Democratic legitimacy in the era of fiscal integration

Abstract While several countries still struggle to return to sustainable growth and Euroscepticism has shown growing strength ever since the 2014 European Elections, Europe is slowly advancing on the path of fiscal integration. This paper reassesses how legitimacy is provided and why the advancing economic and fiscal integration constitutes a ‘genetic change’ of the Union. The second section discusses the functional deadlock emerging from the interaction between demos democracy and redistribution, which invites the EMU to make a fundamental choice between ‘convergence of the identities’ and ‘convergence of the economies’.


Introduction
European Integration, says the adage, advances through crises. Functional integration theory, in its evolution from early functionalism of Mitrany to neofunctionalism (Haas [1958] 2004) to neo-neo functionalism has maintained the assumption that each crisis would create the conditions to integrate new functions at supranational level. The justification for this integrative dynamics relies on effectiveness of institutions and economic advantages; further integration is often described as an 'unintended consequence' (Haas [1964] 2008) of previous measures, a step required to ensure the previous measure's effectiveness in a changing environment. Functional integration, as described, rarely relies on the concept of democratic legitimacy; in fact, mass politics is pictured as excluded from determining integration (Hooghe and Marks 2009). Schmitter (2002) and Haas ([1958] 2004) state that only final integration cycles would require integration of polities rather than policies; output legitimacy is the fundamental driver of EU legitimacy. Yet, the process of integration has somehow accelerated towards the uncharted territories of fiscal integration, mirrored by the rise of Euroscepticism in many countries. The slowing implementation of often-contested austerity measures and the renewed role of inter-governmentalism at European level have stimulated, once again, the debate over the alleged democratic deficit of the European Union. Against this background, this paper argues that the EU, rather than having entered a 'postfunctional' phase of integration, is now starting to deal with a different set of functions which require different sources of legitimacy, whose application is constrained by the relative importance of national identity.

Functional legitimacy
The question whether the European Union is suffering from a democratic deficit has increasingly gained scholarly attention since the decision of the Bundesverfassungsgericht (German Federal Constitutional Court, BVerfG) on the Maastricht Treaty in 1993. The BVerfG stressed that democratic legitimacy 'is not to remain a formal principle of accountability' (Bundesverfassungsgericht 1993, 18). Yet democratic deficit and legitimacy are separated concepts: A political system may suffer of a democratic deficit though still being fully legitimised in front of its citizens, either because of its good policies and/or because legitimacy stems from a different level of representation (other than democratically-elected parliaments). Thus, a conclusive investigation on the alleged democratic deficit of the EU requires a two-steps process. First, it must be verified that the EU is precisely in need of the form of legitimacy that is democratic legitimacy; second, if this is the case, it must be verified whether such democratic legitimacy exists or not.
Legitimacy of the EU as a broader concept has been also object of intense debating. The idea that the EU relies on indirect legitimacy has obtained a wide (although not overwhelming) scholarly consensus, and it is loosely parented with both neo-functionalist and liberalintergovernmentalist approaches. Although the scholarship on EU legitimacy is extremely rich and creative, indirect-legitimacy theses can be clustered around three main cornerstones: the 'regulatory state' argument of Majone (1997); the classical principal-agent framework for international organisations reformulated by Moravcsik (2002); and the outputlegitimacy argument made famous by Scharpf (1970Scharpf ( , 1998. Majone (1997) argues that democratic legitimacy is required only when the functions transferred to the supranational level entail redistributive policies among the participating countries. While this was clearly not the case before the crisis, it could have substantially changed with crisis-led integration (Majone 2014). Moreover, regulatory bodies work more efficiently when 'insulated' from politics, thus strengthening output-legitimacy (Majone 1997, 4). The insulation of decision-making is thus justified, like within nation-states, by the democratic qualification of the principals, so as long as member-states preserve their democratic stance, the EU shall be considered equally democratic (Moravcsik 2002). Similarly, Scharpf (1970Scharpf ( , 1998Scharpf ( , 2009) applies Easton's input-and output-legitimacy to the EU, suggesting that the latter has been thus far legitimated only through output-legitimacy (the effectiveness of its decisions) rather than by input-legitimacy (the democratic election of its leadership). The idea that a 'ranking' of competences exists within democracies has also been affirmed by the jurisprudence of the Bundesverfassungsgericht. In the 2009 Lisbon Judgement, the court proposes, in fact, a list of fields of competence that determine the content of democracy: Particularly sensitive for the ability of a constitutional state to democratically shape itself are (…) fundamental fiscal decisions on public revenue and public expenditure, the latter being particularly motivated, inter alia, by social policy considerations (…). (Bundesverfassungsgericht 2009, para. 252, emphasis mine) A loss of parliament's decision making rights in these fields would not represent a marginal transfer of competences towards an institution legitimated through a classical principalagent relation, but would instead constitute a true deprivation of substantial democracy for the concerned constituencies. As Fabbrini (2013) notes, fiscal and economic policy decisions are substantially adopted today by intergovernmental decision making, while 'the legitimacy of decisions taken on behalf of the EU cannot be a derivative of the legitimacy enjoyed by the governments of its member-states' . Consequently (and in absence of a body truly representative of the European people), the transfer of these competences would spoil the effectiveness of the citizens' political rights to vote.
