Theorising European integration: the four phases since Ernst Haas’ original contribution

ABSTRACT

Ever since Ernst Haas’ ground-breaking work on European integration, scholars have been theoretically divided over who or what are the drivers of European integration, mainly between supranationalism or intergovernmentalism. Cutting across the substantive divides have also been differences in analytic frameworks, including rationalist/rational choice institutionalism, historical institutionalism, and constructivist/discursive institutionalism. These cross-cutting methodological cleavages have played out in successive phases of substantive scholarly theorisation, even as alternative developments in theory and practice involving deepening integration and increasing politicisation have complicated such theorizations. While the first phase (late 1950s-1970s) was methodologically pluralist as it divided between neofunctionalist supranationalism and realist intergovernmentalism, subsequent phases were clearly divided also by methodological approach. The second phase (beginning in the 1990s) mainly divided between historical institutionalist supranationalists and rational choice institutionalist liberal intergovernmentalists. The third phase (beginning in the 2010s) divided between ‘new’ constructivist/discursive institutionalist supranationalists and ‘new’ constructivist/discursive institutionalist intergovernmentalists. The fourth phase (in the mid to late 2010s) takes greater stock of increasing politicisation while dividing between rationalist/historical institutionalist post-functionalists and constructivist/discursive institutionalist post-functionalists. The conclusion asks how these divides could be more fully bridged in order to achieve more substantive and methodological pluralism.

KEYWORDS:

• European integration
• neofunctionalism
• supranationalism
• intergovernmentalism
• constructivism
• discursive institutionalism
• historical institutionalism
• rational choice institutionalism

Introduction

Scholars of international relations and comparative politics have long theorised about the development of the European Union as a ‘polity’ beyond the nation-state, in particular with regard to who were the drivers of European integration. Ernst Haas started such theorisation in 1958 with his masterful book, The Uniting of Europe. In it, he used a methodological mix of neo-functionalist mechanisms of spillover and constructivist ideas about ‘rational’ interests to argue that supranational technocratic actors in and around the Commission were in control of institutions ‘beyond the nation state,’ and as such were the main drivers of integration (Haas, 1958, see also 1964). His theory soon found strong opposition from Stanley Hoffmann (1966) among others, who maintained to the contrary using realist logics that political actors in the Council, bargaining on the basis of contrasting worldviews and national interests, were in command of those supranational institutions, and thus of European integration. The positions they took have resonated ever since, finding echoes in successive phases of theorisation to this day.

The ever-continuing substantive divide on European integration has been between, on the one hand, the ‘supranationalism’ of those who claim that supranational technical actors such as the Commission (and later the ECB among others) in coordination with experts and interest groups dominate the process of European integration and, on the other hand, the ‘intergovernmentalism’ of those who counter instead that member-state leaders in the Council engaged in intergovernmental bargaining have always been in charge of that process. Over time, these scholarly divides have been complicated by the development in theory and practice of both the EU’s deepening integration and its increasing politicisation. Deepening integration has led to greater decision-making interactivity and growing empowerment for all institutional actors – especially the European Parliament. Increasing politicisation through the impact of politics and ‘post-functionalist’ logics on all EU institutional actors has not just become ever more constraining for European integration but equally in some instances empowering.

In addition to these substantive divisions over the trajectory and drivers of European integration have been methodological differences about the analytic framework best able to explain the motivations, powers, and mechanisms behind European integration. These have cut across the substantive divides. Whereas in the first phase of European integration theory in the 1950s to 1970s, scholars like Haas (1958, 1990) and Hoffmann (1966) used a mix of methodological approaches, including rationalist, systemic, and historical/interpretive modes of explanation, by the second phase beginning in the 1990s such pluralism had disappeared. Instead, comparative politics scholars tended to divide according to their preferred ‘neo-institutionalist’ framework, whether rational choice, historical, or ideational/discursive institutionalism (e.g., Hall & Taylor, 1996; Aspinwall & Schneider, 2000; Schmidt, 2010), at the same time that international relations scholars split between rationalist and constructivist perspectives (e.g., Adler, 1997; Checkel, 1998; Christiansen et al., 1999; Moravcsik, 1999; Abdelal et al., 2010).¹

Taken together, these methodological divisions since the 1990s have led to three very different epistemological positions about the nature of agency, change, and power that cut across substantive theories of EU integration: (1) Scholars take a rationalist and/or rational choice institutionalist perspective in which the material interests of rational actors with coercive powers underpin the rationalist logics and bargaining mechanisms that drive European integration (e.g., Moravcsik, 1993, 1998, 1999; Pollack, 1997; Scharpf, 1999; Schimmelfennig 2015a; Tsebelis, 1994, 2016). (2) Scholars take a historical institutionalist perspective in which the institutional logics of path-dependency (or spillover) and incrementalism are the drivers of institutional power as well as of the mechanisms for continuity in European integration (e.g., Sandholtz & Zysman, 1992; Pierson, 1996; Sandholtz & Stone Sweet, 1998; Ioannou et al., 2015). (3) Scholars take a constructivist and/or discursive institutionalist perspective in which sentient agents’ ideas about interests and discursive logics of policy coordination and political communication are at the basis of their persuasive powers and of the mechanisms for continuity and change in European integration (e.g., Christiansen et al., 1999; Schmidt, 2002, 2006, 2020a; Parsons, 2003; Jabko, 2019; Risse, 2018; Wendler & Hurrelmann, 2022).

