Assessing the Legitimacy of the EU's Treaty Revision Methods


Thomas Risse
Otto Suhr Institut für Politische Wissenschaft
Freie Universität Berlin
Ihnestraße 22
Room 105
Berlin 14195


How legitimate are the EU's treaty revision procedures, that is, IGCs and the Convention? Since the rejection of the Constitutional Treaty the recurrence to intergovernmental procedures appears as one way out of the current dilemma. We argue instead that the Convention method is not to blame for the current crisis. On the contrary, it increases the legitimacy of EU constitutionalization considerably.


How legitimate are treaty revision procedures in the European Union? Since the 1980s, the EC/EU has been in a semi-permanent reform and subject to various revisions of its constitutional structure. Several multilateral negotiations were convened in order to amend the founding treaties or to create new policy areas outside the Communities, respectively. The considerable effect that the EU's policies have on the European citizen's daily life necessitates asking whether the method of these adjustments is legitimate. In this article, we focus on two basic models for treaty revisions, namely Intergovernmental Conferences (IGC), on the one hand, and the combination of an IGC with a Convention1 as its preparatory body, on the other. We therefore ask whether and how the Convention method such as the Constitutional Convention adds legitimacy to the original intergovernmental treaty-making and constitutionalization procedure. Yet, since the rejection of the Constitutional Treaty in France and the Netherlands in 2005, the excitement about the Convention method has fallen silent, and a recurrence to good old intergovernmental procedures in turn appears as the way out of the current dilemma. We argue, however, that the Convention method increases the legitimacy of EU constitutionalization considerably by

  • • shortening the lines of accountability from the citizens to their representatives at the EU level through the participation of national and European parliamentarians (better input legitimacy and accountability);
  • • increasing the transparency and publicity of the EU treaty-making process as well as introducing elements of deliberative democracy in this process (better throughput legitimacy);
  • • enhancing the problem-solving capacity of the European Union through better institutions and improved decision-making procedures (better output legitimacy).

At the same time, the Convention method cannot solve the fundamental legitimacy problem of a European Union of 25-plus Member States that deeply affects the daily lives of the citizens: the lack of a fully politicized and transnational public discourse on EU policies and politics as a pre-condition for a developed European democracy beyond the nation-state.

We proceed in three steps. First, we ask why we should bother at all about the legitimacy of European governance. Second, we clarify the concepts of legitimacy and accountability as applied to policy-making beyond the nation-state. Third, we compare the legitimacy of traditional intergovernmental conferences with the Convention method using the Constitutional Convention as our illustrative model. We conclude with some notes on the larger legitimacy deficit of the European Union.

I. Legitimacy in European Governance – Why Bother?

For almost 15 years, since the French and Danish referendums on the Maastricht treaties, the notion of a ‘democratic deficit’ has resounded through the European academic and political landscape. Yet, scholars as well as politicians could not differ more in their assessments. At one end of the spectrum, it is argued that the EU constitutes a federal polity in the making. According to this view, the EU's democratic deficit stems from an institutional setting that does not meet basic standards of governance common for liberal democracies. Consequently, the EU should improve the democratic quality of its decision-making processes through, for example, European-wide referendums, the direct election of the EU Commission president, the empowerment of the European Parliament and the like.2 At the other end of the spectrum, the EU is seen as an international organization that accommodates co-operation between its diverse Member States. In its extreme form, the argument of a democratic deficit claims that immoderate delegation as well as the craving for power of supranational bureaucracies has turned this organization into a technocratic super-state that short-circuits democratic processes at the national level. It thus runs the risk of overstretching the limited acquiescence of EU citizens. Accordingly, the solution is not to improve the democratic quality of EU policy-making, but to reduce the supranational powers of the EU.3