Altogether, the combined arguments of Majone, Scharpf and the Bundesverfassungsgericht may be defined as functional legitimacy: the kind of legitimacy required (input or output) depends on the redistributive or non-redistributive nature of the function under observation. An excess of input legitimacy in policies which do not actually require it would harm legitimacy by hindering their performance (performance deficit), while a lack of input legitimacy in policies that require it would harm legitimacy by not adequately representing the parties, regardless of the contextual effectiveness of the decisions taken (democratic deficit).
The classical indirect legitimacy thesis has been widely criticised. Among others, Follesdal and Hix (2006) believe that (a) regulatory policies often entail distributive and redistributive effects not only between countries, but also within social groups. Raunio (1999) suggests that (b) the fundamental role of governments in the EU institutions has changed the power-balance within nation-states in favour of the executive. McNamara (2002) and Follesdal and Hix (2006) suggest that (c) the perception of the boundaries between regulatory and distributive policies may change, thus creating a democratic deficit. In particular, MacNamaradiscussing the democratic oversight of the ECB -collects evidence against (1) the notion that monetary policy has no redistributive effects (McNamara 2002, 54), and (2) the notion that central banking independence provides better outcomes in fighting inflation than democratic control over monetary policy. Finally, Agné (2007) indicates that (d) the 'chain of delegation' of the EU is both too long and irreversible, constituting a stretching and eventually a violation of the principal-agent framework and creating a deficit of democratic oversight.
Propositions (a-d) account for the large majority of the claims that the EU suffered from a democratic deficit. However, the claims that the EU suffered from a democratic deficit before the crisis are easily countered. Hix's analysis of distributive effects in regulatory policies fails to take into account the reality of EU policy-making. As noted, among others, by Scharpf (2009), the decisions undertaken at European level should not be considered individually but rather as parts of larger packages of regulations which aim at making every participating country better off with respect to the entire regulatory package. Therefore, while individual regulations may make certain countries worse off, this effect is counter-balanced by the negotiation of additional measures that are more favourable to the given country. If the whole country is better off in each round of negotiations, this implies that it disposes of resources to compensate -if it wishes so -the social groups that individually find themselves worse off. Whether this does happen or not is purely a matter of domestic preferences within the Parliament and not a defining feature of the EU construction.
Similarly, Raunio's claim (the EU decisional procedures shift power from the parliaments to the governments) overlooks reality on the ground. In fact, countries wishing to provide a stronger level of scrutiny of executives' decisions by domestic parliaments are not prevented to do so in any way: As Raunio himself acknowledges, certain countries have gone so far as to empower national parliaments' committees to adopt binding positions for their ministries in the Council. In other words, the EU as such does not strengthen the executives vis à vis domestic parliaments, which can exert any control they wish; rather, executives (and not the Parliaments) are held accountable, at EU level, for their actions. In addition, national constitutional courts can intervene to protect parliamentary rights, as done by the German Bundesverfassungsgericht in 2012. It follows that, again, it is a matter of national preferences to adopt such supervision procedures, not an institutional feature of the EU.

The impact of the crisis
As shown, if we narrow down our analysis to the EU as it looked before the crisis, the claim that its institutions were not legitimated is weak. Not only theoretical arguments in favour of democratic deficit before the crisis can be, to some extent, easily dismissed; but also trust in institutions remained relatively stable until 2007, while Eurosceptic parties maintained a low profile in many countries until 2009. In sum, despite a still-lively debate, there is ground to believe that, prior to the 2009/2010 crisis, the EMU was a fully legitimated undertaking (Schmidt 2015).