In scholarly theorising about European integration, such cross-cutting divisions over substance and method define distinct phases of theorisation of the drivers of European integration. The first phase, which took place roughly from the late 1950s to the late 1970s, is mainly defined by the substantive divisions, since both sides used a mix of methodological approaches to support their arguments. These divisions pit Ernst Haas’ (1958, 1964) neo-functionalist supranationalist concentration on the Commission against Stanley Hoffmann’s (1966) realist intergovernmentalist focus on the Council. The second phase, which began in the 1990s but continues to this day, is largely defined by divides between scholars whose theories of neo-functionalist supranational integration through the Commission and other technical actors are supported by historical institutionalist approaches (e.g., Sandholtz & Stone Sweet, 1998; Ioannou et al., 2015) and those whose theories of liberal intergovernmental integration through the Council are underpinned by rational choice institutionalist logics (e.g., Moravcsik, 1998; Schimmelfennig, 2015a). Although at the forefront of European integration theory at the time, these two perspectives were challenged by scholars who substantively pointed to the EU’s growing politicisation and empowerment of the European Parliament and/or methodologically took a constructivist/discursive institutionalist perspective. The third phase of integration theory, which came to a head in the 2010s, sets ‘new’ supranationalists using constructivist/discursive institutionalist methodological approaches (e.g., Bauer & Becker, 2014; Dehousse, 2016; Epstein & Rhodes, 2016) against ‘new’ intergovernmentalists using similar approaches (e.g., Puetter, 2012; Bickerton et al., 2015), with both sides attributing deepening integration to the growing empowerment of ‘their’ institutional actor. This time as well, however, parallel theories developed concerning the EU’s deepening integration through the growing empowerment of the European Parliament as well as the EU’s increasing politicisation and post-functionalist logics. Finally, a fourth phase of integration theory, which gathered steam around the mid to late 2010s took both supranationalists and intergovernmentalists of all stripes to task for failing to deal sufficiently with the post-functionalist politicisation of European integration and the greater interactivity and empowerment of all EU institutional actors, even as they pitted rationalist/historical institutionalist post-functionalists (e.g., Hooghe & Marks, 2009, 2019; Kuhn, 2019) against constructivist/ discursive institutionalist post-functionalists (e.g., Wendler & Hurrelmann, 2022; Hurrelmann & Wendler, 2024) (see Table 1).

Table 1. Theories of European integration: substantive and methodological approaches.

Phases of Integration	Mainstream Intergovernmental Theories/Approaches	Mainstream Supranational Theories/Approaches	Alternative Theories/Approaches
Original 1950s-1970s	Realist Intergovernmentalism (Council in charge) Rationalist/Historical/Interpretive	Neo-Functionalism (Commission in control) Rationalist/Functionalist/Interpretive/Historical	None
Traditional 1990s & on	Liberal Intergovernmentalism (Council in charge) Rational Choice Institutionalism	Supranationalism (Commission in control) Historical Institutionalism	Deepening Integration (EP) & Growing Politicisation Constructivist/Discursive Institutionalism
New 2010s & on	Intergovernmentalism (Council + de novo bodies in charge) Constructivist/Discursive Institutionalism	Supranationalism (Commission + other agencies in control) Constructivist/Discursive Institutionalism	More Deepening Integration (EP) & Major Politicisation + Post-Functionalism
Current Mid 2010s & on	Post-Functionalism (pressures bottom-up on Council + Commission via constraining dissensus) Rationalist/Historical Institutionalism	Post-Functionalism (pressures bottom-up & at top in Commission + Council via constraining/empowering dissensus) Historical/Constructivist/Discursive Institutionalism	More Deepening Integration thru greater Institutional Interaction & Differentiated Integration

Thus, as part of the Special Issue on Ernst Haas’ contributions to comparative politics and international relations, this article reconsiders the history of substantive scholarly theorising about European integration while adding a new focus on the cross-cutting divides in methodological approaches. It argues that after a first phase of methodological pluralism, subsequent phases of scholarly perspectives on European integration were largely divided not only substantively between supranationalism and intergovernmentalism but also methodologically by the analytic frameworks of rational choice institutionalism, historical institutionalism, and constructivist/discursive institutionalism. The article finds that those substantive and methodological divides are only now beginning to be bridged, suggesting that Haas’ original methodological pluralism, abandoned since the 1990s, may be beginning to re-emerge. It concludes with a question about how EU integration scholars may bridge their remaining differences.

Phase 1: Neo-functionalist supranationalism versus realist intergovernmentalism

Ernst Haas’ (1958, 1961, 1964) early work on European integration took a strong position on substantive and methodological questions. Building on David Mitrany’s (1943) earlier theorisation of ‘functionalism,’ Haas called his substantive theory ‘neo-functionalist.’ Neofunctionalism described supranational civil servants in the Commission in dialogue with experts and interest groups as the drivers of neo-functionalist processes of European integration through incremental spillover and learning. As for his methodological approach. Haas was pluralist in his use of an eclectic mix of approaches, in which ‘rational’ interest was constituted by agents’ interpretive ideas about interests developed in epistemic communities, plus a kind of systems theory ‘lite’ through functionalist spillover. Initially, Haas’ neo-functionalist theory predominated. But by the mid to late 1960s, it was challenged by Stanley Hoffmann’s opposing theory of realist intergovernmentalism which, while similarly methodologically pluralist, championed member-state leaders in the Council over bureaucrats in the Commission.

Substantive theories

Substantively, in Haas’ view of the EU in the late 1950s and early 1960s, the main actors driving integration were supranational civil servants in the Commission engaged on an everyday basis in dialogue and interaction with experts and interest groups of various sorts, whose input to member-state leaders engaged in intergovernmental bargaining in the Council also fed into the functionalist processes of incremental change. Neo-functionalism itself was an institution-building process in which policy engagement in one area led to spill-over into other areas through the unintended consequences of the dynamic interactions among technical actors with other relevant actors, which in turn could lead to changing perceptions of interests (Haas, 1958, 1964). But there was nothing guaranteed about the process, which depended upon a range of factors including institutions, modes of decision-making, and ideological/social environment, which meant that among the many supranational organisations at the time only the EU (as the ECSC and then the EEC) actually reached the mark with regard to supranational integration (Haas, 1961).

With his theory of neo-functionalist supranationalism, Haas directly opposed realist intergovernmentalism, with its rationalist assumptions that states represented by their political leaders acting in the name of geo-political national interests were in charge of integration, even as he admitted that they had their role to play. His main opponent was Stanley Hoffmann (1966), who argued that the member-states were in charge of integration, with member-state leaders exercising their (coercive) bargaining power in the Council to protect their national geopolitical interests as shaped by their differing worldviews while building Europe. For Haas, technical actors in supranational institutions engaging in institution-building in the name of the common interest were as, if not more, powerful in institutional and ideational terms than the political actors of intergovernmental institutions.