Yet, as Andrew Moravcsik reminds us, these contributions are generally based upon stylized ideal-types that are rarely met in reality. The debate, therefore, runs the risk of largely exaggerating the severity of the problem. In fact, one might even contend that the ‘near consensual criticism of European integration as democratically illegitimate is [. . .] unwarranted’ (Moravcsik, 2004). So why bother? While the debate on the ‘democratic deficit’ sometimes appears shrill and biased, there is nevertheless a widespread sentiment that the permissive consensus supporting European integration, almost no matter what, shows signs of erosion, which is mirrored in, for example, decreasing turnouts for elections to the European Parliament and protests during European summits. Moreover, the acquis communautaire has reached a degree of intrusion in the lives of European citizens including regulatory policies with distributive effects that we may ask whether the democratic quality of EU policy-making matches the severity of its decisions. Finally, we can observe an increasing alienation and disillusionment of citizens from the policy-making process, both at the national and at the European levels. ‘Europe’ and ‘globalization’ fears are frequently exploited by populist and simplistic arguments in national election campaigns. And as far as the EU is concerned, this problem is exacerbated by the well-known scapegoating behaviour of national policy-makers who frequently blame ‘Brussels’ for costly and unpopular decisions, while claiming the better accepted policies for themselves.

As a result, we have to ask the question how legitimate and accountable the EU decision-making process is, particularly if one is convinced – as are the authors of this article – that scaling down the supranational quality of EU integration is not an adequate solution to the challenges Europe faces in a globalized world. Indeed, the growing awareness of a legitimacy problem among European policy-makers resulted in the post-Nice process and the Laeken Declaration in 2001. The heads of states and governments declared that the Union needed to be brought ‘closer to its citizens’. What is more, intergovernmental methods, and the style and outcome of the 2000 Nice IGC in particular, seemed to have even aggravated the problem. Against this background, the Convention method that had just proven surprisingly successful in the elaboration of the Charter of Fundamental Rights appeared as a panacea to achieve better and more legitimate outcome. Before we evaluate whether the Constitutional Convention did improve the legitimacy of EU treaty-making, however, we need to clarify what we mean by legitimacy and accountability.

II. Legitimacy and Accountability in the European Union

Following Max Weber (Weber, 1980), a social or political order is legitimate if the members consider its rules both appropriate and binding. Legitimacy then generates voluntary compliance with costly rules because of a logic of appropriateness irrespective of either sanctioning mechanisms or the utility of the rule for those who have to comply. In this context, Fritz Scharpf and Michael Zürn have further distinguished between input, throughput and output legitimacy (Scharpf, 1999; Zürn, 1998, 2000). Input legitimacy refers to the probability that those being ruled have some say in the process of rule-making itself. Notions of input legitimacy that stem from models of representative democracy are generally based upon the assumption of a congruence between rulers and the ruled through mechanisms of representation and public contestation. Beyond the nation-state, however, power wielders must be held accountable in a way that institutionalizes crucial features of representative democracy without simply replicating them. Input legitimacy and accountability, thus, are closely linked concepts. According to Ruth Grant and Robert Keohane, accountability ‘implies that some actors have the right to hold other actors to a set of standards, to judge whether they have fulfilled their responsibilities in light of these standards and to impose sanctions if they determine that these responsibilities have not been met’ (Grant and Keohane, 2005, p. 29).

At first glance, the EU does not face problems of input legitimacy or accountability. The governments of the Member States that are represented in the European Council and in the Council of Ministers are all democratically elected. The European Commission is accountable to these democratically elected governments as well as to the equally democratically elected European Parliament. Thus, citizens who do not like what they get in terms of European decisions can sanction their representatives through voting them out of office. Yet, this is an idealized picture. The real problem of input legitimacy in the European Union is the combination of multi-level governance and supranationalism in the absence of strong transnational interest representation. As a result, citizens often have no idea who can be held accountable at what level of governance, particularly when their national governments can be overruled in the Council. This problem is exacerbated by the all too common ‘Brussels made me do it’ attitude of national governments mentioned above. One has to acknowledge, though, that EU treaty-making faces fewer problems of input legitimacy and accountability than everyday decision-making in the EU. First, Member States remain the ‘masters of the treaties’ as a result of which national procedures including referendums for the ratification of international treaties apply. Second, EU treaty-making requires the unanimity of the Member States so that the accountability problems of supranational decision-making are less severe.

