Identity Politics, Core State Powers and Regional Integration: Europe and Beyond

This article concentrates on the path from the development of collective identities to the integration of core state powers. Firstly, we focus on the European experience. We argue that the identities of political, economic, and social elites have been crucial for the evolution of European integration. With regard to mass public opinion, European integration has been made possible by a consensus of EU citizens with inclusive national identities. Most recently, the politicization of EU affairs in many member states has been driven by populist forces mobilizing minorities with exclusive nationalist identities. Secondly, we discuss the extent to which insights from Europe have travelled to other regions of the world. Elites involved in region&#8208;building almost always develop identity narratives linking their national experience to the respective regions. Moreover, there is evidence that the difference between inclusive and exclusive nationalist identifications has also travelled beyond Europe.


Introduction
This special issue explores the relationship between political identities with a territorial dimension, on the one hand, and the transfer of core state powers to the EU and other regional institutions, on the other hand (see introductory article by Kuhn and Nicoli, 2020). Our contribution explores the path from collective identitiesboth that of elites and of ordinary citizensto the integration of core state powers. We argue that, irrespective of functional demands for cooperation and integration, identity politics as well as the degree to which they resonate with the public are crucial ingredients of processes of regional integration. Moreover, identity politics can work both ways, in favour of and against the transfer of core state powers and integration. Firstly, collective identification and mutual trust among political elites enables them to overcome the collective action problems associated with political integration at the regional level and the transfer of core state powers to supranational institutions. At the same time, the lack of a sense of community among elites might still enable some sort of (intergovernmental) cooperation, but it is likely to inhibit integration. Secondly, elite identity narratives linking the nation-state to regionbuilding need to resonate with citizens in order to generate diffuse support (Easton, 1965) for regional integration. At the same time, exclusive nationalist identities can also be mobilized against regional integration, as we are currently witnessing in Europe and in other parts of the world.
Our contribution proceeds in two steps. Firstly, after having clarified our key concepts, we concentrate on the European experience. We argue that elite identities have been crucial for the evolution of European integration, from the beginnings during the 1950s to the Maastricht Treaties. With regard to citizens, European integration has been brought about by the permissive consensus of EU citizens with inclusive national identities (that is, with Europe as a secondary identity). Most recently, however, the politicization of EU affairs in many member states after the Euro crisis on has been driven by populist eurosceptical forces that have been mobilizing considerable minorities of citizens holding exclusive nationalist identities. Thus, identity politics can also work against regional integration. The mobilization of exclusive national identities maps unto a realignment of political forces alongside a cultural cleavage of cosmopolitan versus exclusive nationalist attitudes, which is discernible across Europe.
Secondly, we discuss the extent to which insights from Europe travel to other regions of the world (see also Checkel, 2016). Here, the empirical evidence on the relationship between community-building and region-building is scanty. However, and comparable to the European experience, elites involved in region-building in Latin America, the Middle East and sub-Saharan Africa (less so in North America) almost always develop identity narratives linking their national experience to the respective regions. Moreover, evidence fromprimarilythe World Value Surveys (WVS) suggests that the difference between inclusive national identities (allowing for secondary regional identities), on the one hand, and exclusive nationalist identifications, on the other hand, travels beyond Europe. What is more, the latter have recently been mobilized by the likes of Donald Trump in the USA Rodrigo Duterte in the Philippines and now Jair Bolsonaro in Brazil. This suggests that the cultural cleavage structure underlying these mobilization strategies may also be valid in other world regions. We conclude with some suggestions for further research.

I. The Transfer of Core State Powers, Identity Politics, and the European Experience
By transfer of core state powers, we refer to two features of (regional) integration (see also the article by Bremer, Genschel andJachtenfuchs, 2020, andJachtenfuchs, 2016). Firstly, the core policies of the state such as defence and monetary and economic policies are communitarized in a multi-level system of governance. 1 Secondly, the core competencies of the state are supranationalized through a system of pooling and delegating powers to the regional level (Hooghe and Marks, 2015). States thus voluntarily accept intrusions in their Westphalian sovereignty, that is, deep interference into what used to be their domestic affairs, be it in policy areas or power resources. Scholars defines this as integration in contrast to mere cooperation, where the ultimate decision-making power still resides with the nation-state.
With regard to collective identities and in line with social identity as well as self-categorization theories (Abrams and Hogg, 1999;Tajfel, 1974), we define collective identities as the way in which individuals relate to social groups, including imagined communities such as the nation-state or Europe (Anderson, 1991, see Risse, 2010, Chapter 1, for details). Social identities are composed of two ingredients, namely the special characteristics of the group, and the delineation of borders between the in-group and the outgroup. In this article, we distinguish further between the social identities that are discursively constructed by political elites and the social identification of citizens with larger imagined communities. Lastly, by identity politics we mean the deliberate use of identity discourses for political purposes, for example, to justify or legitimate policies or the transfer of powers and authority onto regional levels. Of course, identity politics can also be a means to oppose regional integration, for example, through the mobilization of nationalism.