In the late 1950s and early 1960s, Haas’ narrative about the neofunctionalist drivers of European integration seemed to best describe what was happening on the ground, as the EU expanded from the ECSC into a much more integrated EEC, and as supranational technical actors gained ever-increasing competencies and power. But there were hard limits to this process, once ‘politics’ reared its head in the form of De Gaulle’s precipitation of the ‘empty chair crisis’ in 1966 over concerns about supranational over-reach (by the Commission and the EP) jeopardising national sovereignty and economic self-interest (mainly agricultural).

Haas’ theory could not account for ‘politicisation,’ as personified by this one charismatic leader. Nor could it deal with nationalism. De Gaulle was the reason why Ernst Haas subsequently declared his own neo-functionalist theory of European integration obsolescent, admitting that the ‘high politics’ of political actors had taken over from the more everyday pragmatic and incremental processes of technical actors (Haas, 1975b). Haas’ declaration we now know was premature. Functionalist processes continued, but in much longer starts and stops than Haas anticipated (see Schmitter, 1970, 2005), and politics became increasingly important as a factor affecting integration.

Methodological approaches

In these early years, the substantive theories of both neofunctionalist supranationalism and realist intergovernmentalism were methodologically pluralist. Haas’ approach combined a wide range of political science's modes of explanation of the time, including rationalist (but not rational choice), historical, interpretive, and functionalist. Hoffmann’s approach was the same with the exception of functionalism. But since the focus of this Special Issue is on Haas’ work, suffice to say about Hoffmann (1966) that his approach included interpretive attention to leaders’ ideas and worldviews, historical focus on nations’ institutional trajectories, and rationalist concentration on nations’ geopolitical interests as expressed through bargaining in the Council.

Ernst Haas’ substantive theory cannot be disentangled from his methodological approach. Toward the end of his career, Haas declared himself a ‘pragmatic constructivist’ (Haas & Haas, 2002). But that is a more recent term. His original methodological toolbox should be considered against the background of the methods prevalent in the 1950s and 1960s in international relations (IR) and comparative politics (CP), which included lawlike, rational choice/rationalist, systemic, historical and interpretive approaches. Haas was clearly set against lawlike (causal and statistically probable) approaches involved in the deductive theorising of general covering laws of human behaviour and of rational choice approaches concerned with rationalist logics of individual (or state) action driven by narrow self-interest with fixed preferences in stable institutions. Instead, his heart lay with Weber on the importance of ideas and the primacy of interests (Ruggie et al., 2005). Interests were defined expansively, as the product of real actors’ ideas about their interests and values constructed in response to material reality, making for ‘pragmatic’ decisions, ‘rational choices,’ or even ‘utilitarian calculations’ based on contextualised shared meanings – as opposed to being narrowly conceptualised as material interests attributed to ‘rational’ actors by (rational choice) scholars. And as a Weberian, his investigative method was arguably a mix of interpretive and historical approaches, with ideal types a primary way of generalising from the particular to the universal (Ruggie et al., 2005). Later, moreover, he would formalise his attention to supranational actors’ consensus-seeking dialogue by theorising the discursive interactions of supranational civil servants and experts in ‘epistemic communities’ (borrowing from Ruggie, 1998) made up of actors ‘who shared a commitment to a common causal model and a common set of political values’ (Haas, 1990, p. 41; see also Haas & Haas, 2002). For Haas, then, technical actors’ ideas and discursive interactions created knowledge, and knowledge had power.

All of this suggests that Haas’ original approach would certainly be seen (and was later seen by Haas himself – see Haas & Haas, 2002) as fitting into the constructivism which developed in IR in the 1990s (Ruggie, 1998) and had entered EU studies by the mid to late 1990s (e.g., Christiansen et al., 1999; see discussion in Risse, 2018). But not all of Haas’ approach fits under these rubrics. His neo-functionalist theory also incorporates another methodological approach that goes beyond technical actors’ (ideas about) interests to focus on the unintended consequences of their actions and the creation of institutions with a ‘life’ (and institutional power) of their own, through ‘functionalist’ spillovers acting as drivers of change separate from actors’ intentions and interests.

Functionalism here is neither structural-functional nor systemic in the sense often theorised during the 1950s and 1960s (whether of Talcott Parsons or David Easton), and spillover is therefore not some automatic process, later misinterpretations to the contrary notwithstanding. Haas’ functionalism is about functional areas of policy in which technical agents’ interest-based ideas and actions created the impetus for expansion from one policy area into another through incremental processes of change that went beyond the actors’ own intentions or even ability to grasp. This is functionalism ‘lite,’ however, since there are no set systemic structures or mechanical functions, no homeostatic equilibrium, or anything of the sort. In this latter respect, Haas’ approach is akin to historical institutionalism of the early 1990s, in which institutional path-dependency and unintended consequences are key to understanding the dynamics of continuity (and change), and where agents’ ideas about interests and values were considered alongside rationalist attention to material interests (Steinmo et al., 1992; see Schmidt, 2010).

