A member state like any other? Germany and the European integration of core state powers

ABSTRACT The EU has integrated core state powers in a largely unsustainable manner. Why is this? In this introduction to a special issue on Germany, we take an in-depth look at national preference-formation. We trace the impact of state elites, as emphasised by functionalist theories, and mass publics and political parties, as stressed by postfunctionalism. We find that across policy fields and with striking continuity over time, Germany acts as a normal member state. The country prefers the regulation of national capacities over the creation of European capacities, and (increasingly) the intergovernmental rather than supranational control of those capacities. Only in existential crises, Germany supports European capacity-building under intergovernmental control. This leads to unstable integration but is not an indicator of hegemonic dominance. Crucial from both a practical and theoretical perspective, there exists no major gap between state elites and political parties or public opinion on German preferences.


Introduction
German preferences and actions towards the European Union have long puzzled scholars and political observers. One group regards Germany as the hegemon and true ruler of the EU, imposing its institutional preferences, economic model and foreign policy positions on the other member states and the European Commission. Another group sees Germany as a compulsive pro-European which defines its interests in European terms and stands ever ready to support a common European solution, if needed with billions of Deutschmarks or Euros.
The disagreement over the normative underpinnings and substantive direction of German policy towards the EU is linked to theoretical debates about the drivers of European integration. For functional theories such as neofunctionalism or liberal intergovernmentalism, the EU is an institution for solving collective action problems and increasing domestic gains (Schimmelfennig 2014;Moravcsik 2018). In this perspective, rationally egoistic governments scale up governance functions to the EU because it pays off with their voters, by creating more jobs, catching more criminals and a better protection against security threats. For postfunctionalism, the logic of community conflicts with the logic of interest (Hooghe and Marks 2009). Voters also care for community, particularly if supranational governance involves redistribution and sacrifice, and challenger parties give voice to these concerns and exert pressure on mainstream parties. In this perspective, rationally egoistic governments try to assure their re-election by relinquishing potential gains from the scaling up of governance functions to the EU if audience costs are too high.
Governments thus have to choose which of the two logics they should follow. Their choice may depend on the policy field, issue salience, the strength of sovereigntist challenger parties, or patterns of asymmetric interdependence, to name but a few. When the issue is technical, low salience, no challenger party in sight but only huge gains from a common European solution, the choice for scaling up is easy. When the issue is close to the core of sovereignty, of high salience, mobilized by strong challenger parties and involves high risks or costs, the choice is much more difficult.
In the post-Maastricht period, the latter type of choices has increased strongly because the EU has moved from market integration into core state powers. In this special issue, we try to assess the influence of functionalist and postfunctionalist factors in explaining German preferences on core state power integration. Consequently, the contributions to the special issue analyse • an extended time period: from Maastricht to the present • different core state power areas: asylum, policing, fiscal capacity, fiscal regulation, defence procurement, security policy • periods and policy fields with differing levels of interdependence and salience • different types of actors: state elites, political parties, public opinion The guiding question of this special issue is to which degree and under which conditions state elites (reacting to functional pressures) and political parties and/or public opinion (reacting to postfunctional concerns of community and sovereignty) shape German preferences on the integration of core state powers. In answering this question, we provide a fresh and systematic look at the old topic of Germany's role in the EU. We also make a contribution to the theoretical debate by introducing and analysing state elites as the carriers of functional interests in core state powers, where business interests are largely absent, and by comparing the respective influence of functional and postfunctional influences on government preferences.
In the following, we discuss case selection and method, in particular the focus on Germany (section 2), our framework for analysing our dependent variable, namely preferences on the integration of core state powers (section 3), our independent variables and hypotheses (section 4), and provide a comprehensive comparative discussion of the findings of our contributions (section 5).
We conclude that Germany is neither willing and able to become a European hegemon, nor is it an unconditional supporter of the EU. Instead, it acts as a member state like any other. When faced with strong sectoral externalities, German governments instrumentally compare the costs of the status quo with the price for change, and seek to minimise the costs and risks incurred from the integration of core state powers. Across core state power areas, German governments thus express a consistent preference for regulation over costly supranational capacity-building. Only in cases in which the creation of some supranational capacities appears unavoidable, primarily in full-blown crisis situations, Germany supports the build-up of state-like capacities on the supranational level, while insisting on intergovernmental control and/or their temporal nature to reduce costs and risks (Fabbrini and Puetter 2016). Regarding the influence of functional or postfunctional factors explaining preferences, we find few instances where functional interests represented by state elites and identity concerns represented by political parties and public opinion actually clash, even though we looked at several broad policy fields over thirty years. Only in security policy, we have a constellation where elites are broadly in favour of more integration whereas mass publics oppose it (successfully). In the other fields, functional and postfunctional forces by and large point into the same direction. Overall, therefore, there is no systematic conflict between elites willing to integrate and mass publics opposing integration in post-Maastricht Germany.

Case selection and method: why Germany?