On a purely theoretical level, it is hard to see how regional integration or the transfer of core state powers onto the regional level (as defined above) can be legitimized or sustained without some sense of community and, thus, collective identification among elites and citizens. Two fathers of integration theories -Karl W. Deutsch and Ernst Haas were deeply aware of the relationship between integration and community-building. Haas defined integration as the shift of loyalties to a new supranational centre (Haas, 1958, p. 16). In a similar way, Deutsch's transactionalist approach saw security communities as bound by a mutual sense of community (Deutsch et al., 1957; for a more recent treatment see Hooghe et al., 2019;Kuhn, 2015). However, if identity is used as a constitutive feature or as an indicator of (regional) integration or the transfer of core state powers, we can no longer investigate how the two relate to each other. Thus, in line with the special issue, we distinguish between the two.
The direction of the causal arrows between identity and integration is less clear. Does regional integration lead to collective identification, is community-building a precondition for regional integration, or do we observe mutually reinforcing processes of integration and identification? Let us look at the European experience.
Identity politics, that is, the mobilization of collective identities for political purposes, has been crucial throughout the history of European integration. Until about the 2000s, Europeanized elite identities have been supported by a permissive consensus of inclusive national identities among a majority of EU citizens. We illustrate this point with regard to the beginnings of European integration in the early 1950s and the Maastricht Treaty of the early 1990s, which introduced the single currency and, thus, initiated a most profound transfer of core state powers to the European level. Over the past 15 years, however, we have observed the development from permissive consensus to a constraining dissensus (Hooghe and Marks, 2009). Most recently, this has resulted in the mobilization of exclusive nationalist identities by populist eurosceptical forces in events ranging from the Euro crisis to the migration challenge. This last example serves to show that identity politics can work both ways, in favour of and in opposition to enhanced regional integration. Indeed, as we argue below, the mobilization of collective identities against regional integration has a long history, even in Europe.
Note that we do not suggest that there were no functional demands for European integration, such as security needs since World War Two or economic reasons. As we have argued elsewhere (Börzel and Risse, 2019), elite identities and the development of identity narratives are crucial for the supply side of regional integration and the transfer of core state powers in terms of overcoming problems of collective action as well as to mobilize public support. We are not concerned with whether political elites actually believe in what they are saying. Our argument about identity politics is about discourse, not belief systems or motivations (see Schmidt, 2002Schmidt, , 2008.

The Beginnings: Integrating External Security to Overcome War and Destruction
If we start the history of European integration with the 1951 European Coal and Steel Community (ECSC) rather than the 1957 European Economic Community (EEC), dominant theories of European integration such as neo-functionalism (Haas, 1958) and liberal inter-governmentalism (Moravcsik, 1998) have a hard time accounting for it. They both assume that economic interdependence serves as a major driver for regional cooperation and integration. In the case of Europe, however, economic interdependence took off only after the first steps towards integration have been taken (for details see Börzel and Risse, 2019). It took until the 1960s for Europe to reach the level of commercial and financial interdependence that had existed on the eve of World War One (Graph et al., 2013).
The history of the ECSC demonstrates that sharing a common history of destruction relating to not one but two world wars united the discourse of the founding fathers of the European integration project (see also contribution by Hoffmann and Mérand, 2020). In their pursuit of a united, peaceful and prosperous Europe, Konrad Adenauer, Alcide De Gasperi and Robert Schuman drew on the European peace initiatives promoted by the Pan-European and other movements founded in the 1920s or the European union, which French Foreign Minister Aristide Briand proposed to the General Assembly of the League of Nations, with the support of his German counterpart, Gustav Stresemann, in September 1929 (Stevenson, 2012;Loth, 2015, pp. 1-19).
Twenty years later, French foreign minister Robert Schuman responded to German Chancellor Konrad Adenauer's proposal to mutualize heavy industry in the Ruhr with his plan for a coal and steel community as a way of preventing further war between France and Germany (Loth, 2015, pp. 20-36;Patel, 2018, pp. 75-76). The Schuman Plan presented on 9 May 1950 placed the Franco-German production of coal and steel under a common high authority to 'make it plain that any war between France and Germany becomes not merely unthinkable, but materially impossible'. 2 In other words, the narrative surrounding the ECSC was not only and not primarily about fostering economic integration but to cement peace between two historical enemies. This was to be achieved by a supranational framework, open to other European countries, which involved the substantial integration of core state powers in the area of security policy. The next step towards securing peace in Europe, consequently, was not a common market for goods but the European Defense Community (EDC) that would place 'army, weapons, and basic production under a common sovereignty at the same time' (Monnet, 1976, p. 401). France hoped to put the rearmament of West Germany under the control of the EDC, which would join the ECSC under the umbrella of a European political community. This provoked the counter-mobilization of more nationalist French identities in the Assemblée Nationale emphasizing the loss of French sovereignty. Only when the ratification of the EDC failed in 1953 did integration efforts shift to the realm of 'low politics'. Four years later, the Treaty of Rome established the European Economic Community (EEC) and the European Atomic Energy Community (Loth, 2015, pp. 36-74).