However, as a response to the European crisis, several institutions and procedures were introduced, which slowly moved the Eurozone into the minefield of fiscal integration. According to the Optimal Currency Area framework, since Kenen's (1969) reading of Mundell (1961) a currency area is in theory enabled to prevent the emergence of crises and to correct existing crises with pure market adjustment if there is high labour and capital mobility, high prices flexibility, and reduced public debts. In the absence of these features, a currency area stuck by crisis would be endangered if appropriate institutions are not in place. De Grauwe (2007) refers, in particular, to fiscal transfers to smooth consumption adjustment and to supranational budgets to grant the effectiveness of public stabilizers even in case of national debt under stress. Direct fiscal transfers are, however, only one example of fiscal integration: Merler and Pisani-Ferry (2012) show that the solution to the crisis must include either a direct fiscal union (implying both fiscal stabilizers and enforced fiscal rules), a full-fledged banking union, or a central bank with a widened mandate. While, in principle, these are different institutions they all imply a step towards redistributive policies. Despite their functional differences, all the aforementioned options require some degree of fiscal integration; no credible solution excludes it. Fiscal transfers between countries, by definition, require fiscal integration. A full-fledged banking union is in need of a common fiscal backstop for the joint deposit guarantee (Pisani-Ferry and Wolff 2012); and -as shown by Sinn (2011) and by De Grauwe and Ji (2013) 1 -a wider mandate for the ECB would have fiscal implications. In other words, a permanent solution to the crisis requires the EU to move towards some degree of fiscal integration, either explicitly (fiscal union) on implicitly (banking union, ECB).
Since 2011, the EMU has shown some progress: it has created a series of bail-out mechanisms, it has introduced strengthened coordination of fiscal and economic policies through the Six Pack and the Fiscal Compact; it has introduced the first two steps of a banking union (still lacking a joint deposit guarantee); and it has changed its monetary policy to the point that an implicit guarantee on public debt is de facto in place; it has introduced an investment instrument to boost growth. Many of these measures have direct or implicit redistributive effects, either on a (theoretically) temporary basis (through loans) or permanently (through ECB's debt purchasing or through EFSI (European Fund for Strategic Investment) grants and contributions. However, none of the instruments introduced since 2011 is tailored to provide shock-absorption or redistributive functions (albeit several instruments proposed, but not yet implemented, have this aim); furthermore, none of the instruments introduced is endowed with a supranational democratic governance: when the governance system is supranational (like in the case of the OMT and of the EFSI) the decisions are taken by Commission or ECB officials without Parliamentary consent; in other instances, either the system is purely intergovernmental, or the Commission and the Council share the decisional burden, but without involvement of the European Parliament.
Moreover, the Eurozone design is not yet in equilibrium and will be modified again in the future. As indicated by the Five Presidents Report (2015) and by several national proposals (including the Bundesbank-Banque de France proposal for a Eurozone Financial Minister, 2016), the policy debate is beginning to align with theoretical prescriptions concerning the need of EU-wide stabilisation instruments (for instance, a European Unemployment Insurance, Claeys, Darvas, and Wolff 2014;Biabiany, Marconi, and Maselli 2015). Rather than a question of 'if' , it is the question of 'when' a degree of explicit fiscal integration will be introduced in the EMU design. 2 Fiscal integration touches the very core of democratic systems, whose Europeanisation -either in the form of a strengthened cooperation or with a full transfer of competences -does not simply represent a marginal incremental change in the EU competences. Rather, it constitutes a paradigm shift in the nature of the EU and in the way the EU itself shall be legitimised; nevertheless, the same shortcomings seem to apply to the measures formally proposed thus far to strengthen the EMU''s fiscal integration, lacking input legitimacy.
This slow but continuous process of fiscal integration generates controversial effects. On the one hand, fiscal integration it is not sufficiently developed yet to re-establish Europeanwide output legitimacy. On the other hand, the first elements of the new economic governance aimed to constrain domestic behaviour and increase the power of insulated institutions in fiscal matters, generating a democratic deficit. In other words, current EU anti-crisis measures, while having limited European-wide redistributive effects (no output legitimacy at EU level), greatly reduce the space of action on economic policy of national lawmakers, constraining their capacity of enacting redistributive policies at domestic level (Scharpf 2012, 22). By doing so without successfully addressing the crisis, the EU has both reduced its output legitimacy (Schmidt 2015) and hindered the input legitimacy of national governments (Scharpf 2012, 18), contributing to the rise of Eurosceptic parties in several countries (Nicoli 2016a). In other words, the crisis has created a 'twin legitimacy deficit' of European integration: not sufficient redistributive policies to achieve output legitimacy, but sufficient progress towards insulated decision making on fiscal policy to fail to reach input legitimacy. While 'performance deficit' drove the neofunctional stage of integration towards the construction of insulated institutions, democratic deficit is likely to be the driving force of the postfunctional phase of integration.