Phase 2: Historical institutionalist supranationalism versus rational choice institutionalist liberal intergovernmentalism

The second phase of European integration theory again opposed supranationalists and intergovernmentalists. It began in the 1990s but continues to this day. Here too, the substantive debates were underpinned by methodological differences. But the methodological approaches had themselves changed significantly following the ‘methodological wars’ beginning in the 1990s among different kinds of ‘new institutionalisms’ in CP, including rational choice, historical, and ideational/discursive institutionalism (Hall & Taylor, 1996; Aspinwall & Schneider, 2000; Schmidt, 2002, 2006, 2010), and in IR between rationalist and constructivist approaches (Finnemore, 1996; Adler, 1997; Ruggie, 1998; Checkel, 1998; Christiansen et al., 1999; Moravcsik, 1999; Abdelal et al., 2010; Risse, 2018). Briefly defined, rational choice institutionalism focuses on the coercive power of rational actors who pursue their preferences following a ‘logic of calculation’ within political institutions, defined as structures of incentives. Historical institutionalism details the institutional power of political institutions described as regularised patterns and routinised practices subject to a ‘logic of path-dependence’ or incrementalism. Constructivism and discursive institutionalism centre on the persuasive power of ‘sentient’ (thinking, speaking) agents who express their ideas through discourse following a ‘logic of appropriateness’ in meaning contexts and a ‘logic of communication’ in discursive interactions within institutional contexts (formal and meaning-based) (see Schmidt, 2008, 2010; Carstensen & Schmidt, 2016).

In this second phase of European integration theory, scholars on both sides took their distance from their predecessors as well as one another in both theory and method. On the supranational side, scholars largely updated Haas’ neo-functionalism while dropping the constructivist/ideational aspects in favour of a historical institutionalism more concerned with material interests. On the intergovernmentalist side, scholars rejected realist intergovernmentalism for liberal intergovernmentalism, and moved from the realists’ softer rationalism to rational choice institutionalism.

Historical institutionalist supranationalism

With historical institutionalist supranationalism, scholars such as Wayne Sandholtz, Alec Stone Sweet, and John Zysman argued substantively that integration moved forward because EU officials were in control, exercising institutional power through their mastery of the EU policy process, their expertise, and/or their interaction with organised interests (Sandholtz & Zysman, 1992; Sandholtz & Stone Sweet, 1998). Building on while going beyond Haas’ earlier theorising of neo-functionalist spillover, historical institutionalist supranationalists saw bureaucratic entrepreneurialism and institutional creep as having increasingly institutionally empowered supranational actors in EU institutions such as the European Commission and the European Court of Justice (Sandholtz & Stone Sweet, 1998).

Methodologically, these supranationalists tended to explain integration in terms of the institutional developments of the EU using historical institutionalist methods married to soft rationalist assumptions about how such development advantaged and/or was promoted by EU institutional actors in pursuit of their (material) interests. They therefore centered their attention on the actual institutions, or ‘macro-structures’, in which political action occurred and emphasised the ‘path dependencies’ that served to lock-in the increasing power of EU institutional actors (see Pierson, 1996). In so doing, they also tended to be more concerned with everyday policy-making than with the ‘high politics’ of the treaties, and with technocratic actors rather than political ones. As such, they opposed intergovernmental explanations of integration that focused on the rationalist, interest-based motives of member states using rational choice institutionalist methods like principal-agent theory.

Rational choice institutionalist liberal intergovernmentalism

With liberal intergovernmentalism, Andrew Moravcsik (1993, 1998) sought to account for why member-states deepened integration even when it appeared to go against their national interests. He did so by reaffirming that the member-states were in charge, imposing their preferences through intergovernmental bargaining in the Council, but argued that domestic socio-economic interests rather than geopolitical interests were the main drivers of member-states’ preference formation. Theorising within a rational choice institutionalist framework meant that liberal intergovernmentalists saw the outcomes of member-states’ rationally driven, interest-based negotiations as determined by their asymmetric bargaining power, or what we will henceforth term, for purposes of conformity with the literature, ‘coercive’ power, defined as relations of control by one actor over another, where these relations allow one actor to shape directly the circumstances or actions of another (Dahl, 1957; Barnett & Duvall, 2005, p. 43, 49; see also Carstensen & Schmidt, 2016, 2018).

This liberal intergovernmentalist approach to integration also sought to counter the ‘supranationalist’ view, including the earlier neo-functionalist approach focused on ‘spillover’ effects (Haas, 1958) and the historical institutionalist approach emerging at the same time as liberal intergovernmentalism. For liberal intergovernmentalists generally, supranational actors were of little concern other than, arguably, as delegated agents to the ‘principals’ in the Council. And within the context of principal-agent theory, they therefore proceeded to sort out who delegates to whom on what, with which kinds of tools and sanctions (Pollack, 1997).

In response, supranationalists argued at the time that the problem with principal agent theory is that there were too many principals as well as too many agents, with those technically definable as agents often acting as principals and vice-versa (Marks & Hooghe, 2001; Schmidt, 2006, pp. 53–4). As a result, they insisted that liberal intergovernmentalists could not answer such questions as who is in the driver’s seat or why principals might willingly give up control (Marks & Hooghe, 2001, pp. 71–77). Moreover, as traditional supranationalists might argue, supranational actors’ independent institutional power enabled them to act even without approval from the alleged ‘principals,’ despite their seeming monopoly on coercive power.

But while these two main theorizations of European integration largely shared the limelight, other scholars were developing alternative substantive theories and methodological approaches that focused attention on issues complicating both sides of the integration story. Some scholars concerned themselves with deepening integration through the growing empowerment of other EU actors, including the EP (Tsebelis, 1994), and of the regions through ‘multi-level’ governance (Marks & Hooghe, 2001). Others charted the EU’s increasing politicisation through the EU’s indirect and often deleterious effects on national politics (Van der Eijk & Franklin, 2004; Mair, 2006). And finally yet other scholars built on the analytic frameworks of constructivism and discursive institutionalism to develop methodological approaches to European integration emphasising the importance of leaders’ ideas, their legitimising discourse, national identities and socialisation regarding the EU (e.g., Christiansen et al., 1999; Parsons, 2003; Schmidt, 2002, 2006; Risse, 2010).