The literature on Germany's role in the EU exhibits a strong tendency to regard the country as a special case, explaining German government preferences by idiosyncratic factors. Throughout European integration history, this scholarly understanding of a German Sonderweg on Europe came in different forms. One classic strand saw German governments as displaying a particularly strong 'European vocation' (Paterson 2010), rooted in the country's difficult history. Europe, in this sense, served as a vanishing point for a country grappling with its own past. In the immediate aftermath of reunification, the German affection for Europe is frequently argued to have reached its apogee. At the time, Helmut Kohl emphasised habitually that deepened EU integration and reunification were 'two sides of the same coin' and that the Euro was crucial in rendering European integration 'irreversible' (Banchoff 1997). In our view, the explanatory power of such understandings of Germany's role in Europe is dubitable even for the pre-Maastricht era, contradicted for instance by German attempts to restrict the size of the EC budget throughout the 1980s (see Howarth and Schild 2021). They certainly no longer hold now. Indeed, many scholars observed a process of disenchantment and 'normalisation' (Hyde-Price and Jeffery 2001) that followed the post-reunification honeymoon period.
More recent accounts of Germany's role in Europe deviate drastically from this classic view. Two strands stand out. First, under the impression of the Euro crisis, various scholars (Nedergaard and Snaith 2015;Matthijs 2016;Schäfer 2016) have sought to explain German government preferences by reference to ordoliberal convictions. The deepseated German desire for Ordnungspolitik, in this perspective, led the Eurozone into crisis (Young 2014). While there is merit to this explanation, the ordoliberalism thesis does not explain why other Eurozone countries without an ordoliberal tradition hold similar preferences. It also cannot explain why German governments ultimately caved in at each decisive turn in the 'chicken game' (Schimmelfennig 2015) that was the Eurozone crisis, be it on the ESM or the third Greek bailout program. In our view, such behaviour is much more in line with a rationalist strategy of cost-minimization than a doctrinal attachment to ordoliberalism. A second strand enjoying increasing prominence focuses on Germany's domineering role in the EU. Characterisations of Germany as 'semi-Gulliver' (Bulmer and Paterson 1989) or as 'embedded hegemon' (Crawford 2007) have long been a mainstay of IR-inspired thinking on Germany's role in the EU. In more recent years, conceptualisations of Germany as the EU's power-maximising 'hegemonic stabiliser' have gained currency (Donnelly 2018;Webber 2019). In our view, Germany is indeed a pivotal member state which has arguably become Europe's economic and political powerhouse. As the Schengen crisis demonstrated, however, Germany fails to upload its policy preferences in case it cannot act alongside a powerful coalition of member states (see Ripoll Servent and Zaun 2021). In contrast to the hypothetical hegemon described in these scholarly accounts, Germany is unable to unilaterally impose its preferences on the rest of the EU.
This special issue puts systemic factors at the centre of its explanatory approach. Instead of resorting to idiosyncratic factors to understand German preferences on European integration, we contend that the root cause for changing German preferences on Europe lies in the transformation of European integration itself, with core state power integration increasingly becoming the dominant site of EU institution-building. As a consequence of this systemic approach, we use Germany as a test case for the explanatory leverage of general theories of European integration. Our aim is to analyse in which cases and under which conditions functional pressures for managing interdependence, which is the explanatory focus of rational functionalist theories (Moravcsik 2018) and concerns for identity and community, which lie at the core of postfunctionalism (Hooghe, Lenz, and Marks 2019), play a role in the formation of German preferences. Correspondingly, we trace the impact that state elites, on one hand, and mass publics and political parties, on the other, have on German preference-formation. Obviously, focussing on a single case risks making Germany a special case instead of treating it as one case in a class of many. In order to minimize these risks, we include a comparative section in this introduction, and the bulk of the individual contributions to this special issue also address the German position in comparison to other member states. In addition, eyeing long time horizons (if possible since the Maastricht Treaty) allows us to track whether German preferences have actually changed over 30 years. It also avoids overestimating short-term outliers or premature findings of, for instance, a German hegemony in Europe.

Institutional design in core state powers: regulation and capacity
German government preferences on the European integration of core state powers are the dependent variable of the contributions to this special issue. Preferences describe 'an ordering among underlying substantive outcomes that may result from international political interaction, [. . .] [i.e.] a set of fundamental interests defined across "states of the world."' (Moravcsik 1997, 519;cf. Frieden 1999, 42) Core state powers, in turn, are the action resources essential for upholding the core functions of sovereign government (Kuhn and Nicoli 2020, 7-9), that is, to constitute a territorial, positive state (in contrast to a solely regulatory state, Majone 1997) in the Weberian sense. At the very minimum, these comprise the control over the means of coercion (police, military, borders) and the raising and spending of public revenue (monetary, tax and fiscal), both realised by an effective bureaucratic apparatus. Core state powers can be integrated on the European level by means of either regulation or capacity-building (Genschel and Jachtenfuchs 2014). Regulation refers to the setting of authoritative rules for the exercise of a state's action resources. Once majority-voting applies in a given field, we assume regulation to be exercised by the EU-level, either supranationally via the traditional Community method or intergovernmentally by the member states alone. Capacity, in turn, refers to the creation and maintenance of standing action resources. These resources can be situated either on the national or on the supranational level. Governmental preferences on institutional design in core state power integration can take on four distinct combinations of regulation and capacity-building (see Table 1).