In short, the beginnings of European integration were originally driven by attempts of interwar and postwar elites to integrate core state powers over external security. Their attempts were legitimized by their shared narratives of overcoming a common past of war and destruction (see also Hofmann and Mérand, 2020;Patel, 2018, pp. 88-90). This identity construction based on 'othering' Europe's own past has continued to shape the integration of core state powers for more than 60 years (for details see Risse, 2010, Chapter 3).

The Euro: Integrating Monetary Policy to Advance the Political Union
Forty-one years after the Treaty of Paris that had established the ECSC and 35 years after the Treaty of Rome, the EU took another giant step toward further integration when the 1992 Maastricht Treaty established the Economic and Monetary Union (EMU) and, thus, the single currency (see Risse et al., 1999;Risse, 2003 for the following; see also Kaelberer, 2004). Germany and France had been at the forefront of those promoting a single currency ever since the Single European Act had come into force. The UK, in contrast, remained on the sidelines and opted out of the EMU at the negotiations on the Maastricht treaty (McNamara, 1998;Moravcsik, 1998, Chapter 6;Verdun, 2000). How can we explain this difference in attitudes? In the following, we concentrate on Germany and the UK (for France see Risse et al., 1999).
Neither economic nor geopolitical or security reasons can account for the variation in elite attitudes toward the single currency in the UK and Germany. The UK was as integrated economically with the continent during the late 1980s and early 1990s as other countries that joined the single currency. Moreover, it could have easily met the Maastricht convergence criteria, and its government shared a neoliberal economic ideology (details in Risse et al., 1999, 160; for the following paragraphs see Börzel and Risse, 2019;Risse et al., 1999, pp. 159-163). The British attitude toward the single currency remained the same over two decades. At the Maastricht summit the British government reserved the right to decide for itself whether the UK would join EMU in 1999. The Labour government under Tony Blair confirmed this position and decided that the UK would continue with the 'wait and see' attitude of its predecessor. While the few British proponents of the euro used interest-based arguments to support their claims, conservative eurosceptics routinely invoked identity-related statements to justify their opposition to the EMU arguing: '[A]bolish the pound and you abolish Britain' (Redwood, 1997, p. 19). Their discourse closely resembles that of the Brexiteers two decades later.
In the case of Germany, its government had agreed to the EMU early on and stubbornly supported the euro throughout the 1990s. The German government never wavered in its support for the single currency, even though most German economists objected to the euro on economic grounds (Risse et al., 1999, 150; for the following see ibid., pp. 163-169). Chancellor Kohl framed the single currency as the symbol of European integration and he identified his political fate deeply with the realization of the euro. He also labelled 1997the year of reference for the fulfillment of convergence criteriaas the key year of Europe, as being existentially necessary for further integration. He even argued that the success of the EMU was a matter of war and peace. 3 In essence, Chancellor Kohl framed the issue in the German political discourse by constructing a powerful equation linking support for the euro to German identity, based on a rejection of the German militarist and nationalist past. By mobilizing Europeanized identities, Kohl managed to overcome the considerable opposition to his policiesthat existed in his own party, among most economic experts, and among the social democrats in the opposition. Part of the narrative of opposition was also framed in more nationalist identity terms, namely Deutsche Mark patriotism.
However, Kohl's framing of the issue also served to silence political discourse on the EMU. It was no longer possible to argue about the pros and cons of a single currency and to weigh up policy alternatives in a neutral way. As a result, even those opposed to the EMU did not dare to address the German consensus on European integration, but framed their criticism in terms of asking for a delay or demanding the strict application of the convergence criteria. In this case then, the Europeanization of German identity largely shaped the definition of economic interests.
In sum, discourse on the euro in major EU member states was framed to a large degree in terms of identity politics and political visions of a European order. Supporters of the project shared a common idea of European integration as a modernization project that would overcome the historical divisions of the continent (Jachtenfuchs et al., 1998). They used the single currency as a means to get closer to this political vision. The euro symbolized a collective European identity, while the Deutsche Mark was constructed as symbolic remnant of a nationalist past.
