From Practice to Principle and Back: Applying a New Realist Method to the European Union’s Democratic Deficit

Abstract

The prospect of a Brexit illustrates that the European Union’s legitimacy deficit can have far-reaching political consequences. In normative political theory, realists take a keen interest in questions of legitimacy. Building on Bernard Williams’ realist writings, I propose a two-step method of normative political theorization. Each step contains both a practice-sensitive phase and a practice-insensitive phase. First, the conceptualization of a norm should draw on conceptual resources available to agents within their historical circumstances. Second, the prescriptions that follow from this norm should take into account whether political order can be maintained. Applying this method to the European Union’s democratic deficit yields, first, based on public opinion research, the norm of European deep diversity and, second, a set of prescriptions for a demoicratic confederacy. Thereby, I demonstrate that this realist method is able to yield political theories distinct from other philosophical approaches. Moreover, I contribute a realist theory to the normative literature in European Union studies.

On 23 June 2016, almost 52% of more than 30 million British citizens voted to leave the European Union (EU). This outcome created a sense of shock across Britain and the European continent. Research clearly suggests that “regaining control” over the borders to stop immigration was an influential reason to vote leave (Hobolt, 2016). The widespread perception of the EU’s so-called democratic deficit is an (implicit) precondition for the persuasiveness of this argument. If the EU were deemed a legitimate democratic decision-maker, then these arguments concerning the sovereignty of the Westminster parliament should not convince the British electorate. EU laws would be authoritative whatever their content. The EU’s legitimacy deficit means that popular dissatisfaction with policies can much more easily translate into a rejection of the political system as a whole. The prospect of a Brexit illustrates that this legitimacy deficit can have far-reaching political consequences for Europe’s political order.

In normative political theory, political realists take a keen interest in questions of regime legitimacy. In recent years, realism has gone through a revival. Many scholars have written on its distinctiveness compared to other approaches in normative political theory (e.g. Galston, 2010; North, 2010; Rossi and Sleat, 2014). A central realist claim is that political theory cannot assume any unanimity on political objectives, let alone morality, because politics necessarily deals with permanent disagreement on these matters (Sleat, 2016a). Realists assume that individuals will always disagree about political ends. Coercive structures are necessary to prevent such disagreements dissolving into (civil) war. For realists, political theorists should aim to theorize values that can maintain, or at least not undermine, a civil order funneling these disagreements. Their central concern is not that normative theory should either be plausible or feasible—the way “realistic” is often used (e.g. Neyer, 2010). This concern is conceptual: political theory should take into account coercion and persistent disagreement when theorizing political norms.

In this article, I theorize EU legitimacy using a new realist method, thereby seeking to contribute to the literature on political realism and the normative literature in EU studies. The realist literature was primarily embroiled in a Methodenstreit (Rossi, 2016). Recently, however, realists have turned their attention to substantive issues (e.g. Jubb, 2015b; Rossi, 2017; Sleat, 2016b). These explorations offer little systematic guidance. Meanwhile, methodological pieces provide important but general guidelines (Hall, 2015; Jubb, 2017). Drawing upon the political thought of Bernard Williams (2004, 2005, 2009), I propose a two-step method of realist political theorization that can inform realist studies on a wide variety of substantive questions. Each step contains both a practice-sensitive phase and a practice-insensitive phase. To demonstrate its potential, I apply this method to the EU’s legitimacy deficit. Few political realists have contributed to this debate (e.g. Beetz and Rossi 2017) even though it touches upon a central concern of realists: domination that is coercion without widespread legitimacy.

The article also contributes to the normative literature on the EU. Its first contribution is to propose a new norm of EU legitimacy: “European deep diversity” (EDD). I draw out three widespread conceptual patterns from public opinion research in assessing EU legitimacy. Subsequently, I argue that Charles Taylor’s concept of deep diversity coherently incorporates these patterns. Rather than a mere analytical description, I build on its inherent normative logic to develop a contextualized norm. As such, it justifies the persistence of distinct democratic communities in a single polity. National political orders remain the locus of legitimate rule in the EU because they enable the pursuit of the common good. In turn, the legitimacy of the EU regime relies on both its ability to secure the conditions of meaningful self-rule for national communities in an age of globalization, and on having itself democratic decision-making procedures.

From this realist norm follows, so I argue, that the EU should take the form of a demoicratic confederacy. The normative ends of EDD require an EU regime with integrated judicial-administrative institutions. These autonomous institutions, however, raise concerns about the regime’s democratic legitimacy. On the norm of EDD, national parliaments remain the authoritative institutions. I argue that its members should take a central role in EU decision-making procedures in a European Senate (ES). Moreover, and in contrast to other demoicrats, this Senate would take the place of the European Parliament (EP). This would constitute a fundamental reorientation of the EU’s institutional architecture. This realist study thus cumulates in a novel normative model for the EU.