Under full member state sovereignty, EU actors' influence is confined to non-binding supporting action. An example is member states' continuous control over their contribution of troops to EU-led military missions. In cases of national control of supranational capacities, the member states back the creation of supranational capacities while retaining regulatory oversight and curtailing the involvement of supranational actors such as the Commission or the Parliament. Among these cases rank, for instance, 'de novo bodies' (Bickerton, Hodson, and Puetter 2015) such as the European Stability Mechanism (ESM), the European External Action Service (EEAS), and some EU agencies (Rittberger and Wonka 2011). In cases of supranational or intergovernmental control of national capacities, the member states resist the creation of supranational redistributive capacities but delegate the exercise of regulatory oversight to the EU-level, either to supranational actors that credibly guarantee mutual commitments or to an intergovernmental process that they regard as better protecting their sovereignty. Whereas the initial Stability and Growth Pact (SGP) conformed to the former variant, the later Fiscal Compact or European Semester verged more toward the latter. Positive state-building, creating sovereign supranational actors that can either replace or compete with existing national capacities, remains relatively rare. To this day, the clearest example remains the European Central Bank (ECB).
During the last decade of crises, the EU often resorted to 'mixed' modes for the integration of core state powers which neither leave national sovereignty intact nor constitute positive state-making. While it is plausible that these modes were chosen because they minimize conflict among member states and with their voters, both the supranational or intergovernmental control of national capacities and the national control of supranational capacities are inherently unstable and unsustainable. The former mode allows for the passing of authoritative rules on member states' conduct. However, as exemplified by the Euro and Schengen crises, it hampers the redistributive risk-and burden-sharing necessary for upholding an equitable distribution of the costs and benefits of cooperation in the face of asymmetrical shocks. The latter mode, in turn, would allow for the sharing of key action resources of sovereign government in times of crisis. Yet, it is ineffective in doing so as it is prone to activating the 'joint-decision trap' which arises in situations of compulsory joint decision-making under unanimity and produces decision-making blockades that lead to suboptimal policy outcomes (Scharpf 2006). In sovereignty-sensitive fields susceptible to national identity politics and mass politicization, the occurrence of this deficiency is even more expectable. Obviously, one should not extrapolate initial preferences from policy outcomes. It is well possible that German governments would have preferred more sustainable institutional designs at various instances. The contributions to this special issue will therefore pay particular attention to whether and when these mixed modes were actually a German preference, or merely the outcome of hard intergovernmental bargaining.

What shapes preferences on core state powers?
While there is wide scholarly agreement that understanding governmental preferences is important for understanding the development of the EU, disagreement persists about the factors shaping those preferences. Theories that share a functionalist understanding of integration, such as neofunctionalism and liberal intergovernmentalism, assume asymmetric interdependence to be the decisive factor underlying interstate bargaining on international institution-building. Postfunctionalism, on the other hand, investigates the effect of citizens' concerns for identity and community on member state preferences. Both factors are of potential relevance for core state power integration, which renders the latter an ideal test case for both theoretical traditions. Strong and asymmetric interdependence in fields such as monetary and fiscal policy or border and migration policy has long underpinned the EU's efforts to integrate these areas. At the same time, these fields are generally perceived as close to the core of national sovereignty, enjoying considerable salience in national public and political debates. Our aim is to assess the rival hypotheses put forth by functionalist and postfunctionalist approaches on the single case of Germany. Correspondingly, we trace the impact of, first, state elites and, second, mass publics and political parties.

State elites
State elites 'are for the integration of core state powers what private business is for market integration: the group of actors most immediately affected and concerned' (Genschel and Jachtenfuchs 2016, 52). The category comprises all unelected professionals that derive their status and income from the handling of national core state powers, primarily civil servants. State elites are not political actors in the narrow sense of the word. While they may hold ideological convictions, they primarily serve the state and the institution that employs them. In the German case, the category comprises administrative officials from the departments responsible for a given core state power area, for instance the federal ministries of foreign affairs, the economy, finance, interior, defence, and the chancellery. The analytical focus on state elites ranks among the novelties and central departures of core state power theory from classic functionalist approaches. The most prominent integration theory anchored in the functionalist tradition, i.e. liberal intergovernmentalism, argues that governments form their preferences largely in response to the interests of domestic commercial coalitions (Moravcsik 1998). In the realm of core state powers, however, it is well possible that economic actors hold preferences on integration that range from diffuse to virtually non-existent (for an exception see Târlea et al. 2019, 39). From the point of view of core state power theory, state elites thus provide the functional equivalent to liberal intergovernmentalism's economic interest groups. In core state power integration, state elites are, first, the actors most likely to hold consistent and intense preferences on both the pooling of competencies on the EU-level and the institutional design of integration. Second, they react to strong and asymmetric interdependence between member states within the respective policy sectors they are responsible for. Third, instead of following a set of ideologically derived core convictions, state elites can be expected to hold a general preference for functional problem-solving. In their view, the pooling of competencies is of no value per se but solely justifiable if it promises significant economies of scale. Fourth, state elites are unlikely to promote a comprehensive integration of core state powers since EU-level capacity-building contradicts their fundamental interest in bureaucratic survival and could ultimately lead to the much-dreaded self-abdication of the state that they depend upon. Instead, they seek to keep the material and sovereignty costs of integration as low as possible.