Another aspect of legitimacy concerns the quality of the decision-making process itself, throughput legitimacy. Three components are of interest here. The first one is normally not disputed either at the EU or at the national levels, namely the legality of the process which is guaranteed by the rule of law in a democratic polity. The second component constitutes another ingredient of accountability, namely that it has to be crystal clear who is responsible for which decision at what level. In democratic polities, the transparency of the decision-making process and publicity shall ensure this ingredient of throughput legitimacy. In multi-level governance systems such as the EU, transparency of the process is often not ensured, no matter how many documents are being posted on the internet. EU treaty-making also suffers from a lack of transparency more often than not, given the tendency to overly secret procedures and negotiations behind closed doors. The third component of throughput legitimacy directly concerns the quality of the process. Proponents of deliberative democracy (Cohen and Sabel, 1997; Elster, 1998; Habermas, 1992) suggest that decision-making processes that systematically allow for arguing, reason-giving and mutual learning rather than hard-nosed bargaining will have a substantially improved chance of leading to better outcomes. The main reason is that arguing and reason-giving provide a mechanism to probe and challenge the normative validity of actors' interests as well as to check the empirical facts on which policy choices are based. Thus, deliberative quality constitutes a significant aspect of throughput legitimacy. There are few empirical findings available with regard to the deliberative quality of EU decision-making. As for EU treaty-making, intergovernmental conferences have usually been analysed in the framework of theoretical models which emphasize bargaining over arguing (Moravcsik, 1998). We will discuss below whether the Convention method has improved the deliberative quality of EU treaty-making.

This last component of throughput legitimacy is directly linked to the third aspect of legitimacy: output legitimacy. This concerns the effectiveness and efficiency of the policy-making process. Ultimately, political decisions are meant to improve the problem-solving capacity of a democratic polity. Politics and governance is about collectively regulating social issues and solving political and social problems. A political order that does not perform well will ultimately be considered illegitimate no matter how democratic the policy-making process. Scharpf has even argued that output legitimacy should be the sole criterion for evaluating EU policy-making, since the pre-political conditions for a democratic polity beyond the nation-state, such as a collective identity, cannot be taken for granted (Scharpf, 1999). This article cannot evaluate the problem-solving capacity of the EU in general. Rather, we concentrate on a comparison between intergovernmental conferences as the primary method of EU treaty-making, on the one hand and the Convention method as an innovative mode of EU governance, on the other. We now turn to an evaluation of the Constitutional Convention according to the legitimacy criteria outlined above.

III. The Convention Method: A More Legitimate Mode of EU Governance?

But why should legitimacy be a problem at all in the case of treaty revisions? As mentioned above, Intergovernmental Conferences (IGCs) comprise representatives of all Member States that unanimously agree on a treaty under international law. When it entails a shift of competences from the domestic to the international level, such a treaty is subject to the various ratification procedures of each Member State, that is, it requires approval by domestic parliament(s) and sometimes even national referendums. And in the case of an abuse of power, heads of state and government may be sanctioned by being voted out of office.

We maintain, however, that IGCs face various legitimacy problems. As to input legitimacy and accountability, the concerns are probably less severe than in cases of day-to-day EU decision-making, for the reasons stated above. Nevertheless and in spite of the fact that EU treaties have to be ratified through national procedures, the lines of authorization and accountability from ordinary citizens to their representatives at IGCs are unnecessarily long. More importantly, with regard to throughput legitimacy, IGCs face severe problems of transparency and publicity. Their deliberative qualities are at least questionable in the light of the horse-trading that went on at the Amsterdam and Nice IGCs. Finally, concerning output legitimacy, most observers agree that the Maastricht IGC has been the last one prior to the Constitutional Convention which produced outcomes such as Economic and Monetary Union or the Treaty on the European Union which enhanced the EU's problem-solving capacity. In contrast, the treaties of Amsterdam and Nice are widely regarded as rather modest accomplishments which failed to make the EU decision-making structures fit for enlargement and which produced many so-called ‘left-overs’ to be dealt with at subsequent IGCs.

How does the Convention method score as compared with traditional IGCs? At the Laeken summit in December 2001, the heads of state and government agreed on convening the ‘Convention on the Future of Europe’ as a forum preparing the 2004 IGC. It was composed of four groups of participants: representatives of national parliaments, of the European Parliament (EP), representatives of the Member States and of the European Commission. When the Convention started its work, it quickly decided to interpret its mandate as excessively as possible and to draft a Treaty for a European Constitution which the subsequent IGC only altered at the margins.