Being parliamentary control on fiscal matters essential for democracy, the Bundesverfassungsgericht notes that -when there is no parliamentary control of fiscal policy whatsoever -we are outside the democratic framework. If no parliament controls fiscal policy, there is no democracy. In principle, then, fiscal integration may endanger the democratic qualification of the EU: transferring fiscal powers to the EU without strengthening its democratic legitimacy would deprive the European citizens of the decision making power over an issue absolutely essential for democracy, as noted by the Bundesverfassungsgericht. It follows that the Euro Area is now in a 'grey zone' , suffering from a 'twin legitimacy deficit': having introduced only some limited forms of fiscal integration, it has been unable to permanently solve the crisis, lacking output legitimacy. However, the EMU is not anymore a simple currency area, precisely because of the (partial) progress towards fiscal integration; therefore, output legitimacy is (in theory) increasingly not sufficient, but input legitimacy is still lacking. In other words, given that the EMU has begun to progress towards fiscal integration, its democratic qualifications would greatly benefit by some more input legitimacy, but it could keep relying on output legitimacy, provided that fiscal integration remains partial; however, precisely because integration has been partial, output legitimacy is vanishing both domestically and at European level. To re-establish output legitimacy, integration should go forward towards the completion of fiscal integration, but if this were to happen, then, output legitimacy would not be sufficient if the EMU is to remain a democracy (at least according to the Bundesverfassungsgericht's understanding of democracy). However, while output legitimacy on fiscal issues would violate the democratic condition of political equality, the introduction of input legitimacy can hardly be handled without raising problematic issues. At the same time, as noted by the Bundesverfassungsgericht in 2009, the simple transfer of fiscal power to the European Parliament would not preserve the political equality criterion, as the European Parliament is elected based on geographical constituencies with a regressive proportionality principle, and there is no second federal chamber where the equality of states is represented (Bundesverfassungsgericht 2009, para. 274 and following).

The issue of identity
More importantly, the transfers of economic and fiscal powers to a supranational assembly would be insufficient if identity is a precondition for democracy when democracy has redistributive implications. From Hume's Treatise onwards (Cohon 2010, 7), identification plays a role in one's attitudes towards solidarity; a concept later reflected in Ferdinand Tönnies' Gemeinschaft, a society built upon shared community beliefs rather than upon rational contractualism. In the second half of the twentieth century, the debate over the societal boundaries of democracy and redistribution has revamped on both sides of the Atlantic. In North-America, the renewed discussion addressed the issue of the origins and limits of redistributional justice, being marked by contributions like Rawls' A Theory of Justice (1971), MacIntyre's After Virtue (1981) and Walzer's Spheres of Justice (1983). In Europe, the debateignited by the Controversial Maastricht decision of the BVerfG -focused on the limits of a democratic transformation of the European Union alongside the emergence of a collective European Identity. In this context, the so-called no demos thesis was developed, among others, by Bryde (1994), Kielmansegg (1994Kielmansegg ( , 1996, and, to a lesser extent, by Grimm (1995), Weiler (1996Weiler ( , 2001 and Zürn (2000). As Zürn writes, the foundation of the no demos thesis is that 'majoritarian decision-making is hardly achievable beyond the national level since it requires some form of collective identity that includes trust and solidarity ' (2000, 195). Some proponents of the thesis, like Bryde (1994), ground the demos into ethnical, linguistic and cultural features, which create links of solidarity among the demos members, thus justifying redistribution through majority voting. Outside the demos, these boundaries disappear, making the deployment of a working democracy impossible.
Weiler's depiction of the no-demos thesis is likely the most articulated: 'democracy does not exist in a vacuum (...) if there is no demos, there can be no operating democracy ' (1996, 523). The nature and the ultimate goal of the European constitutional construct (its telos) is indeed not to proceed towards the creation of a single European demos, but rather to preserve peacefully the distinct demoi. 3 Consequently, a true majority-voting democracy at EU level is not possible, and the departure from unanimity vote occurred with the Single European Act coincides with the beginning of the crisis of the European construction: 'no matter how close the union, it is to remain a union among distinct peoples' (Weiler 2001, 14).
The no-demos argument, as reconstructed by Weiler, is composed of a positive statement ('Europeans are not a demos') and a normative claim ('Europeans shall not become a demos'). While the normative claim is controversial, many would agree with the empirical claim. For instance, the Bundesverfassungsgericht (2009, para. 279) writes: 'The democratic fundamental rule of the equality of opportunities of success ("one man, one vote") only applies within a people, not in a supranational body of representation, which remains a representation of the peoples linked to each other by the Treaties' . In a way, the 2009 ruling on the Lisbon Treaty constituted a compendium of the defining principles of classical democracy, providing a (legally binding) interpretation of the relations between the fundamental political competences sovereignty, parliaments, demos, and representativeness. The Bundesverfassungsgericht appears to share the positive argument of the no demos thesis concerning democracy, including the basic arguments for which democracy (majority-voting) is possible only within one demos. Although the Court does not rule out the emergence of a European demos, especially through institutional reforms, it recognises that such a demos is still yet to emerge; and thus only inter-governmental politics are possible provided the existing institutional balance.