In case state elites cannot thwart a certain degree of core state power integration due to strong sectoral interdependence, they thus hold an a priori preference for regulation over capacity-building. State elites regard this path as less costly. The supranational or intergovernmental regulation of national capacities avoids the creation of rival capacities and allows national elites to even extend their regulatory influence to the EU-level, primarily via the preparatory bodies of the Council and the encompassing system of comitology. Whether state elites prefer the involvement of supranational institutions for monitoring, enforcement and adjudication purposes, depends on their trust in the supranational agents to credibly and effectively ensure mutual commitments. In that view, the 'Union method' that Angela Merkel conceived during the Euro area crises can be seen as a sign of mistrust toward the enforcement willingness and capacities of supranational institutions (cf. Schoeller and Karlsson 2021).
Only when the problem cannot be solved by either the regulation of national capacities or a return to full member state sovereignty, we expect state elites to advocate supranational capacity-building. Generally, we expect the economies of scale from supranational capacity-building to decrease relative to the size and wealth of a given member state. As a large and relatively wealthy member state, Germany usually retains the capacity to act unilaterally, its economies of scale from supranational capacity-building being relatively smaller. German state elites should thus accept the build-up of supranational capacities solely in cases in which both the costs of disintegration and the costs of maintaining a regulatory status quo in a given field are prohibitive. We expect this to be primarily the case in full-blown crises that undermine an otherwise profitable policy regime or threaten the viability of the EU as a whole. In cases in which supranational capacity-building has strong redistributive implications, member states that expect negative net returns from integration will push for intergovernmental control so as to keep the costs of integration and the risk of moral hazard as low as possible. One example was the creation of the ESM, in which Germany pushed for an intergovernmental governance structure and insisted on a national veto. This allows for the construction of a first hypothesis: (1) In the presence of strongly asymmetric interdependence, state elites will advocate the European regulation of national core state powers.
(2) Only when the costs of regulation appear prohibitive, state elites will advocate the creation of European capacities.

Mass publics and political parties
From a postfunctional perspective, the party-political mobilisation of citizens' concerns about national identity and community is the decisive factor underpinning governmental preference formation. Whereas citizens regarded the market-based EU of the old days with either benevolent indifference or tacit approval, the post-Maastricht integration of core state powers has met fierce identity-based opposition (Hooghe and Marks 2009). Voters, and especially those with exclusive national identities, have increasingly come to see the European integration of policy areas close to the core of national sovereignty as undermining their fundamental desire for self-rule within their political and cultural community. Obliging them to 'look over their shoulders when negotiating European issues' (Hooghe and Marks 2009, 5), citizens' increasingly vocal opposition to the deepening of integration has curtailed governments' room for manoeuvre and affected both their preference formation and bargaining behaviour on the supranational level. Crucially, postfunctionalism argues that these mass politics of European integration trump the influence of functional problem-solvers on national preference-formation whenever 'both come into play' (Hooghe and Marks 2009, 18). The analysis of core state power integration over an extended period of time envisioned by this special issue should allow for an assessment of this claim. While the post-Maastricht integration of core state powers has primarily become associated with the emergence of a constraining dissensus, it is also possible that voters come to support integration. For one, voters could perceive the entailed constraints on national sovereignty 'a bitter but necessary pill to swallow' (Kuhn and Stoeckel 2014, 638), offset by the economic benefits of integration. This tendency was observed among citizens of 'programme countries' during the Euro area crisis (Kuhn and Stoeckel 2014). Secondly, citizens could be 'inclusive nationalis[ts]' (Risse 2010, 41) and hold an identitybased preference in favour of core state power integration, even in cases in which it engenders net economic costs. Among the German general public, for instance, both citizens' self-image and the perceived economic consequences of integration need not be so constraining after all. In recent Eurobarometer surveys, a clear and increasing majority of Germans consider themselves to be European to at least some degree and deem EU membership 'a good thing'. Since 2012, these two survey items have consistently recorded all-time highs, reflecting a general mood among the German public that the EU is an achievement to be preserved, even in trying times, and that Germany has been among its foremost profiteers. Third, not all areas of core state power integration are equally sensitive to concerns for identity and sovereignty. Initial empirical evidence suggests that voters (including those of the populist right) have strongly diverging preferences on core state power integration depending on the issue at hand. Whereas debt relief or refugee relocation tend to be contested indeed, many citizens support integration in other areas such as disaster aid and defence (Bremer, Genschel, and Jachtenfuchs 2020).