My argument unfolds in five sections. In the first section, I present the realist political thought of Bernard Williams from which I then derive a two-step method of normative theorization. In the second section, I discuss methodological approaches in the normative-theoretical debate on EU legitimacy to further draw out the novelty and comparative advantages of this realist method. The third section develops the norm of EDD from widespread patterns in public opinion research. In the fourth section, I spell out the normative model of demoicratic confederacy. The final section summarizes the argument.

From Bernard Williams’ Political Realism to a Two-Step Method of Normative Theorization

The first challenge is to offer a coherent account of the (neo-) realist school and its methods (Sleat, 2014: 331–332). I turn to the writings of Bernard Williams, who together with Raymond Geuss is one of the founding fathers of realism in normative political theory (e.g. Rossi and Sleat, 2014), to elaborate on realism’s core features and derive at a two-step method.

Williams’ political thought starts from an interpretative analysis of “politics” with the aim of establishing its conceptual features. Politics, so Williams argues, is the distinct human endeavor that deals with the unavoidable situation of disagreement in a polity. Without politics, disagreement could digress into a state of (civil) war. The First Political Question (FPQ), according to Williams, is how to secure “order, protection, safety, trust, and the conditions of cooperation?” (Williams, 2005: 3). This admittedly broad claim is fruitfully understood as the need for stable coercive structures. Such structures constitute a “functional response” to secure a civic order despite persistent disagreement in a polity (Stears, 2007: 545). These structures do not solve all disagreements but should contain, funnel, and, possibly, transform them, and, as such, they are the foundation of a political order.

However, coercive structures can become part of the problem because rulers can give in to the all-too-human desire to dominate their subjects (Williams, 2005: 10). Rulers can use their power to suppress parts of the population, such as in the case of slavery, or the entire population through either raw force or more manipulative means, as in the case of a dictatorship or 1984-dystopia respectively. Such situations constitute a state of war rather than politics. As Williams points out, even Thomas Hobbes did not believe that a reign of terror was a legitimate mode of rule (Williams, 2005: 4). What should have been the solution to the FPQ then becomes part of the problem.

Williams’ claim is primarily conceptual: domination is not politics. I would add that domination is also an inherently unstable state of affairs. Domination generates resentment, which can act as a justification for (even futile) acts of resistance, such as terrorism. Or, in the case of slavery, the American civil war illustrates that a polity can (temporarily) digress into a state of war.

Enter the normative concept of politics: legitimacy. Williams argues that political rule is legitimate when coercive structures meet the Basic Legitimation Demand (BLD). The BLD is an account for why coercive structures are a desirable solution to the FPQ. Subjects will accept, or at least acquiesce, to collective rule if they can make sense of coercive structures as a political order, which, moreover, attains normative ends, such as prosperity. Meeting the BLD thus transforms coercive structures into (legitimate) authority (Williams, 2005: 4–6). This demand should legitimize the coercive order, thus avoiding disagreements digressing into war. Yet, it can also delegitimize a regime when it is unable to meet its requirements.

An underappreciated form of domination follows from this conception of legitimacy: a community should be able to understand a proposed norm (Williams, 2009: 200). If agents cannot understand it, then the practical upshot is oppression of (parts of) the population. To further clarify, legitimate rulers require a story to be told to their subjects. This legitimation story should draw upon subjects’ cultural background. The latter is not necessarily a national culture, let alone a homogeneous one. Williams’ anthropological point is that legitimacy is part of heuristic processes of “making sense,” which necessarily take place in a historical setting (Williams, 2000). Thus, he argues that political philosophers can play Kant at the Court of King Arthur, but such a project is irrelevant. An appeal to pure human reason is unintelligible to the Knights of the Round Table deeply embedded in a medieval Christian cosmology. A ruler governing in accordance with these norms effectively engages in the oppression because Arthur’s knights would not be able to understand this legitimation story (Williams, 2005: 10–11).

At this point, a final conceptual observation is warranted. How about non-democratic forms of rule? Can a fascist political order be legitimate? Two questions allow realists to critically assess regime legitimacy. First, to establish whether rule is actually political, a realist can analyze whether the fascist regime is violently oppressing part of the subjects pace the earlier example of slavery. If so, the regime has become part of the problem and the fascist regime does not constitute a civic order. Second, for argument’s sake, let us assume fascist beliefs are widely spread across the entire community, hence no oppression is necessary. The next question is how this situation came about historically. Williams proposes a critical theory principle, which broadly holds that coercive power cannot be the sole source of subjects’ acceptance of that same power’s legitimacy (Williams, 2004: 225–232, 2005: 6; see also Beetz and Rossi, 2017). In short, if coercion or manipulation results in subjects’ acceptance of rulers’ legitimation story, such as the 1984 example, then it does not pass the test. However, if this fascist regime passes these tests, then a realist will have to bite the bullet; it is legitimate. On the realist assumption of disagreement, however, it is improbable that totalitarian regimes pass both tests.

The methodological consequence of Williams’ realist thought is that realists should draw upon material from history and the social sciences (Hall, 2017: 286). Taking into account practices is not a limitation in conceptualizing norms but a necessity (Williams, 2009). A realist theory that does not offer a functional response to the circumstances of politics within a historical set of practices is flawed because it is apolitical, and often moralistic, in character. Furthermore, if agents cannot understand the norm, then domination follows because the BLD remains unanswered.