A legitimacy trilemma
In light of the no demos thesis, the lack of shared European identity generates a major issue: either we decide to proceed in securing the EMU with fiscal integration giving up any claim of democratic control over it, thus creating a purely inter-governmental (output-legitimised) fiscal union; or we decide to maintain democracy at the national level, avoiding fiscal integration, and therefore receding from the monetary integration. In other words, the EU is facing a true governance trilemma (Figure 1) which, prima facie, mirrors Rodrik's (2007) setting. We can preserve democratic decision-making, the monetary union, and exclusive national identity, but only two of these features at a time. Pursuing all three features simultaneously would prevent any resolution to the crisis. In fact, democratic decision making and fiscal integration are consistent, but this would require the creation of a European Parliament that is enabled to deal with fiscal and economic policy. This would be in contrast with an exclusive national identity, which assumes that fiscal policy cannot be the object of democratic decision making outside the demos because of its redistributive implications. Similarly, we could maintain an exclusive national identity and proceed towards fiscal integration at the condition that any decision on fiscal and economic policy that is undertaken in the newly established fiscal union is taken by consensus of governments, thus removing democratic decision making from controlling these policies. Finally, we could maintain an exclusive national identity and a majority-voting control over fiscal and economic policies, only if these are not integrated at the EU level. The next sections will discuss each of the alternatives in more detail.
The implications for the European Union and its quest of legitimacy differ under each of the different options. From this point onwards, we enter a rather uncharted territory. While Figure 1 may look familiar, it differs in some fundamental elements from the global economy trilemma proposed by Dani Rodrik (2007Rodrik ( , 2011 and operationalised by Crum (2013) for the Euro Area.
The legitimacy trilemma proposed here, albeit dealing with a comparable set of problems, differs in three fundamental aspects from Rodrik /Crum approach. 4 First, the domain of effectiveness of the legitimacy trilemma is extremely restricted when compared with Rodrik/Crum's own approaches. In section A it was demonstrated that not any level of 'globalization' threatens democracy, but only sufficiently advanced degrees of integration that put redistribution at stake. 5 While this may sound trivial at a first glance, it is worth noticing that in Rodrik's settings, functional integration is never compatible with democracy; the European project seems to be in conflict with democracy since its inception. This paper instead argues (along with Schmitter (2002) that only the final phase of integration, putting redistributive policies at stake, requires a leap forward from integration of policies to integration of polities (Haas [1958(Haas [ ] 2004. Second, the concept of nation state in Rodrik's trilemma is misplaced. A fully sovereign Nation State constitutes already a political system in 'functional equilibrium' (Schmitter 2002) where all key competences of sovereignty are maintained at the same governance level. In a Nation State with full fiscal and monetary sovereignty, very strong sense of identity, and working democratic institutions, there is no trilemma; similarly, if Europe had these three elements together, there would be no trilemma, regardless of whether it acquires the status of a 'State' or not.
In other words, rather than a vertex of the trilemma (like in Rodrik 2007), classical Nation States, being the milieu of redistributive politics in a geographical area occupied by a single 'people' constitute one the trilemma's possible solutions; in classical Nation-States the trilemma is solved by creating a single demos kept together by a strong bond of identity, hence nullifying the no demos thesis. In Rodrik's setting Nation States constitute a vertex of the trilemma, but at this point of the analysis it should be clear that relinquishing national sovereignty without addressing the identity issue does not solve the trilemma at all; it rather generates new problems, as witnessed by secessionist pushes in existing nation-states with sizeable minorities. A proper vertex for the trilemma, instead, requires 'exclusive national identity'; solving the trilemma for integration and democracy requires the creation of a degree a shared identity, overcoming the exclusiveness on the national bond.
In Rodrik's setting, a European Federation would constitute a solution to the same trilemma on a larger geographical area, regardless of the existence of some identity bond between its participants; it is simply a matter of allocation of functions to the right level of governance, as if functional integration could shift seamlessly from non-redistributive to redistributive functions. In the setting here proposed, a European Federation, if not grounded into a 'self-understanding' of a European people, would result in increased detachment, Euroscepticism and resentment against the European integration process, not differently from what already happens in several existing nation states struggling with regional independence movements.
