/dissent and cooperation/non-cooperation. Finally, Natalie Grimm and colleagues deal with what they categorise as a hyperwork society by referring to all four ingredients to point out societal changes of the last decades as combinations of simultaneous dis- and over-integration. Thus, the three most general contributions, as well as the overviews on religion and work, confirm the four ingredients of social integration we proposed.

Importantly, our concept highlights the interrelation among the four ingredients of social integration. In this sense, Fenella Fleischmann and Yassine Khoudja devote attention to the interplay of consensus, conformity and cooperation. Their thorough assessment of individual religious change of immigrants to the Netherlands suggests that the boundaries between Muslim immigrants and the secular host societies of Europe may blur over time. To explain changes in immigrants’ religiosity, the authors refer to cooperation within and across religious boundaries as potential mechanisms. They also investigate conformity—in this case, to norms of religious behaviour—and consider the role of consensus regarding views on gender equality and liberal values.

Three other contributions put an emphasis on the interrelation of consensus and cooperation. Carlotta Giustozzi operationalises consensus through work and family values, and cooperation through the frequency of meeting and helping friends and family, as well as paid work. Based on this operationalisation she finds that in the context of Germany’s male-breadwinner culture, men’s social integration is weakened by unemployment to a greater extent than women’s—which is explained by women’s stronger family and lower work orientations. The paper by Christopher Swader and Andreea-Valentina Moraru deals with loneliness, which, understood as a person’s “actor integration” (Münch 2015, p. 43), refers to a lack of consensus and cooperation with fellow human beings. Surprisingly, they find that people in highly individualised countries, characterised by lower levels of social integration, nevertheless report lower levels of loneliness. This contradicts the often-heard thesis that loneliness is one manifestation of lower levels of social integration. The third article, which highlights the consensus–cooperation nexus, is by Nicole Deitelhoff and Cord Schmelzle. Without denying that some degree and kinds of consensus and cooperation are necessary for social integration, they emphasise that too much of both can be detrimental to a society’s capability to innovate and re-design its structures.

The trust–cooperation nexus—highlighted in classic Rational Choice explanations of social integration such as Robert Axelrod’s (1984) study of the “evolution of cooperation” and emphasised by Delhey et al. (2023, this issue)—is the topic of three articles. Jennifer Silva confirms this nexus by considering the opposite situation. Studying young, disadvantaged working-class adults in a rural coal town in Pennsylvania, she shows how distrust and a lack of cooperation—evident in her interviewees’ social isolation and individualised coping strategies—foster political disengagement and susceptibility to conspiracy theories, alienating them from overarching collective points of reference. Stefan Jünger and Merlin Schaeffer operationalise social integration as community attachment and social trust. They find little empirical support for the notion that social integration is reduced in ethnically diverse neighbourhoods. Jennifer Fitzgerald and her colleagues analyse democratic effects of belonging to place-based communities among young Swedes. They find that feelings of belonging strengthen support for democracy, independent of internal political efficacy beliefs, anti-immigrant attitudes, or trust. These results highlight the positive effects of cooperation experienced in place-based communities, next to established factors such as trust.

Finally, two articles focus on one specific ingredient of social integration. In their paper on the development of globalisation cleavage, Céline Teney and Li Kathrin Rupieper investigate to what extent public opinion in Germany is polarised along issues of immigration, European integration, economic liberalism and the environment. Their analysis speaks of consensus as an ingredient of social integration. Importantly, the authors find no evidence of an increasing polarisation of attitudes. Instead, German public opinion appears to exhibit a higher degree of consensus than the controversial public debates around these issues suggest.Footnote 1Kathrin Ackermann, Julian Erhardt and Markus Freitag focus on the cooperation ingredient of social integration. Their cross-national analysis investigates how volunteering, which is an instance of cooperative behaviour, is shaped by welfare state institutions. Rather than crowding out civic engagement, strong welfare states appear to strengthen civic participation, and thus cooperation, especially among those benefiting from welfare state policies.

As was to be expected, not all logically possible combinations of the four proposed ingredients appear in our compilation of contributions. The number of contributions is smaller than the number of possible combinations, and some combinations occur more than once. Importantly, none of the contributions highlights the nexus between consensus and conformity, which represents the hallmark of a Parsonian understanding of social integration (Parsons 1970 [1951]). This absence may be pure coincidence—but what could it mean if it is not?