Based on Williams’ writings, I propose a two-step process of normative political theorization. My proposal fleshes out a constructive realist method in a systematic and more detailed manner than currently available. Its exact application will differ per project; however, its central logic can guide future realist inquiries. First, the conceptualization of a norm should draw on cultural resources available to the agents to which it would apply. Second, the normative prescriptions that follow from this norm should be, if necessary, adjusted to ensure a stable political order. Each step contains a practice-sensitive and practice-insensitive phase.

The first step of a normative political theory is norm formation. In a Weberian vein, a realist should establish the cultural resources available to agents, thus ensuring the norm can govern them. In short, he should avoid playing Kant at the Court of King Arthur. After establishing these resources, a political theorist should use them in the development of a norm. Notice the switch in perspective, the theorist moves from an internal perspective—what agents can understand—to an external perspective—norm formation. To further clarify, the practice-sensitive phase should ensure that historically situated agents can understand the norm. The demand is not necessarily that anyone already believes in the subsequent norm (Sleat, 2014), but that the norm is intelligible to the members of the political community “now and around here” (Hall, 2017: 287)—wherever the now and here might be. For instance, the aforementioned knights of King Arthur would not have the conceptual resources in their medieval cosmology to understand a Kantian conception of legitimacy. In this step, a realist first establishes conceptual resources in the real-world practice to then develop a coherent norm.

The second step is the formulation of prescriptions. A realist should theorize prescriptions that follow from the norm and draw upon appropriate evidence to ensure they remain political. The realist must ensure prescriptions remain compatible with maintaining order under conditions of disagreement. Attainability factors—such as political will or sufficient resources—are not relevant. The first phase is relatively practice-insensitive because the theorist can formulate prescriptions independent from empirical considerations. The second phase requires practice-sensitivity. For instance, let us assume that a realist establishes that citizens in a polity believe that physical safety is the sole normative end of a political order pace Hobbes. Historical evidence demonstrates, as John Locke foresaw, that an absolutist Hobbesian state is likely to collapse into Hobbes’ feared “reign of terror.” Hobbes’ prescription is not a realist one because it likely results in domination rather than a political order. In this step, practice-sensitivity is vital to ensure that prescriptions remain political.

This combination of practice-sensitive and practice-insensitive phases in each step distinguishes this realist method from other prominent normative-theoretical approaches. I turn to the normative EU debate to demonstrate this method’s novelty and comparative advantage to other approaches.

The Normative Literature on EU Legitimacy: A Methodological Perspective

The EU’s alleged democratic deficit has been a catalyst for normative reflections on EU legitimacy (e.g. Bellamy and Castiglione, 2003; Føllesdal, 2006; Friese and Wagner, 2002). This debate, in part, reflects a sincere contestation of the appropriate normative yardstick to assess EU legitimacy. At this point, rather than analyze the normative content, I present three dominant methodological approaches and contrast them to political realism: principle-case methods, immanent methods, and practice-dependent methods. As these three methods feature prominently in international political theory, the EU legitimacy debate can be considered an exemplary case.

The first approach is the principle-case method (Friese and Wagner, 2002: 343). Many realists explicitly contrast their approach to this one, which they label either moralistic or liberal (see Rossi and Sleat, 2014). This method starts with an argument for a specific conception of a norm—often justice or democracy—which is then applied to the case of the EU (e.g. Lord, 2013). Glyn Morgan, for instance, argues that the value of security should legitimate the EU, which informs his argument for a European superstate (Morgan, 2005). In many such approaches, facts about practices merely demonstrate their normative claims (Neyer, 2012: 20).

A variation of the above method is non-ideal approaches, which focus on the attainment of ideals. Feasibility is their key concern. For instance, Jürgen Neyer (2010) argues that the human right of justification should legitimize the EU. He then illustrates that the norm has become institutionalized in EU practices. His methodological claim is that a norm “is achievable in the sense that its underlying normative principles are already well institutionalized,” thereby “holding up the link between ‘ought’ and ‘can’” (Neyer, 2010: 917). Achievability—feasibility—is thus the central consideration to include practices. Arguably, compromises are acceptable in the light of real-world constraints, yet the norm’s conception remains practice-insensitive.

A core realist critique of the principle-case method is that it posits unanimous agreement on a political aim or norm. It is almost intrinsically partisan because its point of departure is an argument for a particular norm. As such it is no longer a political but a moral norm. By contrast, realists do not posit any agreement on such ends, but rather turn to the conceptual resources of the participants in the practice first. Thereby, they aim to avoid partisanship on both the issues of selecting the relevant aims and their conceptualization. They do not advance a moral ideal but aim to formulate a political norm broadly acceptable in its historical context. As such realist methods hold more real-world promise in theorizing EU legitimacy, considering the EU’s political aims remain deeply contested.