There are good functional reasons for first instituting the single market and afterwards a single currency. However, the supranationalization of core state powers in the security realm and with regard to heavy industries (in the early 1950s, just 6 years after World War Two) and over monetary policies in the 1990s necessitated a sense of community among political elites that could enable them to supply the demand for integration by pooling and delegating authority at the EU level. Europeanized identity narratives constructing Europe's own past of nationalism and wars as the 'other' signaled their preparedness to overcome the considerable problems of collective action involved in supranational integration. At the same time, these discourses served to justify and legitimize European integration to their citizens.
The British opposition to the euro confirms our point. The UK had as many good as bad economic reasons for joining the single currency. Yet, pro-European British elites never used identity language to justify British membership of the EU, but reverted to economic (and political) interests in their public discourse. In contrast, British opposition to the EU, from the euro to Brexit, has always been framed in terms of an exclusive nationalist identity (Risse, 2010, pp. 81-86). We suggest that this, to a large degree, explains why the UK has always been on the sidelines of further transfers of core state powers to the EU level.
However, how and to what extent have the various elite discourses resonated with public opinion and citizens over time? After all, identity politics is meant to create diffuse support and legitimacy for policy decisions, in this case the transfer of core state powers to the EU.

The Permissive Consensus and the Europeanization of Citizens' Identities
European political leaders have been able to invoke identity discourses to legitimate moves towards an ever closer union because they resonated with mass public opinion.
Up until the early 2010s European elites routinely used identity-related arguments to silence major debates about European policies and European integration in general in the various national public spheres. As Milward has argued, European integration strengthened national executives by shielding policy-making in Brussels from national public scrutiny (Milward, 1992). Yet the silencing mechanism worked only because of the permissive consensus in favour of European integration among the publics in most member states (Hooghe and Marks, 2009). The support of most EU citizens has hinged on their Europeanized national identities whereby 'Europe' was added as a secondary identification to citizens' national identities. Over the past three decades a plurality of citizens in most member states have held Europeanized identitiesexcept for the UK. Figure 1, based on Eurobarometer data, demonstrates that the main divide in mass public opinion is between those who include Europe into their national identity (inclusive or Europeanized nationalism) and those who exclusively hold national identities (exclusive nationalism; for a detailed discussion see Risse, 2010, Chapter 2). If we add those citizens who claim Europe as their first or even exclusive identity (very small minorities), those with some degree of European identity make up more than 60 per cent of the population on average across member states. The numbers have not fluctuated much over the past 25 yearsand there is little reason to assume that the picture was any different prior to 1992 (for most recent data showing a similar picture, but with national variation see Krastev et al., 2019, p. 5).
Europeanized identities correlate strongly with other attitudes, among which are support for European integration (Citrin and Sides, 2004;Hooghe and Marks, 2005;Kuhn and Stoeckel, 2014;McLaren, 2006) and 'solidarity among strangers' (Habermas, 2006) in terms of the preparedness to support redistributive policies across the EU (see Kuhn et al., 2018 for details; see also contributions by de Vries; by Nicoli, Kuhn and Burgoon;and by Karstens, 2020). In contrast, exclusive national identities go together with opposition to European integration as well as hostile attitudes toward migrants and foreigners.
We argue that the permissive consensus that has been widely reported in the literature, was at play up to about the mid-2000s and that the Europeanized identity discourse of the political elites reported above tapped into this consensus and, thus, was able to generate successfully diffuse support for European integration among most citizens in most member states. This has dramatically changed since then.

The Constraining Dissensus and the Politicization of Exclusive National Identities
For a long time, identity politics facilitated and legitimized the integration of core state powers. This started to change in the 2000s when European integration became more salient and was contested in the member states. France and the Netherlands rejected the Constitutional Treaty in 2005 following negative majorities in public referenda. In other member states, too, the permissive consensus gave way to a constraining dissensus (Hooghe and Marks, 2009), which has prevented the further integration of core state powers in response to the euro crisis and the challenges of mass migration.
As we have argued elsewhere in more detail (Börzel and Risse, 2018), populist eurosceptical parties and movements, particularly on the right, have increasingly succeeded in mobilizing citizens with exclusive national identities. They have deliberately used identity politics to turn latent attitudes among citizens into manifest political behaviour. Rather than creating anti-EU sentiments or changing collective identities towards a rise of exclusive nationalism, eurosceptical parties have tapped into and mobilized pre-existing attitudes among the considerable numbers of European minorities into protesting against and voting against EU policies and institutions.