With regard to input legitimacy and accountability, the main novelty of the Convention as compared to traditional IGCs concerned the prominent role of members of the EP and of national parliaments. Since parliamentarians are directly elected by citizens, this considerably shortens the lines of accountability from the principals in a democratic polity – the citizens – to the agents in charge of the negotiations. Moreover, and probably more importantly, the participation of European and national parliamentarians vastly increased the plurality of interests represented in EU treaty-making. At traditional IGCs, interest representation takes place via national governments who – in theory – aggregate their respective domestic interests following a logic of ‘two level games’, from the national to the European level (Putnam, 1988). In contrast, the composition of the Convention allowed for a broader representation of interests, not just along national lines, but also along party and ideological lines.

With regard to throughput legitimacy, the Constitutional Convention strongly increased the transparency of the EU treaty-making process. The deliberations mostly took place in public and every single document was publicly accessible on the Convention's website. Thus, citizens were able to follow and to trace back the origins of almost every proposal for the Constitutional Treaty. The information available to citizens was enormous as compared to conventional IGCs. Publicity and transparency are pre-conditions for the ability of citizens to hold their representatives accountable for their decisions.

In addition, we submit that the institutional setting of the Convention served to improve the deliberative quality of the process.4 The Convention method put a premium on reason-giving and arguing as opposed to interest-based bargaining. First, the rules of the game, as laid out by Convention President Giscard d'Estaing at the very beginning, stipulated that Conventioneers were supposed to speak on their own behalf rather than on behalf of their nation-state, party, or whatever group. While such a rule does not preclude interest-based bargaining, it does require that speakers have to appeal to commonly accepted principles and norms in order to make their points. In that sense, it privileges arguing and reason-giving over bargaining. Second and perhaps most importantly, the Convention's decision rule was consensus (not unanimity, since it was never specified what precisely was meant by ‘consensus’). Given the plurality of interests and preferences represented at the Convention (see above), speakers could never be certain about the preferences of their audiences. They could not know for sure whether a particular proposal would meet the interests of a large majority of Conventioneers. Empirical studies of the scope conditions for arguing show, however, that reason-giving matters most in situations in which actors are either uncertain about their interests and preferences and/or speakers do not know the distribution of preferences among their audiences (Ulbert and Risse, 2005). Under such circumstances, speakers have to use persuasive arguments to convince their audience that a particular proposal is indeed in their best interest. The institutional setup of the Constitutional Convention ensured that speakers could rarely be certain about the preferences of their audience and/or which proposal would find majority support. Thus, the Convention's institutional setting privileged deliberation over bargaining. In contrast, conventional IGCs tend to institutionalize bargaining situations leading to the usual horse-trading and package deals (at least during the final rounds), since the preferences of the various governments are usually well-known in advance of the negotiations.

One could object to these claims that the rules of procedure granted the Convention chairman, Valéry Giscard D'Estaing and the Presidium (a sample of the Convention itself) considerable influence over the process and over substantive outcomes. It is certainly true that the Convention's particular outcome, a Constitutional Treaty that comprises as well as recasts the EU's constitutional law, owed a lot to Giscard's subtle steering of the deliberations and his particular use of the rules. However, it is fairly common in complex negotiating systems such as the Constitutional Convention that some functions are delegated to an actor or a group of actors in order to attain a mutually beneficial outcome. Moreover, there is little evidence to suggest that the chairman indeed succeeded in manipulating the Convention in a way so as to favour the pursuit of his own personal interests and preferences (Kleine and Magnette, 2006; for a detailed account of the negotiations see Norman, 2005).

As to output legitimacy, irrespective of the political controversy surrounding the Constitutional Treaty, there are several indications that the Convention indeed produced a better outcome than previous IGCs. First, for the first time since the early 1990s, there are no ‘left-overs’ to be dealt with in subsequent IGCs. Second, most observers agree that the Constitutional Treaty increases the efficiency of EU decision-making in a community of 27-plus members (through the provisions for a “double majority” and so on). Third, the Constitutional Treaty creates a single legal personality for the EU, an accomplishment that the Amsterdam and Nice IGCs failed to achieve. Fourth, the Constitutional Treaty improves the EU's ability to speak with one voice in foreign and security matters (EU foreign minister, EU diplomatic service etc.), even though it failed to introduce qualified majority voting in foreign and security policy.