Political realists are, however, not necessarily concerned with feasibility and being “realistic” in the everyday meaning of the word (e.g. Rossi, 2015; Sleat, 2016a). Realism refers to the conceptual features of real-world politics—coercion and disagreement—which they claim have been overlooked in much normative political theory. In this vein, the two-step method takes seriously the power dynamics of institutions in practice. Such facts matter not for normative reasons but for conceptual ones. As such its novelty lies in taking much more seriously the institutionalization of norms without giving normative weight to their attainability let alone the status quo (Sleat, 2014).

Immanent approaches constitute a second set of prevalent methodologies to theorize EU legitimacy. Their intellectual roots lie in the Frankfurt School. These scholars determine normative commitments in real-world practices to ground critical (Azmanova, 2013) and constructive (Nicolaïdis, 2013) reflections on the EU. To take one example, Jürgen Habermas (e.g. 1999, 2012) adopts his immanent method of rational reconstruction to theorize EU legitimacy (Patberg, 2014). After establishing normative fragments and particulars found in political practices, he reconstructs the norms to which these agents should be committed. His rational reconstruction results in a conception of EU democratic legitimacy grounded in political practices.

Political realism clearly has more affinities with this approach’s practice-sensitivity; however, a moralistic tendency is observable. Realists often ascribe a liberal position to Habermas’ political philosophy (Rossi and Sleat, 2014: 695). On this point, realists might well overplay their hand because his oeuvre displays an acute sensitivity to political and historical facts in his normative reasoning (e.g. Habermas, 1992). However, and here they would have a point, his writings on the EU are often more abstract and liberal. In theorizing constitutional legitimacy, the norm of liberal democracy acts as a filter to select the norms inherent in the practice. In my opinion, this method results in one of the most attractive democratic utopia for the EU: pouvoir constituent mixte (Habermas, 2012). In effect, this method results in the rejection of alternative conceptions of legitimacy found in practices. By contrast, to avoid domination and maintain order, the proposed two-step method remains sensitive to all possible sources of (EU) legitimacy.

Third and finally, practice-dependent methods have been used to theorize EU legitimacy (Sangiovanni, 2013, 2016). These political theorists stand in a Rawlsian tradition (Jubb, 2015a). They often focus on whether a demand for justice arises in the EU context without assuming justice’s relevance per se. In his study on solidarity in the EU, for instance, Andrea Sangiovanni argues that:

… we first need an interpretation of the point and purpose of the institutions that the conception is intended to govern, and the role of principles are intended to play within them. This interpretative step constraints what the content of justice is by telling us what it is for (italics Italics in original; Sangiovanni, 2013: 221).

One important observation is that disagreement plays no role in this study. To be fair, in his early methodological writings, Sangiovanni (2008: 156) does acknowledge this dimension and Williams’ influence. Another observation is that existing legal institutions determine acceptable conceptions of solidarity. Similar to immanent methods, practices impact on norm formation; however, practice-dependent scholars let existing practices determine conceptions of norms.

Unlike prominent practice-dependent methods, realists assume no intrinsic justificatory value in existing practices (nor necessarily deny it!). Coercive structures can be stable for the wrong reasons (Jubb, 2015a). A central question is which (institutional) practices can maintain order in the face of persistent disagreement, while meeting normative criteria for the proposed conception of legitimacy. This question lies at the heart of the second step of theory formation that I propose. The model remains independent from the real-world institutions and policies of the EU; it only refers to them as evidence to determine political institutions able to attain normative ends in the real world.

None of the arguments above imply that the other three methods are inept, but they highlight that a realist method, such as the one proposed, might contribute to our normative debates. The following sections aim to demonstrate this potential by applying it to the well-charted territory that is the EU’s democratic deficit.

A Realist Conception of EU Legitimacy: EDD

The first question for the stage of norm formation is to identify relevant materials to establish significant conceptual resources. In practice, citizens’ beliefs constitute the micro-foundations of EU-regime legitimacy (Zürn, 2016). Therefore, I turn to EU citizens’ evaluations to establish conceptual patterns within them.

A proven source to establish widespread patterns in EU citizens’ evaluations of regime legitimacy is the Eurobarometer (e.g. Zürn, 2016). One particular advantage of the Eurobarometer is that it reaches many EU citizens. This feature is essential to establish widespread patterns. However, the tool has two drawbacks. First, several authors critique its methodology and, related, argue that it purposefully fosters pro-European sentiments (e.g. Höpner and Jurczyk, 2015; Shore, 2000). Despite these critiques, the findings show a strong persistence of national identities, hence the impact of these biases seems limited at least for the purpose here. A second concern is that the nature of large-scale surveys precludes in-depth inquiry into respondents’ answers. Therefore, I draw upon qualitative research into this topic to flesh out and validate interpretations of Eurobarometer data on EU citizenship (Eurobarometer, 2013). Although I certainly do not expect to resolve all interpretive issues, this combination of sources should ensure a reasonable interpretation.

The empirical evidence converges around three patterns in EU citizens’ conceptual frameworks: (1) EU-regime legitimacy is evaluated based on (a) direct and (b) indirect (beneficial) outcomes for citizens, (2) national identity continues to legitimize state rule, and (3) national and EU-regime legitimacy should conform to the standards of liberal democracy.