Public discourses about refugees in particular were driven less by economic or political issues. They should be understood as a clash of competing European and national identities. While this intensive politicization might be new, debates about immigration have always been about 'the other within' (Risse, 2010, pp. 222-224) pitting modern liberal Europe as a multicultural entity that is tolerant towards people of different religions, races and cultural backgrounds against nationalist Europe, which is openly hostile to non-European immigrants. Thus, the discourses about migrants and refugees were largely framed with regard to the in-group/out-group dimension of collective identities (see introduction to this issue by Kuhn and Nicoli). De Vries and Edwards have argued in this context that 'extremist parties on the right tap into feelings of cultural insecurity to reject further integration and to defend national sovereignty from control from Brussels. These parties mobilize national identity considerations against the EU' (De Vries and Edwards, 2009, p. 9). There is ample empirical evidence that support for right-wing eurosceptic and populist parties across Europe is driven by exclusive nationalism and culturally based anti-immigrant attitudes (see Dunn, 2015;Lucassen and Lubbers, 2012;Werts et al., 2013).
Moreover, the distinction between inclusive and exclusive Europeanized national identities maps unto a cultural cleavage that increasingly structures European party systems (see Marks, 2009, 2018;Hutter et al., 2016;Zürn and de Wilde, 2016). 4 Scholars have labelled the cleavage differently, but they all refer to some degree of openness, internationalism and cosmopolitanism at the one end, versus closeness and exclusive nationalism, on the other end. This cultural cleavage has led to a realignment of political forces in Europe and elsewhere. It is orthogonal to the conventional socioeconomic cleavage (left versus right) and has replaced the religious cleavage (Catholicism versus Protestantism) that Stein Rokkan and others explored in Western Europe of the 1950s and 1960s (Lipset and Rokkan, 1967;Rokkan, 1970). For illustrative purposes, we have analyzed various parties and movements in EU member states (see Figure 2). The current identity politics in Europe maps on the cultural cleavage. The Europeanization of citizen identities has not changed. If anything, it has increased in recent years. What has changed is the political mobilization of exclusive nationalist identities by mostly right-wing and nationalist forces (the south-eastern quadrant in Figure 2), while the counter-mobilization of pro-integration forces has been much slower (Macron's En Marche movement in France; the anti-Brexiteers in the UK; the recent rise of the Greens in the polls in Germany). However, the 2019 European Parliament elections seem to have mobilized voters on the cultural cleavage, strengthening both green and liberal as well as authoritarian nationalist forces.
To conclude: the recent European experience shows that identity politics can be both a facilitator and inhibitor of the integration of core state powers. Identification levels with Europe among both elites and citizens have remained largely constant over the past decades. What has changed over time is that political elites have invoked identity constructions to sway public opinion not only in favour, but also against the transfer of national sovereignty to the European level. Moreover, the influence of identity politics on integration is particularly powerful when it relates to the constitutive dimension of the in-group/ out-group distinction. 4 The cultural cleavage is labelled differently in the literature (see Hooghe and Marks, 2018;Hutter et al., 2016;Zürn and de Wilde, 2016). We submit that they all refer to the same cultural cleavage. Identity politics and regional integration

II. Identity Politics beyond Europe: Is There a There There?
To what extent do identity politics help explain the transfer of core state powers to regional levels elsewhere in the world? Of course, the EU is unique in terms of both its scope and level of supranational integration. At the same time, other regions have moved forward with the transfer of core state powers. Eurasia (Hancock and Libman, 2016), sub-Saharan Africa (Hartmann, 2016), Latin America (Bianculli, 2016) and south-east Asia (Jetschke and Katada, 2016) are cases in point. In contrast, North America, the Middle East and east Asia are characterized by mostly intergovernmental regional cooperation schemes, including the North American Free Trade Area (NAFTA), now the US-Mexico-Canada Agreement (USMCA), the League of Arab States, the Gulf Cooperation Council, or the absence of even low-level free trade agreements (east Asia). What is the role of identity politics with regard to these various regional integration schemes as compared to the more intergovernmental cooperation? Does identity politics matter when we take a broad perspective of comparative regionalism?
One problem of establishing causal linkages between the regionalization of identities and regional institution-building is the lack of empirical data permitting cross-regional comparisons. The literature is rather sketchy and, if authors address identity-related questions at all, they are often centred on their particular world region (for an excellent overview see Checkel, 2016).
However, there is evidence that corroborates our arguments that • the relevance of pro-regional elite identity discourses for region-building efforts, • the use of identity narratives by elites that resonate with larger publics, • andmost recentlythe counter-mobilization of exclusive nationalist identities by (right-wing) populist forcesmight also hold beyond Europe. Let us now address each point in turn.