In this article, we argued that the considerable effect decisions on the EU level have on the European citizen's daily life necessarily implies asking whether the treaty revision procedures of the European Union are legitimate. We tried to assess its legitimacy by comparing the Convention method with traditional IGCs. We claimed that the Convention method indeed scores better than conventional IGCs with regard to:

  • • input legitimacy (participation of members of parliaments leading to shorter lines of accountability and to increased plurality of represented interests);
  • • throughput legitimacy (better transparency and increased deliberative quality);
  • • and output legitimacy (no ‘left-overs’, greater efficiency of EU decision-making through institutional reforms).

Ultimately, however, legitimacy is about people's beliefs that a political order is indeed appropriate and worthy of voluntary compliance. One might ask, therefore, whether these rather abstract considerations of the legitimacy of the Convention matter much in the light of the rejection of the Constitutional Treaty by a majority of French and Dutch citizens which put the entire Constitutional project in jeopardy. Was the whole project superfluous and doomed to failure, as some have suggested (Moravcsik, 2006)? Would the legitimacy deficit disappear if elites just stopped talking about it in public? We doubt that and we suggest that the failure of ratification demonstrates a more fundamental problem of EU policy-making that even the Constitutional Convention – for all its merits – failed to address: the large gap between an elite consensus on European integration on the one hand, and the scepticism among many citizens on the other. This uneasiness about the European project cannot be overcome through renewed ad-hoc public relations efforts. The elite silence on Europe makes things worse and will ultimately backfire. At best, it contributes to the increasing alienation of citizens from the policy-making process, both at home and in Europe. At worst, it invites the exploitation of European issues by populist parties and policy-makers on the – right-wing or left-wing – fringes. Rather than keeping silent about the European project, policy-makers who care about European integration should actively contribute to the politicization of European issues. Given the significance of EU decisions for people's lives, European policies must become similarly subject to contestation and controversy in the transnational public sphere, just as domestic policies are in the national public spheres. As long as European elites including the European Commission are too afraid to debate Europe, they will exacerbate the current crisis. Public debates and political controversies are healthy in any polity and the EU is no exception. Moreover, one should not confuse politicization with populist politics. Populist parties are usually exploiting a lack of public debate and they have little to contribute to open and frank discussions in a lively public sphere. Of course, politicization might slow down European integration in response to globalization challenges quite a bit. But if this is the price to pay for increased legitimacy of European governance, so be it. Only if the Convention method is embedded in a lively transnational public sphere will it be able to translate its potential into increased legitimacy in the eyes of European citizens.


  • Research for this article was undertaken in the framework of the NEWGOV Integrated Project (‘New Modes of European Governance’) as part of the European Commission's 6th Framework Program for socio-economic research. We thank an anonymous reviewer for comments.

  • 1

    In 2000, a Convention drew up a Charter of Fundamental Rights. In 2002–03, the European Convention negotiated a Draft Constitutional Treaty. In the following, we will mainly focus on the latter.

  • 2

    See, for example, contributions by federalist policy-makers and MEPs to the debate. For academic contributions see Lodge (1994) and Williams (1991).

  • 3

    This line of reasoning is often voiced in the British debate. In more modest academic contributions, it is maintained that the EU's main task and source of legitimacy is the efficient fulfilment of its delegated functions (e.g. Majone, 2005). A democratic deficit either arises when governmental actors pursue their own agenda and do no longer represent their citizens' interests, or when their ability to pursue national interests is constrained in a way that it undermines legitimate governance. Fritz Scharpf argues that EU Member States are in an institutional configuration which privileges market regulation over social protection (Scharpf, 1999). Since this configuration is difficult to change, Member States find themselves trapped in a situation in which their ability to accommodate neoliberal policies with social protection at the national level is largely restricted – the resulting policy mix is suboptimal and illegitimate.

  • 4

    Case studies on the Convention's deliberations confirm this assertion. See Göler (2006).