The first pattern is that EU legitimacy is evaluated (1) based on an assessment of its outcomes for citizens. In this regard, (1–a) direct benefits for themselves as EU citizens are often emphasized in justifying European integration. Cheaper phone tariffs exemplify this discourse. Citizens recognize the achievements of European integration. Free movement (57%) consistently ranks highest of the Union’s achievements. What citizens primarily expect the EU to deliver on lies in the economic realm, namely, employment (18%) and quality of life (13%) (Eurobarometer, 2013). As other research shows, economic benefits are deemed particularly important in assessing EU legitimacy (e.g. Hooghe and Marks, 2004; Kuhn and Stoeckel, 2014). Therefore, even if only a minority of citizens perceive direct benefits (Eurobarometer, 2013), they evaluate regime legitimacy based on positive achievements. Still, direct benefits are only part of citizens’ conceptual framework in evaluating EU legitimacy based on outcomes.

Member States remain part of EU citizens’ conceptual framework when assessing EU outcomes. They evaluate (1–b) outcomes for themselves as part of Member States. For instance, the EU’s second most often mentioned achievement is peace among Member States (53%) (Eurobarometer, 2013). Some researchers conclude furthermore that EU legitimacy remains nested in its ability to empower national regimes (De Vries and Van Kersbergen, 2007). Moreover, benefits that citizens expect from the EU, such as employment, also remain a responsibility of national regimes (Eurobarometer, 2013). In the recent Euro-crisis, the nation-state (re-)emerged as the authoritative regime to protect EU citizens while trust in the EU dropped (Polyakova and Fligstein, 2016: 64). This supposed retrenchment to the Member State is unsurprising from the perspective of the second widespread pattern.

The second pattern is that (2) national identity continues to legitimize state rule. Cultural identity remains a core conceptual criterion to evaluate regime legitimacy. Despite its aforementioned EU bias, the Eurobarometer shows that a vast majority of citizens continue to primarily identify with the (sub)national community. In all, 42% identifies exclusively as a national citizen, and 47% identifies primarily as a national while acknowledging a secondary EU identity. Furthermore, a stable majority of 52% indicates not being attached to the Union (Eurobarometer, 2013). When European integration is felt as a threat to national cultures, then it delegitimizes EU rule (McLaren, 2002). National identities continue to legitimize sovereign state rule. In a similar vein, despite high levels of support for a common foreign policy, research indicates that EU citizens become upset to know how much has already been “pooled” (Díez Medrano, 2010: 332). Unsurprisingly, the recent transfer of these and other traditional state powers to the European level has been related to greater Euro-skepticism (e.g. Bickerton et al., 2015).

The lack of a widespread EU identity constitutes the other side of this coin. More than half of EU citizens recognize a European dimension to their identity (secondary 47%, and primary 9%); however, only 2% exclusively identifies as a European. Thomas Risse argues that we are witnessing the emergence of an EU identity (Risse, 2010, 2014). However, one blunt observation is that a significant percentage (42%) does not identify with the Union; hence, for the time being, there is no widespread EU identity. More problematically, referring to an EU identity might constitute a category error. The Eurobarometer (2013) concludes that a positive evaluation of the EU results in a positive attachment (72% vs 26%) and a feeling of EU citizenship (82% vs 18%). In the context of recent crises, legitimacy easily dissipates when the EU fails to deliver. These findings suggest that positive sentiments toward the EU rely on a positive evaluation of the outcomes, hence support follows a functional logic rather than an identificatory one (cf. Guibernau, 2011; Risse, 2014). Moreover, EU citizens indicate that this EU identity is “subordinate” to the national one (Antonsich, 2008: 517). Still, they support the EU in light of the attainment of common goals as well as a set of shared democratic values.

The third and final pattern is that (3) regimes’ conformity to liberal-democratic criteria influences the assessment of its legitimacy. Liberal-democratic values are part of EU citizens’ conceptual framework to assess regime legitimacy (Eurobarometer, 2013). On the national level, these democratic values ground a commitment to national self-determination. National sovereignty is often understood as a precondition for democratic self-rule (Beetz, 2015) as became apparent also in the Brexit debates. European citizens value European integration because they associate it with securing the sovereignty of national democracies (De Vries and Van Kersbergen, 2007). Important to note, citizens closely relate liberal democracy to national sovereignty.

EU citizens also evaluate the EU using liberal-democratic values. The Union’s decision-making structures are often assessed using core concepts of democracy, such as representation (Rohrschneider, 2002) and efficacy (McEvoy, 2016). The more positive a citizen’s evaluation of the Union’s democratic credentials, the more positive is his or her evaluation of its legitimacy. Nevertheless, the democratic deficit remains a widespread concern exemplified by the rise of Euro-skepticism across Member States. A further challenge is that EU citizens have distinct understandings of liberal democracy resulting in divergent solutions for this deficit (Nicolaïdis and Young, 2014: 1410). Despite these differences, the conceptual point remains that EU citizens use liberal-democratic values to evaluate EU legitimacy.