Phase 3: ‘New’ constructivist/discursive institutionalist supranationalism versus ‘New’ constructivist/discursive institutionalist intergovernmentalism

The third’ phase of theorising European integration came to a head in the 2010s, despite the persistence of the now ‘traditional’ (phase 2) of theorizations. Traditional historical institutionalist supranationalists still maintained that the Commission used institutional dynamics to push deeper European integration along with self-empowerment (Nieman & Ioannou, 2015; Ioannou et al., 2015) while traditional rational choice institutionalist intergovernmentalists insisted that the strongest won out in the rationalist hard bargaining of the Council (e.g., Schimmelfennig, 2015a; Tsebelis, 2016), But in this third phase, while scholarly views remained substantively divided on the same bases as the ‘traditional’ theorizations, between intergovernmentalists and supranationalists, they were methodologically united in their rejection of rationalist and historical institutionalist analytic frameworks in favour of constructivist/discursive institutionalist frameworks. This new stage of theorisation pitted the ‘new’ constructivist/discursive institutionalist supranationalists, who viewed EU level supranational actors such as the Commission and the European Central Bank (ECB) as continuing to drive integration through their greater ideational role in policy design and innovation, against the ‘new’ constructivist/discursive institutionalist intergovernmentalists, who insisted that the more actively engaged, consensus-seeking member-state governments in the deliberations of the Council had retaken control. The new intergovernmentalists argued that member-state actors were in charge of EU governance as a result of their powers of persuasion in intergovernmental deliberation while the new supranationalists contended instead that supranational actors were in control of EU governance as a result of their powers of ideational innovation, institutional entrepreneurialism, and administrative discretion (Schmidt, 2018). Here, we begin with the new intergovernmentalists, mainly because the new supranationalists have had more to say in response.

‘New’ constructivist/discursive institutionalist intergovernmentalism

With the ‘new’ intergovernmentalism, scholars such as Uwe Puetter, Christopher Bickerton, Dermot Hodson, and Sergio Fabbrini argued that the EU’s member-state leaders in the European Council had become much more legislatively active than in the past through their unprecedented leadership role exercised through consensus-seeking deliberation rather than the interest-based bargaining theorised by traditional intergovernmentalists (Puetter, 2012; Bickerton et al., 2015; Fabbrini, 2015). This involved much more shared authority and joint control at the EU level than was considered possible by traditional intergovernmentalists, whether in the original ‘realist’ view – in which member-states’ bargaining focused on protecting national sovereignty and interests (Hoffmann, 1966) – or the revisionist ‘liberal’ view – in which member-states’ hard bargaining and brinkmanship served as a conduit for domestic socio-economic interests (Moravcsik, 1993; see also Schimmelfennig, 2015a).

Moreover, unlike those traditional intergovernmentalists who theorised the Commission as delegated ‘agents’ which the Council ‘principals’ could control, the new intergovernmentalists presented the Council as actively seeking to take back control from the Commission through the creation of de novo supranational agencies (e.g., Hodson, 2015). In conjunction with this argument, the new intergovernmentalists at the same time contended that member-states’ new activism sidelined the bureaucratic entrepreneurialism and institutional creep theorised by traditional supranationalism.

For the new intergovernmentalists, finally, the mistake of traditional intergovernmentalists and supranationalists alike was to assume that the process is all about the pursuit of hard power, whether coercive power through interest-based bargaining in the Council or institutional power through budget maximising for the bureaucracy. Instead, new intergovernmentalists maintained that the decision-making process in the Council since the Maastricht Treaty of the early 1990s needed also to be understood in terms of the exercise of ideational and discursive power (Carstensen & Schmidt, 2018), with member-states seeking to arrive at consensual agreements through deliberation (Bickerton et al., 2015). In the Council in particular, they highlighted the deliberative processes of negotiation that lead to agreements resulting from persuasion rather than power politics (Puetter, 2012).

As such, new intergovernmentalists would reject liberal intergovernmentalists’ single-minded focus of analysis of the Eurozone crisis response as a rational choice game of chicken, in which the stronger member-states’ preference to avoid the breakdown of the euro area was combined with efforts to shift the costs to the weaker euro members most in trouble (Schimmelfennig, 2015a; Tsebelis, 2016). Instead, new intergovernmentalists emphasised ideational/discursive power, mainly in terms of the consensus-based agreements forged in the Council. After all, each and every member-state bought into the story of excessive public debt and failure to follow the rules, pledged themselves to austerity, and agreed repeatedly to reinforce the rules of the SGP in exchange for loan bailouts or bailout mechanisms (Fabbrini, 2013; Schmidt, 2020a, Ch. 5). Missing in this analysis (with the exception of Fabbrini 2016), however, is the recognition that this was not a deliberation among equals, since Germany held outsize power to pursue its own interests. And Germany’s power was not only ideational, following from German insistence on the reinforcement of the ordo-liberal rules of the Stability and Growth Pact; it was also institutional, as German opposition to doing anything delayed any decision on Greece until the markets threatened the very existence of the euro; and it was coercive, resulting from Germany’s political veto position in any agreement as a result of its economic weight as the strongest economy in Europe (Carstensen & Schmidt, 2018; Schmidt, 2018).

A major problem for the new intergovernmentalist approach in general, then, is that it has done little to theorise other kinds of power in the context of constructivist deliberation. The very use of the terms deliberation and consensus-seeking seems to imply that member-states do not engage in the power relations and bargaining posited by the traditional intergovernmentalists. But without considering coercive power in the processes of deliberation or, better, contestation, we can’t explain why Germany’s preferences won out, especially initially (Schimmelfennig, 2015a, 2015b). That said, by only positing coercive power and rationalist bargaining (as per traditional rationalist intergovernmentalism), we can’t explain why Germany conceded, over and over again, to things it had initially resisted not only during the Eurozone crisis but equally during the Covid-19 pandemic (Schmidt 2015, 2018, 2020a; Carstensen & Schmidt, 2018 – but see Lundgren et al., 2019 for a rationalist intergovernmentalist explanation of how Germany ‘traded gains and concessions’ and ‘exchanged wins and losses’).