Finally, a third reason to believe that the legitimacy trilemma here proposed is better suited to model the Euro Area problems than the Rodrik/Crum setting, is that the setting proposed here is grounded on well-established theoretical work. Fiscal integration (of some kind) as a solution for un-optimal currency areas has been discussed since Mundell (1960Mundell ( , 1961, Kenen (1969) and several European Commission reports (Werner 1970;McDougall 1977;Padoa-Schioppa 1987;Delors 1988) and constitutes one of the cornerstones of today's debate on European economics; the idea of democratic legitimacy and redistribution is widely discussed since Majone (1997) and adopted by the Bundesverfassungsgericht in 2009; the same Bundesverfassungsgericht contributed, since 1993, to the lively debate on the demos thesis.
For these reasons, even if Rodrik's work may provide some insights concerning the longterm outcomes of the trilemma's pairs, it is insufficient to provide a correct understanding of the crisis of the European Union. A discussion of the trilemma's solutions follows in the next subsections.

National identity and democratic principle: a 'disintegrative spill-back'?
The first option for resolving the trilemma implies the choice of democracy and exclusive national identity, therefore rejecting fiscal and monetary integration. Is the dissolution of the Eurozone a viable option? As discussed below, theory and historical experience seem to suggest that the dissolution of monetary unions would not move the institutionaleconomic system into a stable equilibrium. Further institutional adjustment would be needed, endangering the European Union and the Single Market.
What are the implications of a Euro-breakup? Besides the short term effects of a similar decision (which according to several studies -UBS [Deo, Donovan, and Hathaway 2011] and Bertelsmann Stiftung [Belke 2012] -might acquire a global magnitude several times higher than the collapse of Lehman Brothers), removing the Euro would endanger the long term stability of the EU as a whole, possibly unleashing a dis-integrative dynamics of the EU. The EMU, in fact, was introduced for specific functional reasons. While there was a sincere political willingness to push forward European integration, there was also a clear macroeconomic concern behind EMU inception: achieving a sustainable monetary stability in Europe to prevent the collapse of the Single Market. Both attempts to create a quasi-fixed exchange rate system after Bretton Woods (the 'Monetary Snake' and the European Monetary System) had collapsed; a common currency was (and is) required if fixed exchange rates (needed to prevent exploitation of the common market rules, Meade 1957) and capital mobility (at the heart of the single market agenda in [1985][1986][1987][1988][1989][1990][1991][1992] are to be kept together. In other words, the Euro is the only possible solution to Mundell's (1961) famous 'Impossibility Trilemma' in favour of free movement of capital and fixed exchange rates (thus repelling independent monetary policy). The collapse of the EMU would raise the same question again, namely how to create monetary stability in Europe, without a viable solution. 6

Monetary union and national identity: 'convergence of the economies'
The second choice with respect to the legitimacy trilemma concerns monetary union and exclusive national identity, leading to an inter-governmental type of fiscal integration (or executive-federal in the words of Habermas). Clearly, it would produce a shift from input (democratic) legitimisation to an output legitimisation of fiscal policies. Indirect legitimacy for insulated institutions worked as long as the EU could be qualified as a regulatory, nonredistributive polity, but adding the redistributive functions would constitute a fundamental change of the Union and, if pursued without democracy-enhancing reforms, would constitute a net loss of democratic stance (Majone 2014). Nevertheless, such arrangement may work in practice. In this setting, the formal equality of states, including their veto powers on fiscal matters, shall be preserved in respect of exclusive national identity. However, the credibility of commitments is still essential: Credibility must cover not only the willingness of peripheral countries to adopt reforms, but also the willingness of core countries to pay. If each fiscal transfer were to need 28 majorities in 28 national parliaments, the whole system would end-up being completely non-credible in face of financial markets and of the voters. Moreover, the whole 'grand-bargaining' still needs to be protected by electoral politics. Fiscal and economic integration (with legal responsibilities for both net recipient and net contributor countries) is thus again insulated and protected by unilateral changes through constitutionalisation in a fiscal treaty with redistributive features. Indeed, the 'insulation logic' observed by Majone constitutes not only a requirement to achieve effectiveness avoiding decisional traps, but also a strategy to ensure compliance (Wolf 1999, 341). The goal is precisely the de-democratization of a given competence. The respect of national sovereignty turns a purely fictional stratagem into an insulation setting. Clearly, the respect of the no demos thesis is apparent: In order to grant to each national demos equal and non-majoritarian rights on fiscal and economic policies, these policies must be excluded from the fields of action of democratic policy-making and being constitutionalised.
Such an arrangement, albeit eventually effective, would face two major problems on the legitimacy side. First, constitutionalisation of fiscal policy strengthens' Vivien Schmidt's (2006) paradox: the EU adds a new 'policy without politics' , and MS lose their core policy for politics to deliberate on.