The analysis now moves from an internal perspective on legitimacy to an “external” one, hence I move from the practice-sensitive to the practice-insensitive phase of this step. These three widespread patterns in EU citizens’ conceptual framework, so I propose, can be aligned in a normative conception of “deep diversity.”

Canadian political philosopher Charles Taylor (1993) introduces the concept of deep diversity to describe the Canadian reality of multiple cultural communities in a single polity with distinct and potentially clashing conceptions of the aims of Canada’s political order. As an analytical category, deep diversity has been used in studies of the EU (e.g. Fossum, 2003; Nicolaïdis and Pélabay, 2009). However, unlike Canada’s case of three historically bound nations, the EU consists of more than 28 (sub)national communities with distinct historical trajectories that have not been part of the same polity before. Therefore, the diversity in the EU is “deeper” than in Canada (Nicolaïdis and Pélabay, 2009).

Crucially, for Taylor, deep diversity is not merely an analytical-descriptive category, but performs normative work in his democratic thought. As he puts it, “To build a country for everyone, Canada would have to allow for … deep diversity, in which a plurality of ways of belonging would be acknowledged and accepted” (Taylor, 1993: 183; italics added). Moreover, each tier of the political order should conform to liberal-democratic principles, hence liberal democracy trumps communitarianism in his analysis (see also Fossum, 2003: 339). In this normative vein, EU scholars have also drawn upon deep diversity when proposing normative theories of EU legitimacy (Eriksen and Fossum, 2007: 7–11).

I propose EDD as a contextualized variation of Taylor’s normative interpretation of deep diversity. Following Taylor’s logic, this norm of deep diversity justifies the persistence of distinct democratic communities in a single polity. Understood as such, it can align the three conceptual patterns in citizens’ evaluation of the EU. In line with patterns (2) and (3), nation-states should remain the primary locus of democratic rule. EU citizens identify with national democratic procedures, which result in distinct conceptions of the common good constructed from a plurality of visions. These diverse ways of belonging should be acknowledged and accepted, or at least be tolerated. In line with pattern (1–b), the EU regime should fulfill necessary and, where agreed upon, useful functions for maintaining self-rule of national communities. The EU regime does not generate legitimacy independent from the national tiers, but in service to them. Benefits for individuals can bolster its legitimacy (1–a), but not replace its original source. In line with pattern (3), the EU regime should conform to liberal-democratic principles. Whatever their exact formulation, these principles should not endanger the persistence of national democracies.

Although EDD should govern the present EU polity, it constitutes a practice-independent norm of regime legitimacy. EU citizens do not have to consciously subscribe to EDD; rather, it coherently incorporates prevalent conceptual patterns. Many citizens might upon reflection recognize (many of) their normative commitments in the norm of EDD, and might subscribe to it. For instance, moderate Euro-skeptics committed to some forms of European cooperation but opposed to political unification, such as French Gaullists, could quite easily accept this norm. Still, it does not cover every particular conviction; thus, space exists for disagreement. Furthermore, despite having an understanding of the norm, extreme Euro-skeptics and Euro-federalists might remain committed to their cause. On the norm of EDD, they should be able to pursue their political agenda through democratic channels. This accommodation of disagreement underlines that EDD remains political according to a realist definition. The next step is to assess whether it can inform the design of a legitimate political order for the EU polity.

A Realist Normative Model for the EU: Demoicratic Confederation

In this section, I spell out a normative model that can meet the normative demands of EDD while also constituting a functional political order. Europe’s coercive structures should conform to the norm of EDD. In addition, following a realist conception of politics, these structures should maintain order and funnel disagreement. Rather than two separate phases, I will systematically asses each set of institutions. First, I propose institutions that conform to EDD’s normative demands. Second, based on historical and social scientific evidence, I analyze whether these institutions can maintain order without collapsing into domination.

On the norm of EDD, national democracies should remain the political regimes pursuing conceptions of the common good in the EU polity. At the same time, EU citizens should respect other European conceptions of the common good as normative equals even if they disagree because these conceptions are, or at least should be, the result of democratic processes within national polities. Implicit in this logic is the requirement that national regimes are fully functioning democracies that can effectively implement collective decisions (Offe and Preuss, 2006). This ability to shape the polity also impacts the regime; therefore, it reflects at least in part a national conceptions of the common good. In sum, Europe’s national regimes’ democratic pursuit of common good legitimizes their authority.

The historical record demonstrates that state-bound democracies provide an institutional framework to pursue the common good within the context of persistent disagreement. Conceptions of the common good remain connected to a (sub)national public sphere, which offers the deliberative infrastructure essential for democratic conflict resolution (e.g. Bellamy and Castiglione, 2013; Crum, 2005). Institutional conditions for will-formation, such as voting, have been established at the European level; however, vernacular languages restrict opinion-formation to the (sub)national level. Despite the challenges posed by polarization and populism, the conditions for conflict resolution in mass democracies remain in place at the national level.