‘New’ constructivist/discursive institutionalist supranationalism

With the ‘new’ supranationalism, scholars such as Michael Bauer, Stefan Becker, Renaud Dehousse, Rachel Epstein and Martin Rhodes (Bauer & Becker, 2014; Dehousse, 2016; Epstein & Rhodes, 2016) might agree with the new intergovernmentalists against traditional supranationalists that institutional power and leadership by technical actors in the Commission has indeed diminished. But in exchange, these new supranationalists contended, intergovernmental political leaders’ greater legislative activism in the Council had enabled all supranational technical actors – whether the Commission, the European Central Bank, or other de novo bodies – to gain even greater institutional powers of enforcement than in the past. Moreover, in an ironic twist, according to the new supranationalists, these selfsame technical actors had, through the exercise of ideational power, developed and proposed to intergovernmental political leaders the policy initiatives they themselves then enforce (Schmidt, 2018).

Thus, in contrast to the traditional supranationalists’ focus on the Commission’s self-interested drive for institutional empowerment, these new supranationalists emphasised the Commission’s ideas and institutional entrepreneurship to make European integration work better, whether or not this served its specific power and interests (Bauer & Becker, 2014; Dehousse, 2016; Epstein & Rhodes, 2016). Moreover, they argued not only that supranational actors had gained greater authority and responsibility but also that with this had come an unprecedented institutional autonomy of action (especially the ECB) and discretion in applying the rules (in particular the Commission).

In the Eurozone crisis, for example, the new supranationalists argued that the European Commission was ‘the unexpected winner of the crisis’ (Bauer & Becker, 2014) and that supranational actors more generally – in particular the ECB as well as the Commission – had ‘availed themselves of the discretionary powers with which they were formally or informally vested to adopt decisions that did not reflect the policy preferences of all national governments, notably those of Germany’ (Dehousse, 2016). Moreover, these supranational actors were themselves responsible for coming up with the ideas for new empowering institutions in coordinative discourses with think tanks and experts as well as in consultation with member-states in the Council. This included the European Semester, which was long prepared by the Commission, and Banking Union, spurred by the ECB (Bauer & Becker, 2014; Dehousse, 2016; Rhodes and Epstein 2016; Schmidt 2015, 2020a).

New supranationalists, then, like the new intergovernmentalists and unlike the traditional supranationalists or intergovernmentalists, were also mainly focused on ideational innovation and discursive interactions as the mechanisms through which technical actors were able to prevail even against the wishes of Council political leaders. As for ideational/discursive power, they saw it exercised through actors’ innovative ideas and persuasive discourses, whether the ECB radically reinterpreting its mandate to legitimate its increasingly expansive monetary policy or the Commission reinterpreting the rules of the Stability and Growth Pact (Schmidt, 2016, 2020a; Carstensen & Schmidt, 2018). Such supranational ideational/discursive power could also be seen in the pandemic response by both the Commission and the ECB, whether the Commission’s repackaging of its ideas about the green transition, the digital transformation, and combatting inequality as the goals of the Resilience and Recovery Fund or the ECB’s unlimited bond-buying (Schmidt, 2020b).

One omission of these ‘new’ theorizations of European integration, much like the traditional theorizations, is that they did not do much to theorise about how the two institutional actors interacted with each other. And yet, such interaction has long been in discussion in EU studies, including in the ‘deliberative supranationalism’ evident in comitology, where the member-states’ delegates sit on Commission committees in regulatory activities (Jeorges and Neyer 1997), in the ‘supranational intergovernmentalism’ of decision-making in Common Security and Defense Policy (Howorth, 2012), and in the interactions of EU agencies and supervisory bodies such as those involved with the ECB in supervising banking activities, where actors representing both supranational and intergovernmental institutions interact on an everyday basis (Christiansen this issue).

Another omission is that such ‘new’ approaches overlooked alternative substantive theories focused on the continued deepening of integration as well as the major increase in politicisation. By the 2010s, an even more extensive literature than in the 2000s existed on the growing empowerment of the EP (e.g., Héritier & Reh, 2012; Hix & Høyland, 2013; Crum & Curtin, 2015) and on the realities of increasing politicisation, as seen in rising Euroskepticism and anti-system populist parties (e.g., De Wilde & Zürn, 2012; Kriesi et al., 2012; Mair, 2013; Hutter et al., 2016; Zürn, 2019). Moreover, theorizations of politicisation also gave rise to post-functionalist theories of integration (Hooghe & Marks, 2009), which were to gain traction in the subsequent phase of theorisation.

Phase 4: rationalist/historical institutionalist post-Functionalism versus constructivist/discursive institutionalist post-Functionalism

As we have already seen, largely missing from the scholarly literature on the drivers of European integration, traditional and new, have been deepening integration, in particular with regard to the empowerment of the EP, and politicisation. While this may have been understandable in the case of Haas’ earliest works, at a time when the EU’s development was still to come while politics (other than ‘high politics’) seemed pretty far away from EU preoccupations and national citizens’ concerns, it became less and less so over the decades. But for a very long time, the debates between supranationalists and intergovernmentalists passed over these issues. These oversights have in many ways been remedied in the most recent phase of integration theory.

The fourth phase of European integration theory begun in the mid to late 2010s mainly addresses the politicisation of European integration while implicitly critiquing Haas’ neofunctionalism by calling itself ‘post-functionalism.’ Although this approach originated in 2009 (Hooghe & Marks, 2009), it is really only in the mid to late 2010s that it picked up steam (e.g., Hooghe & Marks, 2019; Kuhn, 2019). But here, too, such theorising of post-functionalist politicisation has been methodologically split between methodological approaches using rationalist/historical institutionalism and ones using constructivist/discursive institutionalism (e.g., Crespy & Vanheuverzwijn, 2019; Jabko, 2019; Schmidt, 2020a, 2020b; Wendler & Hurrelmann, 2022; Hurrelmann and Wendler 2023). What is more, these two sides differ on the implications of such politicisation for European integration, with the former mainly predicting a politicised constraining dissensus and the latter showing the possibilities of overcoming such constraints and even positing a politicised empowering dissensus. But whatever the implications of the different post-functionalist approaches, they left under-theorised the deepening integration involving the institutional actors driving European integration. Alternative theories of European integration, here again, pick up the slack, in this case by considering EU actors’ increasing dynamics of interaction, whether through processes of failing forward (e.g., Jones et al., 2016, 2021), of institutional actors’ interactive dynamics in moments of crisis (e.g., Schmidt, 2018, 2019, 2020b; Riekmann et al., 2021; Schimmelfennig, 2023), or as a result of differentiated integration (Schimmelfennig and Winzen 2019; Leruth et al., 2022).