Postfunctionalism stresses that attitudes harboured by citizens merely provide a latent potential for political mobilisation, in need of an activation by party-political entrepreneurs and primarily by Eurosceptic challenger parties. When forming their preferences on core state power integration, governments thus do not respond to public opinion directly, but particularly to the positioning of pivotal challenger parties. For the purposes of this special issue, this means that measuring the evolution, salience, and distance of different parties' positions on regulation and capacity-building provides a reasonably reliable measure for the politicisation of core state powers integration, especially in the face of scarce issue-specific and long-term data on public opinion (compare the contribution by Freudlsperger and Weinrich 2021). Mainstream parties, as the parties that routinely alternate between government and opposition (De Vries and Hobolt 2012), have long held integrationist positions in most EU countries (Aspinwall 2002) and are thus most affected by the increasing politicisation of core state power integration (De Vries 2007). In the German case, for instance, the rise of the right-wing challenger AfD can be expected to have had a politicizing effect on mainstream parties', and especially the conservative Christian-Democrats', preferences towards the integration of core state powers (Meijers 2017). The AfD opposes the European integration of core state powers for reasons of national sovereignty and favours stripping the EU of all competencies that reach beyond the internal market. In a politicised environment, mainstream parties would thus look over their shoulders, weigh the audience costs of an integration initiative, and come to support visibly costly initiatives, especially in the form of supranational capacity-building, solely when they can be absolutely sure of citizens' support for integration. Our second hypothesis thus argues: (1) Mainstream parties oppose the integration of core state powers if challenger parties succeed in politicising its material or sovereignty costs among the wider public.
(2) Mainstream parties support supranational capacity-building only if citizens regard the material and sovereignty costs involved as justifiable or even desirable.

German preferences on core state powers: Findings
In the following, we provide an overview and assessment of the findings of the various contributions to this special issue, as summarised by Table 2 below.

State elites
The findings of the contributions in this issue lend strong support to our first hypothesis according to which state elites attempt a European regulation of national capacities in cases of strong asymmetric interdependence and resort to the creation of European capacities only if the costs of such regulation are prohibitive. 'Prohibitive' costs usually meant a complete breakdown of regulation in the respective issue area. Asylum policy is a case in point (see Zaun and Ripoll Servent 2021). Since the 1990s, Germany constantly advocated a European regulation of national asylum systems. Although German strategies changed significantly and frequently over time, the desire to upload German regulations to the EU remained constant over the last 30 years. The underlying goal was to reduce the number of applicants and to shift adaptation costs to other member states.
Only during the 'refugee crisis' of 2015, when the Common European Asylum System (CEAS) had de facto broken down and Greece as well as Italy were manifestly neither willing nor able to uphold its rules, Germany supported a modest reinforcement of the European Asylum Support Office (EASO). Even then, however, Germany sought to keep the costs of supranational capacity-building low. From the German perspective, a reinforced EASO was still a support office for member states with insufficient or dysfunctional national capacities. The German Federal Office for Migration and Refugees was to remain the strong national capacity to deal with refugees, unlikely to ever need EASO's support. German preferences regarding European asylum policy matched those of the strong regulators of Northwestern Europe, i.e. states such as the Netherlands, the UK or Sweden. For a long time, this group of countries had experienced substantive immigration, which led them to develop a dense and elaborate system of domestic regulation, substantive administrative capacity for dealing with immigrants, and considerable practical experience over time. All these countries feared 'secondary movements' of asylum applicants after their entry into the EU, and particularly into the border-free Schengen area because they were (not always correctly) convinced that their asylum systems were generous and attractive. A second group of countries consisted of traditional emigration countries, largely from Southern Europe, which had developed little regulation or administrative capacity to deal with asylum applications because, for the longest time, they had so few.