2.2 Conflicts and Multi-Level Dynamics of Dis- and Over-Integration

Besides various combinations of the four ingredients, the contributions to this special issue speak to three further features that characterise our concept of social integration. The first one is a conflict–theoretical approach to social integration. At first glance, conflict appears to many—including social scientists—to be the opposite of integration. But this narrows conflict to destructive confrontations and confuses integration with harmony. Deitelhoff and Schmelzle oppose such misunderstandings and demonstrate how the productive management of conflicts can promote social integration. Their prime example is democracy as a political mode of shaping society collectively—despite conflicts of interest, or even by using these conflicts to reach collectively accepted, balanced decisions. They show that in culturally pluralistic societies a maximisation of consensus is not a feasible way of arriving at decisions. At the same time, social groups that show little tolerance of disagreement challenge the functioning of democracy. Plurality needs democracy, but those who tire of plurality quickly prefer undemocratic forms of political debate. Schwinn (2023, this issue), too, uses a conflict–theoretical analytical framework into which he sorts inequality- as well as culture-based conflicts about social integration.

What Deitelhoff and Schmelzle as well as Schwinn ignore with their focus on political conflicts over social integration are non-politicised problems of social integration—i.e. many phenomena that fall into Robert Merton’s (1938) typology of anomy; for instance, all kinds of deviance, especially crime, or individual exits from society, such as drug abuse or migration to places looking more promising for one’s life chances. A complete frame of reference for analysing social integration would have to look at this part of the spectrum as well, as do the contributions by Swader and Moraru as well as Silva.

A second feature of our concept of social integration, which resonates with the contributions to this special issue, is its multi-level perspective. We proposed intersecting social orders from interactions and small groups to society as a whole; and we suggested that the level and kind of social integration of a lower level order can affect the social integration of society. In his contribution, Schwinn also takes a multi-level perspective, but makes a committed plea against using the societal order as a reference point for the analysis of social integration. For in his neo-Weberian perspective, society is nothing but the sum of all social events and does not represent an independent level of order—certainly not one that frames lower levels. He views social integration as a multi-level process, beginning at the micro-level of social categories and social relations and extending to the meso-level of social milieus and interest groups.Footnote 2 Schwinn views the interrelationship of these levels analogously to Reinhard Kreckel’s (1992) proposal of a “political sociology of social inequality” based on conflict theory. Perceived inequalities of social categories constitute integration problems at this level; as soon as these individual experiences are communicated and shared among those similarly worse off, integration problems emerge at the next higher level. Finally, shared experiences in social milieus can give rise to organisations capable of collective action. Conversely, the compromises found in these struggles reshape the inequalities of social categories, so that the dynamic interplay between levels may quickly switch from dis- to over-integration. These layers of social integration are Schwinn’s main contribution to a further conceptual elaboration of social integration.

Ackermann et al., Swader and Moraru, and Grimm et al. provide further examples of multi-level conceptualisations of social integration. All three contributions point to the welfare state as the macro-level social order, which in its interplay with individuals is a decisive factor of social integration. Grimm and colleagues add the organisations of collective bargaining and work organisations on the meso-level between the welfare state and individuals. This focus on the welfare state is another version of a multi-level perspective, which is useful for studies of social policies as instruments to promote social integration.

A third important characteristic of our concept is that we do not contrast integration and disintegration but disintegration and over-integration—so that integration is a matter of the right balance. The mainstream understanding, which assumes that social integration is an end in itself or that it has only beneficial societal effects, implies that more integration is always better than less integration. Although this makes it easier to assess and measure social integration using a simple additive logic, it comes at the cost of denying the negative side-effects of social integration. Delhey et al. document how, up to now, all relevant quantitative studies of social integration—including their own—follow such a more is better logic, which, in our understanding, misses the sociologically decisive point. Implicitly, many contributions to our special issue—also those applying a qualitative empirical approach—tend to share the mainstream assumption. For example, Giustozzi conceptualises unemployment as a problematic shift towards social disintegration for those affected. However, her findings show how unpaid family work can compensate for this loss, at least for women. In today’s work society, some may actually consider unemployment a short-term relief from over-integration through paid work (see also Grimm et al. 2023, this issue), a phenomenon that is nowadays captured analytically under the heading of work–life balance. To be sure, seeking short-term relief from over-integration caused by an imbalance of paid and unpaid work comes at a cost, as we spell out in Sect. 2.5, even if it is a different kind of integration problem from the one caused by over-integration through paid work.