In addition, a democracy also requires effective coercive power to implement the common good. The sovereign state has successfully fulfilled this function in the past. Moreover, this centralized political order remains subject to national opinion and will-formation processes. As is evident, this institutional arrangement successfully implemented Europe’s distinct consensuses on social justice (e.g. Esping-Andersen, 1990; Ferrera et al., 2000) and liberal democracy (Nicolaïdis and Young, 2014: 1410). Autonomous state-bound democracies can effectively pursue conceptions of the common good within the context of persistent disagreement. This normative end can thus be attained in a political way.

EDD, however, also legitimizes a European political order to secure the conditions for the meaningful pursuit of national common goods in an age of globalization. The widely shared counterfactual scenario is that without some form of institutional cooperation, Europe’s nation-states would have been unable to maintain peace, and even if peacefully coexisting, they would have become economically uncompetitive (e.g. Habermas, 1999; Morgan, 2005). As such, the Union enables national self-determination by, on the one hand, creating peaceful and stable conditions in the polity and, on the other hand, offering a bulwark against external political threats and economic pressures. In a different vein, the liberal-democratic underpinnings of EDD reasonably impose a need to protect the democratic opinion- and will-formation at the national level. The recent concerns about the developments in Hungary and Poland illustrate that the liberal-democratic character of national arrangements cannot be taken for granted (Müller, 2015). These functional preconditions of national democracies necessitate and legitimize a European order.

To fulfill these functions, the EU regime requires judicial-administrative institutions with autonomous decision-making powers. Even according to a statist liberal-intergovernmental logic, effective cooperation between the Member States requires autonomous EU institutions to ensure credible commitments (Moravcsik, 1998: 67). This logic holds for policy domains essential to ensure interstate peace. Economic prosperity and safeguarding democracy, however, require more structural cooperation. While the governance of the single market does not necessitate unification, it does require more than an intergovernmental network (Fabbrini, 2015: 93–123). A European regulatory regime can fulfill this task (Majone, 2005). The maintenance of liberal-democratic rule in Member States is also plausible, but it would require a monitoring committee with the power to sanction. Such a committee could strengthen current oversight, while also taking into consideration national particularities (Müller, 2015). From this perspective, the creation of a supranational judicial-administrative apparatus is not only a historical reality (Lindseth, 1999) but a normative achievement.

The autonomous status of Europe’s administrative apparatus raises concerns about the supranational tier’s democratic legitimacy. A core challenge for EDD crystallizes: how to respect democratic pluralism, while also securing the preconditions of meaningful national self-determination through democratically legitimate EU institutions? To rephrase this question, which democratic form should the EU take?

In different historical circumstances, maybe in the future, a sovereign European superstate might constitute a legitimate institutional choice. A centralized European democracy could funnel conflicts through democratic procedures and rule uniformly. However, EDD legitimizes the EU regime because it is necessary to sustain national democracies rather than to transcend them. A federal superstate is likely to privilege particular national arrangements over others. The Economic and Monetary Union (EMU) arrangement is such a federalized arrangement (Bartolini, 2005). It privileges some countries over others, as became obvious during the Euro-crisis (Johnston and Regan, 2016; Streeck, 2014).

The intergovernmental model (e.g. Majone, 2005) is also not desirable. Generally, indirect legitimacy is insufficient to democratically legitimate autonomous EU institutions with increasingly far-reaching powers. More problematically, the intergovernmental Councils, often in collaboration with the European Central Bank and the European Commission, place insurmountable pressure on democratically elected governments. In 2011, the technocratic Monti-government replaced the elected Berlusconi government under European pressure. More recently, in 2015, the Greek referendum on bailout packages was effectively ignored. On the norm of EDD, the European tier should enable national self-determination, not make it obsolete. In sum, neither a (federal) European superstate nor intergovernmentalism is a legitimate model on the norm of EDD.

Demoicratic theory provides one of the most attractive democratic logics to institutionalize EDD’s normative demands. Demoicrats submit that Europe’s national peoples—the demoi—can together democratically legitimate EU decision-making power. This ideal, so empirical evidence suggests, might attain public legitimacy (e.g. Beetz, 2015; Cram, 2012), but disagreement exists on how to institutionalize it (Beetz, 2015: 39–40; Ronzoni, 2016). Some “federal” demoicrats argue that the EP and national governments are necessary to legitimize EU decision-making (Cheneval and Schimmelfennig, 2013; Nicolaïdis, 2013). Other “intergovermental” ones argue that national governments primarily legitimize EU decisions with auxiliary roles for the EP and national parliaments (Bellamy, 2013, 2016); in line with demoicracy’s central tenet of “demoi governing together,” EDD suggests that national parliaments as representatives of Europe’s peoples should legitimize EU decisions rather than a supranational parliament or national executives. The creation of an ES ensures national parliamentarians can influence EU decision-making; hence it is a desirable innovation. Moreover, it should replace the EP because the latter undermines national democracies. As such, I claim, my realist method cumulates in a distinction model: a demoicratic confederacy.