The Dublin system which stipulates that asylum applications must be processed in the country of first entry puts strong pressures on the Southern countries which were not equipped to deal with a large number of applications. Hence, they wanted to change the status quo. Eastern European countries had equally weak regulation and capacity but also few applicants. While they had to accept the CEAS as part of the acquis when they entered the EU, they had no incentive to change it substantially (Zaun 2017). This preference constellation in which each group of actors seeks to shift the burden to the respective others hands a strong position to Eastern European states able to realize their preferences by simply sticking to the status quo. A similar pattern holds for fiscal policy. German state elites have consistently advocated a European regulation of national budgets since Maastricht (see Schoeller and Karlsson 2021). As in the case of asylum policy, the intention was to upload the German system to the EU level. The goal was to achieve budgetary risk reduction at the national level rather than risk-sharing at the European level. The creation of large fiscal capacities was already resisted before Maastricht, and the rejection of a 'transfer union' which was highly salient during the Eurozone crisis figured already prominently in parliamentary debates in the early 1990s (see Freudlsperger and Weinrich 2021). For German state elites, the EU should remain fiscally weak, and fiscal capacities were to be created only in exceptional, tightly confined cases, with limited means, and preferably for a limited time that is necessary to solve the problem at stake. Since the 1980s already, Germany was at the forefront of those that sought to reduce the size of the EU's multiannual budget relative to the gross national income. It accepted the European Stability Mechanism (ESM) only when the Euro was in existential crisis. It accepted it as a permanent institution solely to the extent it deemed necessary to salvage the currency union and because similar crises were deemed possible in the future. Ideally, however, the ESM should not spend German taxpayer's hard-earned money but lend to states in need only if coupled with tight conditionality. The Franco-German proposal for a recovery fund which led to the 'Next Generation EU' package adopted at the marathon July 2020 European Council under the German presidency is not an indication of a fundamental shift in German preferences. Its debtfinanced grant element is motivated by the largest economic crisis after the Second World War which was symmetric rather than asymmetric and caused by a natural disaster rather than moral hazard, cheaper than the alternative of substantial transfer payments, subject to national parliamentary vetoes, and limited in time and purpose (Genschel and Jachtenfuchs 2021;Howarth and Schild 2021).
At first sight, the preference constellation in fiscal policy is reminiscent of asylum policy. German preferences align with a group of Northwestern states arguing for tight fiscal regulation coupled with little and limited fiscal capacity. Scandinavian countries, Germany, the Netherlands and Austria among others want to upload their regulatory model (and as a consequence their growth model) to the EU, reduce moral hazard and avoid large and most importantly long-term fiscal transfers to the South. Southern member states have increasingly demanded financial support measures after 2010 as they were most hit by the Eurozone crisis. Contrary to asylum policy, however, Eastern European member states sided with the North in their rejection of those mechanisms for fear that financial flows that now go East would become re-directed toward the South. French preferences are almost the direct opposite of German preferences. France has long pushed for financial support mechanisms even before Maastricht (Howarth and Schild 2017). This is exemplified by a long series of proposals from the 1970s, including plans to grant the ESM a banking license, a substantial Eurozone budget, and a recovery fund much larger than the eventual Franco-German compromise of May 2020.
In the field of defence, preferences of German state elites also reflect patterns of interdependence but lead to different positions in defence procurement and in military cooperation. In defence procurement (see Biermann and Weiss 2021), Germany is among the member states with a (relatively) competitive and export-oriented defence industry. It has an interest in creating a supranational armaments agency to open up nationally segmented markets in defence products by means of supranational regulation. France, another country with a strong defence industry, was equally in favour of opening up national markets but without supranational regulation in order to keep control over national defence procurement. Countries with weak defence producers expectedly resisted supranational market-making in order to protect those producers and the jobs attached to them. It is evident, however, that the intensity of German preferences in this sector is much lower than in asylum or fiscal policy, and that a willingness of state elites to create a supranational market is not always greeted with applause by producers (as is illustrated, for instance, by the recent lawsuit of a German shipyard against the decision to have a battleship built by a Dutch consortium, cf. FAZ of 20 January 2020).
A similarly low preference intensity can be observed in military cooperation. To paraphrase Tobias Bunde's analysis (2021) only slightly, German state elites consistently want a European military force that does not do anything. While the security environment is becoming less benign, and the UK, which has for many decades prevented any attempt of creating European military capacities which could be seen as potentially interfering with NATO, has left the EU, this preference for defence integration which is largely symbolic and serves as a political tool furthering European integration rather than military ends has remained constant over many decades. France prefers operative supranational capacities under national control or minilateral (i.e. differentiated) capacities with fewer participants in order to make them operational. Smaller Western states also have preferences for differentiated supranational capacities under national control in specific fields. Eastern European states usually have preferences similar to the traditional position of the UK: NATO is the core of territorial defence, and anything that seems to weaken NATO is rejected. After decades of conveniently hiding behind the UK's strong preference against an EU role in territorial defence, Eastern European countries' apodictic stance now provides Germany with another much-needed cover. In the field of crisis management, German state elites indeed prefer an enhanced role for the EU but their low preference intensity is reflected by a continuous support for regulatory means of integration that keep member states' command of military capacities intact.
The German preference for the creation of a supranational police office is almost the mirror image of its preferences in military cooperation. While the increase of transnational crime since the 1970s created a functional pressure for more intense cooperation among police authorities taken up by all EU member states, the German preference for institutional design in this field has long been an outlier. Whereas most other member states preferred intergovernmental cooperation or at best a supranational capacity under national control and confined to specific issue areas, Germany constantly argued for a broader scope and a true supranational police authority, a 'European FBI' in the words of Helmut Kohl. After Kohl's departure, German governments continued to push for further integration in this field where functional necessity arose. However, attenuating its fervent advocacy for a Europol with operational capacities since the mid-2000s, Germany's position fell more and more in line with other more reticent member states.