Grimm and colleagues address the issue of over- and disintegration through paid work more explicitly. They link our concept of social integration to the changing nature of paid work in the late twentieth and early twenty-first century, arguing that both the manifest and the latent functions of paid work for social integration weaken in times of rising unemployment and employer-driven flexibilisation strategies. These trends carry the risk of social disintegration, as more and more people lack access to the socially integrative functions of paid work, owing to unemployment or low-quality jobs (Grimm et al. 2023, this issue). At the same time, a dynamic is set in motion, in which persons need to organise their lives increasingly around paid work to make ends meet, leading to an unbalanced state of employment-centred over-integration.

Deitelhoff and Schmelzle (2023, this issue) make a determined plea—not only analytically but also normatively—to view too much social integration as a problem. Over-integration is one of the dark sides of social integration that we address in Sect. 2.4. The article by Groh-Samberg et al. (2023, this issue) implies that over-integration within social milieus may foster disintegration at the level of society—an observation shared by Koenig (2023, this issue). For Schwinn (2023, this issue) the latter is a typical result of the multi-level dynamics of social integration and thus no surprise to a sociological observer.

2.3 The Causal Mechanisms of Social Integration

We now turn to the first of the three questions raised in the introductory chapter. Has there been a turn to causally identified and mechanism-based explanations in recent studies of social integration?Footnote 3 Our reading of the contributions to this special issue suggests that the answer might be no. None of the empirical contributions makes a deliberate attempt to identify causal effects in the generation of social integration using experimental or quasi-experimental tools of causal inference. Nor are they concerned with the in-depth reconstruction of the mechanisms that generate, sustain, or destroy social integration.

Instead, most articles focus on correlates of social integration to evaluate theoretical claims about causal mechanisms. For instance, several papers emphasise the importance of individuals’ socio-economic position for social integration and the role of the welfare state in balancing individuals’ different capacities to experience social integration via paid work (Groh-Samberg et al. 2023, this issue; Grimm et al. 2023, this issue; Giustozzi 2023, this issue; Ackermann et al. 2023, this issue; Swader and Moraru 2023, this issue). But they do not offer a strong causal analysis of how these mechanisms generate social integration. Some articles at least implicitly sketch mechanism-like theoretical arguments. Coming from a macro-sociological and cross-national perspective, Delhey et al. (2023, this issue) find that social integration flourishes in more affluent societies with low economic inequality and widely shared post-materialist (as opposed to traditional) values. They also report an important non-finding: population heterogeneity in terms of religion, ethnicity or language is unrelated to a society’s level of social integration. This echoes the results reported by Jünger and Schaeffer (2023, this issue) who take a much more granular approach based on 1 × 1-km census grids in Germany but also cannot relate ethnic heterogeneity—in this case in the form of residential segregation patterns—to a lack of social integration. Together with findings by Fleischmann and Khoudja (2023, this issue) on the gradual assimilation of immigrant religiosity over time, these contributions converge on the important conclusion that social integration is perfectly possible under societal conditions of cultural diversity.

More generally, social integration under conditions of social pluralism seems to depend on three prerequisites, which Deitelhoff and Schmelzle lay out in their theoretical contribution. The first prerequisite is a dilution or small-scale processing of conflicts so that no single one can have a major disintegrative effect in society. The second prerequisite is an embedding or procedural regulation of conflict that prevents unregulated radicalisation up to the use of violence. The third prerequisite entails an “adaptation and appropriation” of conflict based on democratic values, which welcomes disagreement and dispute instead of insisting on “the only right solution”. In sum, Deitelhoff and Schmelzle’s three-component model presents a versatile middle-range mechanism that fits well into Schwinn’s and our frame of reference (Grunow et al. 2023, this issue) and is supported by several key empirical findings in this special issue.

Making social integration work in diverse environments requires favourable and fair economic conditions, as well as a broad consensus on liberal values. Ackermann and colleagues as well as Swader and Moraru provide evidence for the importance of institutions—specifically, a generous welfare state—for the development of cooperative behaviour and the alleviation of loneliness. Besides state institutions, Fitzgerald and colleagues’ findings point at a potential interplay between cooperative social relations in place-based communities and a consensus on democratic norms. Similarly, though not making strong causal claims, the qualitative analysis provided by Silva shows how social isolation—i.e. the absence of cooperative relationships—is connected to growing distrust in others and a turn towards conspiracy theories, which can be interpreted as turning away from a democratic consensus on the workings of the political system. Teney and Rupieper argue that both opinion polarisation and the cleavage between winners and losers of globalisation have been overstated in public debates. They emphasise the role of political entrepreneurs who successfully mobilise voters with pre-existing anti-globalisation attitudes.