Second, precisely because the new-born fiscal policy is insulated, it would be legitimated on the output side. However, the benefits (output) should be positive across the different constituencies in order for the agreement to remain legitimated. In other words, output legitimacy requires horizontal effectiveness of the implemented policies. This requirement generates two problems: how to deal with (1) variations of constituencies' interests and preferences after the constitutional agreement, and (2) time inconsistency issues related to differentiated timing in the emersion and/or perception of benefits?
In an intergovernmental setting, we can assume that, if an insulated policy was agreed upon in the first place, the ex-ante expectation is that it would produce positive results in all the parties agreeing upon it (either directly or as a part of a 'grand bargain'). However, if that policy ends up producing welfare losses in a given constituency, output legitimacy is not achieved-as suggested by Scharpf (2012Scharpf ( , 2014 the lack of horizontal effectiveness during the Eurocrisis hindered the output legitimacy of the EMU. Of course, no constituency would (in theory) agree with a 'quasi-constitutional' agreement that is structurally against its interests. For a single fiscal or economic decision to be output-legitimised, therefore, a relevant level of ex ante economic convergence is required, not only because it would be unlikely that the insulated institution would be endowed with powers of taking a decision structurally against the interests of a constituency, but also because if this were to happen, that decision would still not be output-legitimised in the losing constituency. Now, provided that fiscal and economic integration are a requirement for further convergence in the EMU, we also face a problem of time inconsistency: can we legitimise economic and fiscal integration with a source of legitimacy (output legitimacy) requiring, to be effective, horizontal effectiveness, and therefore ex-ante convergence, which in turn requires ex-ex-ante fiscal and economic integration? Temporal expectations play here an essential role. Agents may support losses in actual democratic stance only if concrete advantages show up timely. The higher the loss, the faster results shall be delivered. 7 Naturally, the 'timely' definition may vary across regions and individuals: Some may accord high values to short terms gains, while others may prefer long-term advantages. However, these arrangements are credible only if we focus on the long term, that is, if their content does not depend on the national political cycle. 8 It follows that, in order achieve an appreciable degree of output legitimacy in the long run, an insulated fiscal and economic Union would have to deploy substantial instruments to foster convergence in the short term, thus enhancing long-term horizontal effectiveness (and thus widespread legitimacy) of a common insulated fiscal policy.

Monetary union and parliamentary democracy: 'convergence of identities'
There is a third possible outcome of the trilemma, a federal integration pattern combining supranational democracy and monetary union. Several concerns can be raised, however, despite the clear political appeal of such an option. These concerns rely on a simple observation: creating a shared European identity is a complex task. There is little doubt concerning the fact that Europeans are not (yet) a 'people'; integrating further may strengthen the hand of hard-line Eurosceptic parties, and -not differently from many democratic fiscal federations like the UK, Spain and Belgium -the EU would the suffer widespread secessionist pressure. Even functionalists with rather federal feelings, as Haas ([1964] 2008) and Schmitter (2002), consider a quick federalisation of the EU unlikely to happen without traumas. The former, in his analysis of international integration patterns, considers that federal integration would succeed only after the instauration of a 'liberal nationalism' . In other words, after a conflicting phase between original national identities, the supranational authorities may succeed to build up a true federation overcoming the nationalist oppositions with the creation of a strong and shared body of beliefs and values. In addition, Schmitter considers that governments would resist with some forms of 'power protectionism' boosting national identities.
Identities, however, are not fixed in the hyperuranium: they can evolve over time and they are constructed through a dialogical process; an appropriate configuration of the political arena may provide the space for a shared identity to emerge endogenously. If the power of an elected assembly (such as the European Parliament) over fiscal issues is increased, then political parties will increase their efforts to take control of the assembly, investing more in collecting votes, mobilising consensus and building up organisational structures. Because of the organisational efforts and the search for consensus, two effects are likely to arise: (1) increased political polarisation, and (2) increased participation rates in the elections. At this regard, Hix, Noury, and Roland (2007) recognise that a relevant increase of the parliaments' powers to shape the outcomes of the political process can indeed increase their legitimacy. Furthermore, a strengthening of the Spitzenkandidaten process, and a stronger link between the results of the European Elections and the economic policy enacted at EU level, may contribute to the process: the introduction of a true parliamentary control over fiscal and economic policy at EU level would create the incentive for the construction of collective preferences. For instance, amending existing provisions on the Economic Dialogue within the SGP legislation, and endowing the European Parliament to take an active role in the drafting of the Annual Growth Survey, not only would not require treaty amendment (only secondary legislation change) but also would contribute to set up the institutional conditions for the emergence of an endogenous dynamics of identity-building through the political process.