Overall, the contributions to this special issue support our first hypothesis: In cases of high interdependence, German state elites consistently supported the supranational regulation of national capacities in the respective fields. Only in extreme cases of a potential system breakdown as during the Eurozone and Corona crises, they began to support the creation of supranational capacities. The sole exceptional case where these conditions were not met is their advocacy for a 'European FBI' -but Europol is also a significantly smaller resource than the ESM or NextGenerationEU with a firepower of hundreds of billions of Euros.
On the most salient issues in the last decades (asylum and fiscal policy), Germany has continuously exhibited a strong preference for the supranational or intergovernmental control of national capacities. Only as a measure of last resort has it ever advocated the creation of supranational capacities. Since the 2000s, and since 2010 at the latest, German preferences have slowly and partially shifted from supranational regulation or supranational capacities to more intergovernmental forms of regulation. This is most clearly visible in the field of fiscal policy where Germany increasingly lost trust in the Commission's ability or willingness to enforce budgetary rules (see Schoeller and Karlsson 2021). A similar distrust of the spending discipline of supranational agents as well as concerns for the budgetary privileges of the Bundestag led Germany to demand a more intergovernmental control of both the ESM and NextGenerationEU (Gnath, Guttenberg, and Redeker 2021). In asylum policy, too, Germany held on to its national veto for fear of losing control over the redistributive allocation of refugees. In the run-up to Amsterdam, it gave up its initial preference for supranational decision-making by qualified majority. Military procurement and police cooperation are exceptions from this tendency towards intergovernmentalism, arguably because the underlying redistributive conflicts are much weaker and much less salient. Chancellor Merkel's advocacy of a more intergovernmental 'union method' in Bruges in 2010 thus seems to be an indicator of a slow but durable shift in German preferences.

Mass publics and mainstream parties
At first sight, the issues analysed in this special issue appear like great test cases for the explanatory leverage of postfunctionalism. Decisions on how to tax and spend, where to use soldiers, whom to arrest and whom to grant asylum are of fundamental importance to the collective identity of a political community. They lead to redistribution and require personal sacrifice. Many of the issues analysed here were also subject to major public debates with high salience. Where, if not here, should the postfunctionalist hypothesis that '(m)ass politics trumps interest group politics when both come into play' (Hooghe and Marks 2009, 18) be vindicated? However, even our data from several decades and over a range of policy areas does not allow for an easy competitive theory testing of functional arguments on interdependence and postfunctional arguments on identity. Only fiscal regulation, fiscal capacity, asylum policy and military cooperation underwent extended periods of strong politicization (understood as high salience, high controversy and increased actor mobilization, Wilde, Leupold, and Schmidtke 2016, 4). Police cooperation and military procurement are typical cases of a permissive consensus: mostly low salience but often with rather high levels of support among the general public (particularly police cooperation).
If the constraining dissensus means that mass publics constrain governments in actions they would engage in if there was no such public dissensus, such a situation is difficult to find in Germany since Maastricht. As the contributions on fiscal regulation (see Schoeller and Karlsson 2021) and fiscal capacity (see Howarth and Schild 2021) demonstrate, there was hardly any dissensus between what state elites wanted and the attitudes of the broader public. Over decades, German state elites and in particularly the ministry of finance, have consistently advocated tight budgetary supervision and opposed the creation of fiscal capacities out of a genuine belief in the analytical and normative appropriateness of such policies and would arguably have done so even if the public did not have strong views. In asylum policy, the rise of challenger parties on the populist right in the mid-1990s and after 2015 did perhaps reinforce the German government's preference for keeping refugee numbers low but did not transform a hypothetical liberal position towards refugees into a restrictive one. Independently of the party-political composition of the German government, it consistently battled for a restrictive position on asylum (see Zaun and Ripoll Servent 2021). The only example where we might see a genuinely constraining dissensus is military cooperation where German state elites might have supported more European capacities with substantial German participation without the consistently strong antimilitarist attitudes of the public (see Bunde 2021).
If there is a constraining dissensus in fiscal policy, it weighs more heavily on the centreleft (i.e. on the Social Democrats and the Greens) than on the centre-right (i.e. on the Christian Democrats and the Liberals). The latter's standard ideological position of a market-based regulatory state on the European level is almost indistinguishable from the beliefs of the general public. The centre-left, on the other hand, is ideologically closer to a position of transnational solidarity and the build-up of European fiscal capacities. In fiscal policy as in other fields, the SPD and the Greens have consistently expressed a moderately higher openness for redistributive capacity-building than the mainstream parties of the centre-right (see Freudlsperger and Weinrich 2021). This reluctant support for supranational capacities is not necessarily reflected by the preferences of the centreleft parties' supporters: 75% of Green voters and 82% of Social Democratic voters rejected the introduction of Eurobonds in 2011 (see Howarth and Schild 2021). Then again, the story on the centre-left might be more complicated than the simple pattern of willing politicians and unwilling publics. Already in the late 1990s, the 'third way' Social Democrats of Gerhard Schröder had called for fiscal restraint on the EU-level and a further reduction of German budgetary contributions (Bremer 2019).