The Ubiquity of Elite-driven Regional Identity Narratives To begin with, efforts at region-building are usually accompanied by social constructions of regional identities. Yet, as Checkel points out, elite discourses establishing regional identity narratives are one thing, but claiming causality between regionalized elite identities and regionalism in terms of the transfer of core state powers to regional institutions is much more demanding (Checkel, 2016, pp. 561-564). Acharya has probably made the strongest claims for this connection with regard to the south-east Asian experience (Acharya, 1997(Acharya, , 2004(Acharya, , 2009; see also Katzenstein, 2005). South-east Asian elites developed the narrative of the Association of South-east Asian Nations (ASEAN) way of emphasizing diversity to legitimize ASEAN. The ASEAN way exemplified identity construction by an elite that was explicitly set up against the European experience (as the 'other') of strongly legalized as well as supranational regionalism (for a detailed discourse analysis, see Yukawa, 2018). The emphasis on preserving sovereignty and diversity as well as on informal networking and communication is enshrined in ASEAN institutions. Diversity was also used initially as an argument for tolerating various types of regime. In the meantime, ASEAN has incorporated some human rights instruments (Jetschke, 2015). Acharya's account probably comes closest to our argument about how the identity discourses of political leaders signal their preparedness to commit to region-building to their counterparts in other countries and, at the same time, how they are used to legitimize region-building to their citizens. Acharya's argument is corroborated by Katzenstein and Hemmer's explanation why there is no NATO in south-east Asia (Hemmer and Katzenstein, 2003). Accordingly, the USA preferred multilateralism in Europe based on a strong sense of community, while it opted for bilateral security ties with Asian states during the Cold War, in the absence of collective identities. Acharya's (and Katzenstein's) claims receive further support by the consistency between the identity discourses of the elite and the institutional design of ASEAN and other regional institutions. In the meantime, however, the discourse appears to have changed considerably (see Yukawa, 2018 for details), in line with the gradual evolution of ASEAN's institutional design, which emulated many parts of EU institutions and extended ASEAN's reach toward ever more areas of policy but with limited supranationalism (for details see Jetschke andMurray, 2012, andJetschke andKatada, 2016).
A similar story related to identity can be told with regard to Africa. The Organization of African Unity (OAU) was created based on a strong postcolonial elite identity emphasizing African independence and strong non-interventional norms (Checkel, 2016, pp. 562-563;Williams, 2007; see also contribution by Eze and van der Wal in this special issue). From the beginning the OAU charter contained the slogan 'Try Africa First' in its preamble. There was a strong sense that postcolonial Africa needed to take its fateboth development and security problemsinto its own hands. After the Somalia disaster in the early 1990s, Ghanaian economist Ayittey coined the phrase 'African solutions to African problems' (Ayittey, 2010) which then became the slogan of the newly founded African Union (AU) in 2002. AU interpreted the postcolonial norm of sovereignty differently from the OAU. Rather than relying on external interventions to solve its security problems, the AU embraced the possibility of military intervention in its member states to deal with war crimes or coup d'états (see Tieku, 2004;Söderbaum, 2004;Williams, 2007). Once again, the elite discourse on these changes is consistent with our argument that identity narratives signal credible commitments to cooperation and integration to other states and are also meant to sway public opinion in the desired direction.
Postcolonial experiences also appear to have shaped the discourse surrounding the formation of the Andean Community of Bolivia, Colombia, Ecuado and Peru in 1969. Tussie has argued that a strong sense of otheringagainst the USA and the neoliberal Washington consensusaccompanied the formation of the Southern Common Market (MERCOSUR) in 1991 (Tussie, 2009;overview in Bianculli, 2016). Likewise, leftist elites that came into power in the 2000s promoted the Bolivian Alliance for the Americas (ALBA) and the Union of South American Nations (UNASUR) as post-neoliberal integration projects (Bianculli, 2016). Even the Eurasian Economic Union (EEU) was legitimized in terms of identity as a regional alternative to the EU, notwithstanding the fact that it copied the EU's institutional design (see Hancock and Libman, 2016).
In contrast, the lack of a collective elite identity may explain the absence of regionalism in east Asia despite their high degree of economic interdependence and their manifest security dilemma. Unsolved issues of historical justice and restitution have prevented China, Japan and South Korea from forming a regional identity, for example, based on a shared memory, which could have provided the necessary trust for building regional institutions (Ikenberry and Moon, 2007;Morris-Suzuki et al., 2013; see also Börzel and Risse, 2019). The North American Free Trade Agreement (NAFTA, now US-Mexico-Canada-Agreement) is also consistent with our argument, as there have been no elite attempts to construct a North American identity or to transfer core state powers to the regional level (Duina, 2016).