On the norm of EDD, in principle, national parliaments should bestow democratic legitimacy upon the European regime. Sincere disagreements exist about policies and instruments, in large part because the preconditions for meaningful national sovereignty differ significantly between, for instance, the Baltic and Mediterranean countries. To remain political, a EU demoicracy should legitimize coercion and funnel disagreement. Europe’s parliamentarian tradition (Piattoni, 2015: 9) requires that these legislative bodies should authorize and scrutinize executive bodies. National parliaments are thus the authoritative democratic institutions in the EU polity.

Institutional innovations, however, are necessary to ensure their effective influence in EU decision-making (e.g. Dawson and Witte, 2016: 219–221). Several avenues have been discussed in the literature, such as the “Danish model” of parliamentary scrutiny of executives before EU negotiations (Neyer, 2010: 914), a virtual third chamber made up of all national parliaments (Cooper, 2005), a right of legislative initiative for national parliaments (Kröger and Bellamy, 2016), turning the Conference of Parliamentary Committees for Union Affairs of Parliaments of the European Union (COSAC) into a constitutional actor (Neyer, 2014), and an ES (Dawson and Witte, 2016).

An ES is the most promising route to fulfill EDD’s normative demands and meet realism’s conceptual requirements. Such a democratic assembly offers national representatives an effective route to assert their authority, while accommodating disagreements (Dawson and Witte, 2016: 220–221). Similar to the German Bundesrat and US Senate, the ES would consist of multiple representatives from each national parliament. Three to five representatives per Member State should ensure a plurality of national voices. Each country should have the same number to reflect the strict equality of Europe’s demoi. The mode of nomination or election can differ in accordance with national conventions. The representatives should be national parliamentarians. Their dual membership ensures that they remain embedded in national opinion-formation processes. This design also accommodates cross-national debate creating space for transnational consensuses. Thus, a Senate ensures that national disagreements and cross-national ones are taken into account in European legislative debates.

Maybe most radically, EDD supplies no justification for the EP, while empirical evidence raises concerns about its normative desirability. Some demoicrats argue that a directly elected EP acts as an agent of the demoi as long as European parliamentarians are elected in national elections (e.g. Bellamy, 2013). However, an EP and an ES result in unnecessary double representation of the demoi, while it would also constitute another veto-player in an already dense institutional field (Lord, 2013). A serious normative concern is that the EP has been shown to undermine democratic life at the national level (Chapter 5 and 10; In: Brug and De Vreese, 2016); therefore, it directly conflicts with the norm of EDD. Therefore, I propose to replace the EP with an ES.

To some extent, I suggest a return to the pre-1979 situation in which the EP was not yet directly elected. In its tasks, however, this ES would take on and possibly have more tasks than the current EP. An ES would take on the role of the EP because a powerful legislator reflects the institutionalization of democratic values of EDD. Without entering into fine details, the European Council, Council of the EU, and Commission could continue to play executive and co-legislative roles in the EU. The recent exclusion of parliaments from decision-making procedures exacerbates existing concerns about executive dominance (Kreuder-Sonnen, 2016; Somek, 2015). The ES should take decisions on all matters of common concern and scrutinize these executive bodies.

This normative model constitutes a demoicratic confederacy, which is distinct from the federal and intergovernmental counterparts in demoicratic thought (Ronzoni, 2016). The absence of a supranational parliament is the reason to call the system a confederacy rather than a federal system, let alone a superstate. The normative model is not an intergovernmental one because EU decisions rely on national parliamentarians for democratic legitimacy rather than on national executives. This realist method thus points toward a novel position in an extensive body of knowledge.

Conclusion

Drawing upon Bernard Williams’ political thought, this article proposes a two-step realist method for undertaking normative theorization. Each step contains both a practice-sensitive phase and a practice-insensitive phase. This method is distinct from those prevalent in the normative literature in EU studies, as well as international political theory. Through its application, realist studies might contribute novel political theories. To demonstrate this promise, I applied this method to the EU’s legitimacy deficit.

This application results in two contributions to the normative literature in EU studies. A first contribution is a grounded norm of EU legitimacy: EDD. Although in need of contextualization, Charles Taylor’s normative interpretation of deep diversity can make sense of widespread conceptual patterns in EU citizens’ understanding of regime legitimacy. As such, EDD constitutes a realist norm to evaluate EU-regime legitimacy.

A second contribution is the normative model of a demoicratic confederacy. On the norm of EDD, national political orders should ensure the democratic establishment of common goods in the EU polity. European judicial-administrative institutions, however, should ascertain the preconditions for these democracies in an age of globalization. Rather than national executives or a supranational parliament, I argue that national parliamentarians in an ES should become the central body in decision-making at the European level. This demoicratic confederation constitutes a novel position in the normative debate on the EU’s institutional design.

This article covered a lot of ground. Some scholars might object to certain interpretations of empirical evidence. Any theoretical abstraction from empirical phenomena remains a road wrought with pitfalls. However, normative studies that aim to remain relevant in a constantly changing world cannot but venture into these murky waters in search for answers.