The above numbers on citizens' support for Eurobonds also show that public opposition to supranational capacity-building need not be stirred by a Eurosceptic challenger party (as Hypothesis 2.1 had predicted). While opposition to a Eurozone 'transfer union' emerged from within the ranks of the mainstream centre-right since the early 1990s, the AfD was founded in 2013 only. In line with hypothesis 2.2, the public's rejection of Eurobonds rather speaks in favour of a cost-based assessment of the risks and potential benefits of supranational capacity-building that also permeates citizens' preferences on core state power integration. German citizens' preferences on core state power integration are usually well-aligned with the preferences of state elites, in most cases even beyond partisan divides. The case of asylum policy is, again, telling about this general alignment between elite and citizen preferences. Although asylum application numbers and the salience of the topic were low in the early 2000s, an 'unholy coalition' (Zaun and Ripoll Servent 2021) of Christian Democrats, Social Democrats and Greens fought hard against more liberal asylum provisions in the European Convention. German political elites, the case shows, rarely need to be pressured by the general public into seeking to keep the costs of core state power integration at the conceivable minimum.

Conclusion: a normal member state
Our analysis of German preferences towards the European integration of core state powers since Maastricht reveals a number of patterns and insights on the role of Germany but also on integration theory.
First, Germany is not and probably never was an unconditional supporter of more integration but rather an instrumental one. Helmut Kohl gave strong and arguably idealistic verbal support for European integration. But many of his proposals were vague and contradictory (Kohler-Koch 1998) while the more concrete rejection of a 'transfer union' was a standing formula of his chancellorship. From Kohl to Merkel, German preferences can largely be explained by the desire to deal with interdependence in a low-risk, low-cost manner. This meant a regulation of national capacities rather than the creation of European capacities, and (increasingly) the intergovernmental rather than supranational control of those capacities. Across issue areas, Germany preferred the regulation of national capacities because it hoped to upload its domestic model to Europe. In fiscal policy, it was successful because Germany was part of a group of likeminded creditor states with an extremely strong negotiation position. In asylum policy, it failed because Germany was less pivotal, its coalition weaker and its opponents could realize their preferences by sticking to the status quo. In security policy, the existence of NATO massively lessened the functional pressures towards scaling up governance functions. Only in cases of existential crisis, Germany supported the creation of substantial European capacities (ESM, NextGenerationEU). This support, however, was linked to time limits (NextGenerationEU) and/or intergovernmental instead of supranational control (ESM, NextGenerationEU).
Second, Germany never was a hegemon in a theoretically meaningful sense and is unlikely to become one. German preferences and strategies matter because of the country's size but it is not powerful enough to impose its preferred policy regimes unilaterally on the rest of the EU. Also, a country whose long-term preferences are only slightly exaggerated by the slogan of 'no transfers, no refugees, no soldiers' is neither able nor willing to provide a disproportionate share of the provision of collective goods as a decent liberal hegemon is supposed to, most strikingly in the field of security. The capacities Germany supported in areas of high political salience were without exception either temporary or emergency fixes for a Union on the brink of collapse. After a decade of existential crises, the long-standing German preference for the intergovernmental regulation of national or European capacities is clearly a hindrance to making the EU more resilient. It shifts adaptation costs to states which are already weak and which are not only unwilling but also unable to comply with those rules. It results either in economic hardship and political instability (in the South after the Eurozone crisis) or in non-agreement (in asylum policy). However, these results are not caused by Germany forcing its regulatory model onto the rest of Europe but are the outcome of conflicts among several larger blocks in the EU of which Germany is just one, albeit important, participant.
Third, there are no clear cases of a constraining dissensus in which mass publics and political parties diverged massively from the preferences of state elites. Security policy may be the exception, with anti-militarist attitudes continuously widespread among voters and centre-left parties. But anti-militarist attitudes are not the same as sovereignty concerns. And with the low costs of maintaining the status quo due to the presence of NATO and the strong socialization of defence elites in the transatlantic alliance, the preference intensity among German state elites for creating operational European military capacities is low anyways. Clearly, challenger parties on the right changed the public debate on refugees and reduced the policy space for the government, as did public mobilization against giving German taxpayers' money to EU member states on the verge of default. In both cases, however, the long-term preferences of state elites were largely in line with public pressures. At the same time, in low-salience issues such as police cooperation and defense procurement, the old permissive consensus continued. Among the postfunctionalist factors underlying preference-formation, mass attitudes could have had a stronger constraining effect on the positions of centre-left parties, where the Greens and the Social Democrats lean more towards positive state-building in fiscal and asylum policy, whereas the Christian Democrats' vision of a market-centred regulatory state is less constrained. Overall, we have found no major divergence between functionalist and postfunctionalist explanatory factors over a long period and across a substantial range of core state power areas. Such a rare occurrence of conflict between interest and identity is theoretically meaningful. Functional pushes for a scaling up of government functions rarely conflict directly with postfunctional concerns, and not all instances of core state power integration are salient enough for mobilizing those concerns. Identity may conflict with interest less frequently than expected.