In the other cases of regional integration, the transfer of core state powers to regional institutions was accompanied by elite discourses establishing regional identities. Moreover, the institutional design of the regional institutions (strong supranationalism in the cases of sub-Saharan Africa and the Andean Community, less so in the cases of ASEAN, MERCOSUR, UNASUR and EEU) is at least consistent with identity narratives. Last not least, as we have argued elsewhere (Börzel andRisse, 2016, 2019), standard theories of international cooperation and integrationwhether neoliberal institutionalism (Keohane, 1989), liberal intergovernmentalism (Moravcsik, 1998) or neofunctionalism (Haas, 1958) do a rather poor job in accounting for the beginnings of regional institutional-building in south-east Asia, sub-Saharan Africa or Latin America. These theories posit the existence of a causal path from economic interdependence via solving likely conflicts, enabling (further) economic exchanges, and insuring credible commitments to regional institution-building (Haas, 1958;Mattli, 1999;Moravcsik, 1998;Stone Sweet and Caporaso, 1998). Yet, in south-east Asia, Africa and Latin America, there was only limited intra-regional economic interdependence when the various regional institutions emerged. At the least, we have to add interdependence for security provision to the equation to gain causal leverage (Börzel, 2016). This helps account for the sub-Saharan African experience (Hartmann, 2016) and partially for the south-east Asian one (Nesadurai, 2008).
In sum, we observe identity discourses referring to particular regional experiences and histories in almost every instance of regional cooperation and integrationfrom Europe to south-east Asia, sub-Saharan Africa and Latin America. We grant that the causal link between elite discourses and region-building is sometimes hard to establish and has to rely on correlational evidence as well as the rejection of alternative explanations (for example, economic interdependence). However, and similar to the European experience, regional elites and leaders appear to believe that their identity narratives resonate with citizens. Otherwise, they could simply justify their decisions on economic (fostering economic interdependence) or security grounds (dealing with negative externalities of violent conflict, for example, in sub-Saharan Africa). In this context, it is also interesting to observe that elite identity narratives with regard to region-building are established irrespective of regime type. Democratic and authoritarian leaders alike have invoked pan-Africanism, pan-Arabism, the Asian way, or Eurasian identity discourses.

Regional Identities and Mass Public Opinion
Is there any evidence that elite discourses resonate with ordinary citizens or mass publics? What about a sense of regional community among citizens? Unfortunately, we lack the kind of sophisticated data for other regions that we now have on Europeanized identities among citizens in the EU. This is a huge lacuna awaiting further research. In the meantime, we would like to point to at least some studies suggesting that Europe and the EU are not so special after all.
Firstly, Roose used 2003 data of the International Social Survey Programme to compare levels of regional identification in Europe with that in other world regions (Roose, 2013). The data for the EU15 show that 54.2 per cent of respondents feel close or very close to their continent. Respective mean values for three South American countries are 61.7 per cent and for South Africa even 70.1 per cent, but for three east Asian countries this is only a meagre 34.8 per cent (Roose, 2013, p. 287). Note that there are almost no regional institutions in east Asia (see Roose, 2013, p. 287). What is more, Roose finds that those identifying with their region share socio-structural characteristics across continents; namely, that 'people with better professional positions, higher income and better education tend to identify more with their continent' (Roose, 2013, p. 292).
Secondly, the World Value Survey (WVS) provides another dataset we can use to estimate the degree with which people identify with their region. The WVS data has sparked quite a debate on explaining the degree to which citizens across the world hold cosmopolitan or supranational identities (see for example, Jung, 2008;Norris, 2000;Pichler, 2012). To begin with, the studies confirm the socio-structural characteristics of those identifying with their continent or with the world at large, with some important differences. While Norris, 2000, points to age differences, Jung, 2008, argues in favour of a life cycle effect; on data for Europe see Fligstein (2008). Jung, in particular, points out that most people across continents identify predominantly with their nation-state, while those with supranational identities are distinct minorities (approx. 25 per cent on average worldwide) almost everywhere (Jung, 2008). Pichler demonstratessomewhat paradoxicallythat cosmopolitan orientations are particularly strong in (non-Western) countries and continents that are less globalized (Pichler, 2012). Last, not least, these studies confirm findings from European surveys that citizens holding transnational identities beyond their nation-state are also more liberal in their attitudes toward foreigners than those with exclusive national identities.
A major weakness of these studies is that they do not take into account the fact that most people hold multiple identities, even with regard to the territorial dimension. That is, people can strongly identify with their nation-state and with their region. As argued above, the main division in Europe is between those holding exclusive nationalist identities and those adding Europe as a secondary identity; citizens who identify only with Europe form a rather small minority almost everywhere in Europe. Yet the studies quoted above mostly focus on those citizens with strong supranational identities (except for Jung, 2008). The findings are, thus, less relevant for cross-regional comparisons with the European findings.