Rationalist/Historical institutionalist postfunctionalism

With rationalist/historical institutionalist post-functionalism, Liesbet Hooghe and Gary Marks (2009, 2019) developed a theory that builds on neo-functionalist approaches while adding a rationalist political component to explain the increasing politicisation of the EU’s multi-level governance. Against the predictions of earlier neofunctionalist scholarship (Schmitter, 1969), they argue that politicization in the EU has not led to increased integration but rather the opposite. Because these post-functionalists see politicization processes as having largely unfolded on an axis of identity and territorial cleavages in EU politics, they expected it to exercise ‘downward pressure on the level and scope of integration’ (Hooghe & Marks, 2009, p. 21), leading to a ‘constraining dissensus’ on both the deepening and widening of the EU polity (Hooghe & Marks, 2009, p. 2019).

In their post-functionalist theory, Hooghe and Marks (2009, p. 2019) argue that transnational interdependence, which triggers a demand for European governance in a specific area, results in mismatches between the requirements of effective governance by the EU (or ‘function’) and the institutional status-quo of the nation-state member (or ‘form), which speaks in particular to concerns related to national sovereignty, identity, and democratic legitimacy. They predict that where European integration becomes a matter of public debates in mass arenas resulting in societal controversy, a constraining dissensus will emerge, and thereby stymie further integration. This is because politicization activates concerns about collective identities and translates into new transnational political cleavages that disrupt traditional party politics on the left/right axis (Hooghe & Marks, 2019; see also Hooghe & Marks, 2009). As such, their theory provides an important addition to neo-functionalist theories of integration as well as a corrective to intergovernmentalism and supranationalism, since they add the national level into the mix both in terms of institutions (historical institutional/functionalist path-dependencies) and politics (rationalist political logics).

But while this a compelling theoretical argument, in practice it suffers from two drawbacks. First, it is does little to theorize EU level interactions other than to suggest that the Council decision-making is constrained by bottom-up politics. As such, it seems to suggest that politicization operates the same way across member-states, with the more salient the issue, the greater the constraints on policymakers. Empirically, however, politicization due to mismatches between form and function is not automatic, since it occurs only if political agents themselves raise their concerns publicly, and it affects governments differently, with some better able to deal with it than others (even when controlling for institutional contexts and material factors) (Wendler and Hurrelman 2022; Hurrelman and Wendler 2023). Second, post-functionalism presupposes that politicization is largely negative in terms of results, leading to constraints on political elites that thereby hamper further integration. One could instead argue, however, that under certain circumstances, politicization can be positively empowering, as citizens demand solutions, and policymakers supply them, as they accommodate or innovate in terms of European integration (as discussed below). This said, one could alternatively contend, as have constructivist/discursive institutionalist intergovernmentalists, that politicization has led to a ‘destructive dissensus’ rather than engendering a constraining dissensus, let alone an empowering one (Hodson & Puetter, 2019).

Constructivist/Discursive institutionalist post-Functionalism

With discursive institutionalist post-functionalism, Achim Hurrelmann and Frank Wendler offer a theoretical complement to rationalist/historical institutionalist post-functionalism by adding a discursive institutionalist perspective (Wendler & Hurrelmann, 2022; Hurrelmann and Wendler 2023). This approach overlays the dynamics of ideas and discursive interactions onto the rationalist and historical institutionalist assumptions built into post-functionalism, while also taking account of material and political factors. Thus, they argue that ‘political elites do not face form-function mismatches as a quasi-objective, technical challenge, but as a discursively constructed conflict that can range between relatively limited policy-specific questions to more fundamental challenges to the form and rationale of European integration’ (Wendler & Hurrelmann, 2022). And they critique the post-functionalist assumption that politicization automatically leads to a ‘one size fits all’ constraining dissensus on European integration when mismatches occur between ‘function’ (effective governance at the EU level) and ‘form’ (institutional status quo at the national level).

Instead, Wendler and Hurrelmann (2022) break down the different elements involved in politicization with regard to differences among kinds of ideas as well as discursive interactions. Thus, they argue that politicized issues that stay within the coordinative discourse of the policy sphere – which are most often policy-related (especially economic or technical) – are easier for governments to deal with than those that spill over into the communicative discourse of the political sphere, where a clash with programmatic ideas about political values and partisan ideologies or with philosophical ideas about democratic legitimacy or national sovereignty can prove more difficult for governments to reconcile. But high salience (or form-function mismatch) does not have to constrain EU integration, so long as government elites respond with strong communicative discourses of their own (Wendler & Hurrelmann, 2022).

Finally, although Hurrelman and Wendler critique the post-functionalist assumption that politicization is always constraining in the same way in its bottom-up effects, they don’t seem to go farther than to suggest that it can be overcome in a variety of ways. But couldn’t politicization also be empowering? And what would empowering mean? When defined not only in terms of heightened salience or actor expansion but also in terms of the actual enlarging of available policy options, politicization can also lead to an ‘empowering dissensus’ (Oleart, 2020; Bouza & Oleart, 2018) that responds to the concerns of citizens and civil society while putting political pressure on European decision makers to take collective action. We could call this ‘positive’ politicization (Schmidt, 2019). To illustrate, during the Eurozone crisis politicization among EU institutional actors may have constrained but did not stop more positive ideational innovations and discursive dynamics that led to reinterpretations of the rules (Crespy & Vanheuverzwijn, 2019; Schmidt, 2020a). Moreover, the discursive repertoires of EU ‘governance’ changed in ways that ensured that political contestation may have constrained but did not derail EU institutional actors’ increasing flexibility, innovation and experimentation (Jabko, 2019). And certainly the response to the Covid-19 pandemic represented another instance of positive politicization after the initial period of constraining dissensus, where Commission and Council engaged in a mutually empowering dissensus or, better, an ‘enabling consensus’ on taboo-breaking measures (Ferrara & Kriesi, 2022).