2.4 Functional and Normative Baseline of Social Integration

Turning from social integration as the phenomenon to be explained, our next question about the functional necessity or normative value of social integration treats it as an explanatory variable for other desirable social outcomes. Indeed, empirical research on social integration generally tends to start from the premise that social integration is a desirable state of affairs. The implicit justification is often a functional one and follows a utilitarian logic. It is explicitly addressed in the contribution by Delhey and colleagues who highlight that societies that score high on indices of social cohesion also tend to score high on indicators of subjective well-being, happiness and life satisfaction (in a similar vein Swader and Moraru 2023, this issue). They are also less likely to spiral into violent conflict. From a cross-national perspective, therefore, social integration is indeed a good thing, which is linked to a wide range of desirable societal outcomes.

Although few articles state this assumption explicitly, many appear to share it. For instance, Silva’s empirical analysis of political alienation among young members of the US-American working class vividly testifies to the negative consequences of social disintegration not only for marginalised individuals themselves but also for society, as personal suffering is individualised and collective action for social change rejected. Giustozzi and Ackermann and colleagues can be read as mirror images to these findings, showing that closer social relationships and more volunteering are beneficial for both individuals and society.

Even beyond such functional justifications of why social integration is of value, there are—at least implicit—normative assessments of its desirability. For instance, the articles on religion and identity are in one way or the other motivated by the question of how social integration can work under conditions of cultural and religious diversity (Fleischmann and Khoudja 2023, this issue; Jünger and Schaeffer 2023, this issue; Koenig 2023, this issue). The stated concern that mass immigration, conflicts over values or residential segregation may harm social integration echoes the worries of early sociologists about a dissolution of the social bond, implying that this bond is inherently desirable. Similarly, Fitzgerald and colleagues tacitly assume that a positive reference to larger collectives—which is what feeling at home means—is necessary for pro-democratic orientations to develop. Also here, then, social integration appears as an end in itself if one shares a basic normative preference for democracy over alternative forms of political order (see also Deitelhoff and Schmelzle 2023, this issue).

Thus, although few of the papers assembled here explicitly reflect on the functional necessity or normative desirability of social integration, many contain at least implicit assumptions about positive effects of social integration, and corresponding negative effects of social disintegration. However, the picture is not that simple, as we suggest below.

2.5 Acknowledging the Dark Side of Social Integration

Our concept of social integration has proven especially fruitful for investigating the dark side of social integration, i.e. its unintended or tacitly accepted side effects for society, its individual members and particular social groups.

Contemporary western societies all rest on social cohesion among members with unequal social status, which implies that some social groups carry higher burdens and costs of established modes of social integration than others (see for example Delhey et al. 2023, this issue; Grunow et al. 2023, this issue; Grimm et al. 2023, this issue; Giustozzi 2023, this issue). The extent to which the welfare state alleviates these costs and burdens, and thereby generates and stabilises modes of social integration, varies between countries and over time, as the contributions by Ackermann and colleagues as well as Swader and Muraro show. We suspect that this is one reason for the rising political concern over the state of social integration in contemporary western societies.

As religion has always held an ambivalent place in sociological thinking on social integration, it is perhaps not surprising that the contributions to this topic contain more reflection on the dark side of social integration. Koenig, referring to the prejudices religious and secular majorities harbour towards religious minorities, spells out the dangers of over-integration. However, these dangers are also visible in the behaviour of members of religious minorities themselves, for example, if they self-segregate or uphold values incompatible with the liberal mainstream. That these two dynamics of over-integration, which operate on the meso-level of society, may be mutually reinforcing adds an important insight into the mechanisms of the dark side of social integration. At the same time, this example shows the importance of adopting a multi-level perspective to social integration.

Other papers address the costs associated with social disintegration, especially in the form of a lack of cooperative social relationships. The two analyses by Silva and by Fitzgerald and colleagues suggest that the integration of actors into the political sphere rests on feelings of belonging and attachment to larger social collectives or communities. Although Fitzgerald and colleagues highlight the positive impact of place-based feelings of belonging on pro-democratic orientations, Silva reconstructs the political alienation and rejection of a basic democratic consensus that goes along with the social isolation and marginalisation of her respondents. Hence, in the political sphere disintegration is associated with considerable costs (see also Sachweh 2020).