The closest one can get to cross-regional data resembling the Eurobarometer surveys cited above are cross-tabulations from two WVS questions (see Figure 3). The WVS 2010-14 wave asked in almost all regions whether citizens saw themselves as part of their country as well as their respective region. 5 The WVS even used the various regional organizations in their questionnaire, which, of course, raises the question of whether respondents were aware of their existence. In any event, the data show that Europe is not peculiar. In almost all regions of the world, large majorities of citizens feel part of a regional community. What is more, most people hold dual identities in almost every region. That is, they strongly identify with their nation-state and with their respective region or regional organization. Figure 3 plots the regional identification levels for those in the selected countries who also strongly identify with their own country. Overall, the numbers are rather similar across regions (Egypt constitutes the one exception). 6 The data for European, Latin American and south-east Asian countries are rather similar with majorities of 50 to 80 per cent strongly identifying with their nation-state and with their respective regional community. In the case of selected African countries, the numbers are even higher (see dark grey and grey bars for those who strongly agree in Figure 3). In contrast, exclusive nationalists (those who strongly identify with their country, but do not feel part of their region, make up between 20 and 30 per cent across countries. Once again, Europe is not exceptional in this regard. We can use these data as a preliminary indication that inclusive national identities (one's own nation-state plus region) are common across the globe and that large majorities everywhere identify with their country and their region. Moreover, those holding exclusive national identities (those disagreeing with the statement about their region in Figure 3) are visible minorities almost everywhere. If further evidence corroborated these findings, we would be able to conclude that elite identity constructions with regard to region-building do indeed resonate with wider publics beyond Europe. As a result, we should be able to find a permissive consensus in support of regional integration beyond Europe so that the elite identity narratives, which we found for region-building, would indeed resonate with mass public opinion elsewhere in the world. 6 One explanation may be that Egyptians did not know about the AU when they were asked about it.

Mobilization of Exclusive Nationalism beyond Europe
If pro-regional identifications can be used to legitimize regionalism and the transfer of core state powers outside Europe, what about the counter-mobilization of exclusive nationalism by populist forces? Indeed, the onslaught on global and regional governance is not confined to populist organizations in Europe. There are some indications that identity politics serves as a mobilizing force beyond Europe. US President Trump and his attacks on globalism and the liberal international trade order, but also on regional cooperation schemes, such as NAFTA, together with his withdrawal from the Transpacific Partnership Agreement, serve as prominent examples. Trump and his followers in the Republican Party map perfectly on the south-eastern corner (right-wing exclusive nationalism) of the cleavage matrix (see Figure 2). The same holds true for Brazilian president Bolsonaro and for Philippine President Duterte. Moreover, this type of populism is not confined to democratic systems, as the examples of Vladimir Putin in Russia or of Recep Tayyip Erdoğan in Turkey document. In each of these cases, right-wing political elites did not create but mobilized exclusive nationalist identities, thereby turning them from attitudes into political behaviour. Last not least, there are also some left-wing populist governments (such as that of Maduro in Venezuela or of Morales in Bolivia) trying to mobilize nationalist forces on the left against regional integration. These forces are shown in the south-western corner of Figure 2. They all constitute part of a larger attack on the liberal international order of which regional integration is part and parcel.
The evidence presented here is only illustrative. The examples mentioned indicate that our argument on the possibility of mobilizing popular identities in favour of and against regional integration holds beyond Europe. Future research is needed to explore the relationship between elite narratives, mass public identification levels, and the mobilization of the latter in favour of and against regionalism, as well as the overall effects of these forces on region-building.

Conclusions
This contribution has focused on the causal path leading from identity to regional integration in terms of the transfer of core state powers to regional institutions as well as the pooling and delegation of authority at regional levels. We have argued that Europe and the EU are not unique as far as the identity-region-building conundrum is concerned. Elite identity discourses constructing regions appear to be ubiquitous even in North Africa and the Middle East, where they have not resulted in thick regionalism. Citizens around the globe appear to hold multiple identities, being able to identify with their nation-state and their respective region at the same time. We can also observe the political mobilization of identities and of identity politics beyond Europe.
At the same time, we have suggested that the functional story of regional integration starting with (economic or security) interdependence, then the transfer of core state powers to regional institutions and, ultimately, the emergence of regionalized identities (whether via Haas' neofunctionalism or Deutsch's transactionalism) may have to be reversed (see also Hooghe et al., 2019). Evidence from Europe, south-east Asia, sub-Saharan Africa and Latin America suggests that the attempts at regional institutionbuilding by elites are usually accompanied by identity narratives linking national to regional communities. Moreover, regions themselves are not pre-existing territorial entities but social constructions that pertain to some adjacent territorial spaces. In other words, elite identity narratives constitute these regions in a fundamental sense. Functional demands for regional integration notwithstanding, we argue that elite identity narratives are necessary to supplying regionalism in the sense of communicating credible commitments to other political leaders on regional integration. At the same time, these identity narratives are also necessary to generate citizens' support for integration. While elites do not create regional identities among citizens, they tap into pre-existing community orientations, reconstructing and localizing them into coherent identity narratives that link the respective nation-state to the region. Only then can regional institutions emerge to serve functional needs for economic or security cooperation.