These examples also point to the fact that, however significant a corrective this fourth phase of ‘grand theory’ on European integration is to previous theorization, by putting politicization front and center it takes the focus off of the institutional actors driving integration. Alternative theories here too offer insights. Some theorists have softened the sharp divides between intergovernmentalists and supranationalists by combining approaches. For example, even one of the major contributors to the traditional liberal intergovernmental side, Frank Schimmelfennig (e.g., 2015a), has begun theorizing about the functionalist and post-functionalist dynamics of the EU in tandem with rationalist bargaining by member-states, albeit largely keeping the two sides separate (Schimmelfennig, 2023, this issue). Another set of scholars (Jones et al., 2016, 2021) have added a neo-functionalist component to rationalist liberal intergovernmentalism in order to describe the neo-functionalist dynamics of ‘failing forward’ through which intergovernmental bargaining in the Eurozone crisis has led time and again to incomplete agreements and failed reforms carried out by the Commission that soon required new intergovernmental bargains. This latter synthesis has largely remained within the traditional camp, however, with the rapprochement of intergovernmentalism and supranationalism using a mix of rational choice and historical institutionalism. As such, it suffers from the same problems as rationalist/historical institutional post-functionalism, because it has difficulty taking into account the ideational innovations and positive successes in negotiations with regard to the pandemic (Rhodes, 2021).

But be this as it may, these recent theorizations also point to what other scholars using constructivist/discursive institutionalism as much as rationalist/historical institutionalism have been theorising: that EU institutional actors have been engaged in an ever-increasing dynamics of interaction in which all such actors have experienced growing empowerment, each with its own role to play, and with the contestation prevalent among them during the Eurozone crisis (e.g., Schmidt, 2020a; Riekmann et al., 2021) followed by greater cooperation during the Covid-19 crisis (e.g., Schmidt, 2020b; Schmmelfennig 2023) Differentiated integration offers additional insights into these dynamics by charting the increasingly complex set of interactions resulting from the growing differentiation of EU policy domains, with different institutional actors as well as member-states involved (Schimmelfennig and Winzen 2019; Leruth et al., 2022).

Conclusion

The benefit of EU integration theories focused on supranationalism and intergovernmentalism in phases one through three is that the different sides lend major insights into the many different motivations and powers of ‘their’ EU actor/institution vis-à-vis the other EU actors/institutions from the perspective of one analytic framework, rational choice institutionalism, historical institutionalism, or constructivist/discursive institutionalism. The drawback is that they have been naturally more focused on demonstrating the significance of their given EU actor/institution while following rationalist, historical, or constructivist logics rather than in shedding light on the overall picture, which has changed dramatically due to growing politicisation, the empowering effects of deepening integration, and the increasingly complex dynamics of interaction among EU institutional actors. The benefit of post-functionalism is that it calls attention to the politicisation which has affected EU integration from the bottom up as well as at the top from a combination of perspectives, rationalist/historical institutionalist or constructivist/discursive institutionalist complementing the former. The drawback is that it is less concerned with EU institutional actors themselves.

The challenge for integration theorists, then, is how to escape their remaining substantive and methodological silos, to develop theories accounting for how EU actors/institutions have all become more empowered, more dynamically interactive, and more political while exercising many different kinds of power – coercive, institutional, and ideational/discursive. As this article suggests, recent innovations in integration theory show that integration theorists have indeed begun to break down the silos, bridging their differences by increasingly theorising EU institutions and actors in interaction using a mix of analytical frameworks. As such, perhaps it is time to recognise not only that a pluralism of actors/institutions now drives European integration, but that this can be explained using a pluralism of approaches. This would however additionally demand bridging the final gap between rationalist/historical institutionalism and constructivist/discursive institutionalism.

The question is how? Haas’ own work offers some clues, since its methodological pluralism uses different modes of explanation to elucidate different aspects of European integration. We could call such pluralism analytical eclecticism (Sil and Katzenstein 2010) or ‘seizing the ‘middle ground’ (Adler, 1997), and leave it at that. But I would rather think of such pluralism as requiring the recognition that rationalist, historical institutionalist, and constructivist/discursive institutionalist approaches all provide different perspectives on reality, with their own ontologies and epistemologies (Schmidt, 2008, 2010; see also Risse, 2018). Mixing them together indiscriminately is therefore not advised. However, each can provide insights the others do not, often in a complementary fashion, such as with ‘failing forward’ marrying neo-functionalist and rationalist approaches, or with discursive post-functionalism adding the ideational/discursive to analyses in terms of rationalist veto players and functionalist logics. Moreover, many scholars recognise the usefulness of other analytic frameworks even though they may not use them. For example, Hooghe and Marks (2009, p. 13) argue explicitly that politicisation is ‘constructed’ and, that public opinion on Europe is malleable, which would suggest that they would have no objection, say, to constructivist investigations of the reconstruction of identities in ways that could turn the constraining dissensus into an empowering one. But even where achieving complementarity among approaches is difficult, each can equally be used to problematise the results of the other – as was done at the end of each of the sections in this essay. In short, methodological pluralism suggests using as many approaches as appropriate, by sometimes using the results of one methodological approach as useful complement to or background information for another; other times, by problematising them. In the end, much depends on the questions asked and the empirical issues under study as to which methodological approaches are most fruitfully utilised.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Vivien A. Schmidt

Vivien A. Schmidt is Jean Monnet Professor of European Integration and Professor Emerita of International Relations and Political Science at Boston University, Honorary Professor at LUISS Guido Carli University, and Visiting Fellow in the Schuman Center, European University Institute

Notes

1 I leave out sociological institutionalism here because it is largely the equivalent of constructivism in international relations (see Finnemore, 1996) at the same time that it has until recently been much less present in comparative politics, and is in any event in part subsumed under discursive institutionalism (see Schmidt, 2010).