Grimm and colleagues and Giustozzi problematise the central role of paid work for social integration. Importantly, these papers demonstrate that a mode of social integration centred on paid work takes for granted that some groups in society—especially women and immigrant workers—provide unpaid or low paid work in the care and household sector. This leads to a situation where both social integration and inequality are strengthened at the same time. Paradoxically, not only those who benefit support and stabilise these modes but also those who are disadvantaged by them. This paradox is rooted in the fact that on the micro- and group level, the co-interaction ingredients cooperation and conformity can conflict across social domains (i.e. labour market and family) and with co-orientation ingredients (consensus and trust). As Schwinn (2023, this issue) argues, such conflicts can only destabilise social integration at the societal level, or lead to change, if communicated and politicised effectively.

On one side of this paradox, actors continue to orient themselves towards paid work and the access to commercial goods it provides as a key mechanism for social integration (Grimm et al. 2023, this issue; Swader and Moraru 2023, this issue). On the other side, this integration mechanism is not available to all, as Giustozzi shows. According to her findings, paid work is considered more important, and thus more accessible, for men than for women whereas the family is considered to be more important for women than for men. Because both paid and unpaid work generate social integration (Jahoda 1981), women still feel socially integrated (Giustozzi 2023, this issue; Turner and Turner 2013) and thus continue to accept the unequal status quo.

Another force in the social integration paradox is that the now dominant one-and-a-half or dual-earner family models create gaps in the care domain of society, which the male breadwinner model used to generate for free. Now, care work is to a large extent low-paid work, mostly provided by women and immigrant workers. The former group accepts these jobs because they enable work–family reconciliation, the latter group because of low access barriers. Both examples show that strong social integration cements existing inequalities.

2.6 Societal Macro-Structures and Social Integration

We now turn to a topic that emerged as important, if not uncontroversial, in several contributions to this special issue, even if we did not explicitly address it in our concept of social integration. In the introduction, we stressed our interest in the social integration of societies, not of lower levels such as particular social spheres, organisations or families. This analytical reference to society can be specified in more detail, we learned, if we first consider the fundamental social orders of Western modernity with a view to social integration (social inequality, capitalism, cultural diversity and functional differentiation). The significance of nation states for social integration must then be addressed, as well as the problem of methodological nationalism.

2.6.1 Social Inequalities, Cultural Diversity and Capitalism as Challenges of Social Integration

When sociology began to problematise the social integration of the emerging modern society in the nineteenth century, this was equally prompted by four formative social orders (Schimank 2015): (1) the capitalist economy and its impact on society; (2) a dominance of labour-market inequalities in interplay with other inequality structures; (3) cultural diversity; and (4) functional differentiation of social spheres such as science, politics, education, or art, each with its own guiding values setting the tone. Inherent in each of these social orders of modernity are tensions that challenge social integration.

One of these orders—functional differentiation—however, raises problems of systems integration, not social integration. Problems of systems integration may only indirectly spill over to social integration. The most important way of how this can happen is when the welfare state does not effectively balance integration risks and burdens for large groups in society, as several contributions to this special issue mention (Groh-Samberg et al. 2023, this issue; Grimm et al. 2023, this issue; Giustozzi 2023, this issue; Swader and Moraru 2023, this issue).

How do the contributions reflect on the other three sources of threat to social integration? Most contributions refer explicitly or implicitly to two of these social orders: inequality and culture. In his theoretical framework, Schwinn, following Weber, views social integration as the outcome of conflicts between societal groups generated by social inequalities experienced and politicised as unjust—in particular against the background of the modern cultural idea of equality—on the one hand, and by cultural clashes such as religious differences on the other. Obviously, contributions dealing with religion or political ideologies emphasise the cultural dimension, whereas contributions whose topic are social milieus or classes accentuate the inequality dimension. However, social milieus are not only characterised by different incomes, wealth and educational credentials but also by different ways of living and ideas of a good life. At the same time, different religious groups or political camps differ with regard to their economic situation, educational level and educational opportunities. Thus, both social orders, inequality and culture, and the tensions they generate for social integration overlap and can reinforce each other, as is stressed by Schwinn and Groh-Samberg et al. and as also referred to in several of the empirical studies (Ackermann et al. 2023, this issue; Fleischmann and Khoudja 2023, this issue).