Moreover, two indirect effects may emerge. First, economic preferences are an important part of our collective identities: an appropriate European-wide fiscal and economic policy may foster economic convergence, thus generating a re-alignment of economic preferences and creating the foundation of a shared identity. Second, those who enjoy the benefits of integration may also start to feel more European; along with classical neofunctionalism (Haas [1958] 2004) as integration produces benefits, identities and allegiances to Europe are built. Following this path, even if an early decision on 'non-democratic' fiscal and economic integration is made, the ensuing convergence of the economies may entail convergence of (part of ) our identities as well, making easier to introduce democratic provisions later in the process.

Conclusions
While the pre-crisis neofunctional dynamics of integration is driven by the need of addressing a performance deficit, thus re-establishing output legitimacy though more insulation, the Eurocrisis -by introducing a redistributive component of integration -has unleashed a postfunctional integration dynamics, driven instead by the need of addressing the Union's democratic deficit. The paper has shown that while talking of a democratic deficit before the crisis may have been redundant according to both theoretical reasoning and empirical evidence, the wave of crisis-led integration that occurred since 2010 is changing the picture. The EU is not simply facing the need for further democratisation: Instead, we show that the Union is trapped in a policy and legitimacy deadlock from which there are no easy ways out. Fiscal integration (or monetary dis-integration) is unavoidable; and it will substantially transform the Eurozone into a redistributive polity. Depending on the institutional setting, the ensuing conflict could either be qualified as a national struggle or as a societal struggle.
Similarly, inaction will equally fuel social conflict, as indicated by the rising populism in many EU member states.
The analysis suggests that the EMU is caught in a 'functional trap': none of the possible choices is costless, and it is still unclear what would constitute the resting point of both further integration and disintegration solutions on the trilemma. Dismantling the Monetary Union not only could push the continent towards a new financial disaster, but also release disintegrative dynamics which ultimately would end only with the collapse of the Single Market for goods and capital; once unleashed, the dis-integration dynamics may be hard to stop, due to the incremental nature of European functional integration. Proceeding with output-legitimated fiscal integration (thus avoiding federalism and the political costs it entails) would require a substantial convergence of the economies, so that horizontal effectiveness is achieved. However, such a solution may have huge costs in terms of democratic legitimacy, may fuel national conflict and may constitute a solution not lasting sufficiently long to achieve true economic convergence. In fact, it is unclear for how long 'fiscal policies by stealth' , mediated through the ECB balance-sheet, can be maintained. 9 Even if outputlegitimated fiscal integration were introduced, the risk that short-and medium-term losers of such process would rebel before accessing long-term gains is high. Finally, a federal integration pattern would need convergence of the different identities to achieve effective input legitimacy. Constructing a fiscal federation without a degree of shared identity is risky as it will require a democratically legitimised sharing of resources among communities that do not have a truly unified public sphere, identity, or sense of common belonging. Experience shows that this is hard to achieve without political conflict and potential break-ups even within long-standing established communities. Finding a source for a shared European identity seems a central priority and a condition for further integration. Alternatively, the EU risks facing the emergence of a dysfunctional system, hostage of minority blockages in national or even regional parliaments and characterised by recurrent instability on the financial markets. Ultimately, such sub-optimal solutions may not be sustainable and also may end up provoking the collapse of the EMU.

Notes
1. With symmetrically different results concerning the direction of fiscal transfers in comparison to Sinn (2011). 2. Implicitly, it is already in place through the ECB balance sheet and the upcoming Single Resolution Fund. 3. 'Europe is not yet a Demos in the organic national-cultural sense and should never become one' (Weiler 1996, 528). 4. Rodrik (2007) trilemma features globalisation, democracy, and the nation state; Crum (2013) operationalisation features the Euro, democracy, and national determination. 5. This may change across time and societies. As noted by Moravcsik (2002, 614), although monetary policy might have redistributive implications, it was possible to create the ECB because the redistributive function of monetary policy was widely rejected at the time of its creation. 6. Nicoli (2016b) provides a detailed reconstruction of the process of economic integration as a series of sequential trilemmas. 7. Empirical studies on the European Identity and the legitimacy of the institutions demonstrate that there is an extremely sensitive component in the individual attitude towards EU institutions in relation to short-term achievements (see Bruter 2003, 116).
8. In practice, as long as multiple constituencies exist, and as long as these constituencies hold different interests, it may be difficult to find an agreement that satisfies the long-term commitments condition and the short-term effectiveness condition for any constituency. 9. At this regard, the Bundesverfassungsgericht pointed out, in its decision of 21 June 2016, that the legality of the OMT is subject to conditionality and must be reversible once the crisis is finished. Mr. Draghi himself pointed out in several occasions that the ECB action is temporary and that governments have to take the lead in fiscal integration.