Capitalism is another topic addressed by contributions to this special issue. However, we note a disregard of a related and important aspect, namely economisation pressures. Capitalism comes into view as a generator of inequality. This aspect is of course an important part of Schwinn’s Weberian framework. It is also explicitly addressed in some empirical studies, most prominently in Silva’s contribution. Implicitly, capitalism and its specific shaping of labour market inequalities is a background variable in other contributions such as the one by Grimm et al. What is missing, though, is capitalism as a source of economisation pressures all over society (Schimank and Volkmann 2021). Again, this is a spill-over problem for social integration, originating in systems integration, i.e. the inadequate working of the capitalist economy. Whenever the situation of the capitalist economy deteriorates, this leads not only to wage pressure and unemployment but to declining tax revenues and higher social spending by the State. This then affects all State-supported non-economic sectors of society owing to pressures to save money. In addition to public administration, this applies to the police, the courts, and welfare services ranging from health, care and education to cultural institutions and sports promotion. Cutting costs in all of these State activities, each of which contributes to social integration in its own way, quickly brings about its erosion.

2.6.2 The Role of the Nation State

Another important macro-structural point of discussion that the theoretical (Schwinn 2023, this issue; Deitelhoff and Schmelzle 2023, this issue) as well as the conceptual (Delhey et al. 2023, this issue; Grimm et al. 2023, this issue) and empirical (Ackermann et al. 2023, this issue; Swader and Moraru 2023, this issue) contributions to this special issue address is the significance of nation states for social integration.

Whereas the conceptual and empirical contributions emphasise the nation state especially as a welfare state, Schwinn (2023, this issue) understands the nation state more fundamentally as a constitutive framework that closes off and holds together the multi-level architecture of social integration from above. It is quickly recognisable that for Schwinn statehood is a substitute concept for his deliberately weak concept of society that merely denotes the sum of all social activities and does not attribute any formative or even ordering force to society. The nation state undoubtedly represents such a force. But Schwinn tends to inflate the State to an authority above other societal value spheres, such as science, education, religion or the economy. To be sure, the education system is financed and regulated by the State; but whether social integration through education succeeds ultimately remains dependent on how the educational system, in particular the pedagogical profession, shapes teaching. A concept of national society would be analytically broader and could trace in more detail how the various value spheres, including politics, contribute to social integration.

Deitelhoff and Schmelzle also appear to think primarily in terms of the nation state in their description of democracy as the appropriate mode of political production of social integration, as many of their reflections and examples underline. However, as early as post-World War I, with the League of Nations, and then intensified after World War II with the United Nations, global democratic procedures have been institutionalised as a means of peacefully contesting interests between states. The European Union is probably the most developed project of the joint pursuit—and not merely of the reconciliation—of interests. Analytically, Deitelhoff’s and Schmelzle’s considerations can easily be applied to the supranational upscaling of democracy. Empirically, however, important preconditions for a productive management of conflicts of interest do not exist supranationally. At present, there is a growing discrepancy between the limited possibilities of democratic conflict resolution on the one hand and the increasing necessity for it on the other, especially in dealing with climate change and geopolitical instabilities.

Our take on social integration goes beyond the contributions by Schwinn and Deitelhoff and Schmelzle in one important aspect. It equips us to reflect critically upon established “national” or “European” modes of social integration and their interdependencies with other countries and world regions. National consumption styles and divisions of labour that stabilise social integration in Western societies have repercussions for individuals and resources in other world regions. Yet, although climate change, for example, poses a major challenge for societies worldwide, recent decades have witnessed a resurgence of nationalist developments that bespeak an inward-looking search for solutions to global challenges. Furthermore, researchers implicitly or explicitly conceptualise and measure social integration as a phenomenon that is bound to the nation state, and many of the contributions to this special issue still take the nation-state logic for granted. This, however, obscures the fact that much of what enables social integration in one part of the world has consequences for social integration in other regions. Paying greater attention to the transnational interdependencies in which different national modes of social integration are entangled—also with regard to their dark sides—is therefore a critical challenge for future research.

3 Outlook

In sum, the contributions to this special issue demonstrate that both theory and empirical research on social integration have advanced far beyond the discussions we have seen in earlier decades, not least since the predecessor volume edited by Jürgen Friedrichs and Wolfgang Jagodzinski (1999). Both conceptually and empirically, we have a better grip on the various facets of social integration, including its macro-to-micro and micro-to-macro interactions. We are also better at measuring the correlates of social integration and able to provide answers to a broad range of salient societal questions. Needless to say, much work remains to be done. Our synopsis of the contributions assembled in this special issue and the broader literature suggests several crucial areas to explore in the future. We would like to emphasise two of these areas: the problem of theoretical and methodological nationalism, and the lack of causally identified tests of the mechanisms of social integration.