VoxEU Column EU institutions EU policies

Is a new treaty needed?

That the EU needs a new treaty is clear from two strands of evidence: the data showing a marked slowdown in EU decision-making since the May 2004 enlargement, and evidence based on ‘revealed preference’ reasoning.


The EU needs a new Treaty. We make the case based on two strands of evidence. The first is based on data – the slowdown in EU decision-making since the May 2004 enlargement. The second is based on ‘revealed preference’ reasoning concerning the men and women who are most in touch with the realities of EU decision-making – the leaders of EU member states. (This column summarizes the case we make in our recent CEPR Policy Insight.)

The facts: flow of EU lawmaking

A recent publication by a team of French scholars documents the sharp drop in the flow of EU legislation pre- and post-enlargement.1 As the top panel of the diagram shows, the May 2004 enlargement was accompanied by a sudden slump in the number of EU laws adopted. Of course enlargement is not the only suspect – 2004 saw a new European Parliament and Commission being installed. However, drop off was much larger in 2005 than the one that occurred after the previous Parliament-Commission switch-over (see bottom panel). Very recent figures from research published by two doctoral students show that the flow of legislation in 2006 recovered to something close to its pre-enlargement level.2 However this aggregate 2006 figure hides the fact that a significantly larger share of the legislation concerns areas where decision-making in the Council of Ministers is by majority voting rather than unanimity. Since the types of laws that are subject to majority voting – Single Market measures and the like – tend to be less controversial, this shift heightens the impression mentioned above that decision-making in the enlarged EU has had particular problems addressing controversial issues. Evidence from surveys of EU insiders (undertaken by the same doctoral students) confirms the general conclusion that reaching consensus has become much harder since the enlargement.

Revealed preference evidence

Counting the number of laws passed is a poor way to evaluate the impact of enlargement on the EU’s decision-making capacity. The basic problem is that there is no natural metric-stick for lawmaking. Some EU Directives have sweeping effect, while others are highly technocratic. The un-quantifiability of decision-making output does not mean that it cannot be gauged.

A few hundred people are directly involved in EU decision-making. Surely they know whether enlargement has hampered EU decision-making. But how do we get at this ‘private knowledge’? Merely asking them won’t work since insiders have an interest in strategically manipulating their unverifiable statements.

Economists get around such problems with ‘revealed preference’ reasoning. The idea is that the choices that people make in difficult situations reveals their true thoughts, a strategy we shall apply the question of Europe’s need for a new Treaty. The key is that almost all people who really know whether enlargement requires a reform of EU decision-making procedures report to Europe’s political leaders – the European Council – so the Council members must know the truth. We can find glimpses of the truth by looking at their choices in difficult situations.

In June 1993, EU12 leaders said that the Central and Eastern European nations would eventually become EU members. Everyone knew that EU institutions had to adapt and that this would require a Treaty to change the rules. In the EU, Treaty’s are drawn up in so-called InterGovernmental Conference, IGC for short. The European Council instructed the IGC held in 1996 to consider the institution reform linked to enlargement, namely Council of Ministers’ voting rules and the composition of the European Commission.

The IGC96 negotiations proved difficult. The IGC produced the Amsterdam Treaty, but failed to resolve institutional reform issues. In the end, EU members agreed to disagree on the institutional reforms by defining the “Amsterdam Leftovers”, namely Council of Ministers voting rules (vote allocation and areas subject to majority voting), and Commission composition (in particular whether every member should have a Commissioner). From the revealed preference perspective, however, the difficulty is useful. Hard-fought negotiations like the IGC96 act as a sort of ‘natural selection’ on agenda items. Very soon all participants realise which issues are urgent and obvious and which issues are ‘filler’ – things that address the concern du jour, or help buy the political support of wavering governments. Hard bargaining and thinking between 1993 and 1997 whittled down the list of ‘must do’ institutional reforms to the Amsterdam Leftovers.

The next attempt was in the IGC held in 2000 that lead to the Nice Treaty, the prime goal of which was to implement reforms that would maintain the Union’s democratic legitimacy and its ability to act in the face of enlargement, especially as concerns Council voting and Commission composition.

As we have documented extensively elsewhere, the Nice Treaty failed to adequately reform EU decision-making rules.3 This failure, however, was not something EU leaders could admit. As part of the final deal, the Council had to solemnly declare that the reforms in the Nice Treaty were sufficient. Then French President Chirac promised the European Parliament that the Nice reforms would be enough to allow the EU to function effectively and legitimately even after enlarging the club from 15 to 27. Importantly, the Nice Treaty reforms were not to come into effect until enlargement actually happened in May 2004.

The flaw in the Nice Treaty reforms – especially the Council voting rules – were not immediately apparent. And even when the flaw became widely recognised, EU leaders could not explicitly admit their failure. The small EU members, who sacrificed much power in the Nice Treaty deal, sold Nice to their parliaments as a painful but necessary reform – the price of Eastern enlargement.

A new tactic was needed. They set up a ‘convention’ with the 2001 Laeken Declaration and gave the conventioneers a long list of questions among which was hidden the issue of institutional reform. One of the 56 questions is: “how we can improve the efficiency of decision-making and the workings of the institutions in a Union of some thirty Member States?” That, of course, was supposed to be the job of the Nice Treaty.

This question is incredibly useful from a revealed-preference perspective. The negotiations that lead to the Nice Treaty reforms were extremely difficult – as any change in a power-sharing arrangement must be. Yet, in 2001, EU leaders asked the Convention to reform the Nice Treaty reforms even before the Nice Treaty reforms had been implemented. This shows us that the men and women best placed to know believed that the Nice Treaty reforms were a failure.

This change in tactics threw up a smokescreen that made it difficult for observers to accuse EU leaders of having bungled the Nice Treaty reforms.

The Convention, chaired by former French President Giscard d'Estaing, came up with a quite different set of voting rules for the Council of Ministers. Getting a law passed would require yes votes from 50% of member states that represented at least 60% of the population.

The Constitutional Treaty drafted under Giscard’s direction was fed into an IGC held in 2003 under the Italian Presidency. As in 1996 and 2000, the voting rules proved contentious and the draft presented by the Italian Presidency to the European Council in December 2003 was rejected. Importantly, during the Italian negotiations, member states were explicitly given the option of retaining the Nice voting rules instead of the new ones in the draft Constitutional Treaty. Although the same nations had agreed to the Nice rules unanimously in December 2000, they rejected them in 2003 – again before they had even been tried.

The final compromise in the Constitutional Treaty retained the Nice Treaty rules up to November 2009 (to assuage Poland and Spain who were to lose so much power under the Constitution) and it modified Giscard’s double majority scheme modestly. It was grudgingly accepted unanimously by EU25 leaders in June 2004, but rejected by the people – French and Dutch voters in particular.

Using the revealed preference perspective, the Constitutional Treaty’s reforms tell us that EU leaders believed that the Nice Treaty rules could not work. They could have simply refused Giscard’s double-majority scheme in the IGC2003, stuck with the Nice reforms, and taken the rest of the Constitution. They did not. They were given the same choice under the 2004 Irish Presidency and the outcome was the same.

Unfortunately, the Nice voting rules are the ones governing EU decision-making today. This fact cannot be changed without a new treaty. Hard bargaining since 1996 has pared down necessary institutional reforms to just two – reform of the Council voting rules and reform of the Commission’s composition. While the new treaty may touch on a variety of issues in order to provide a politically balanced package, there is little choice when it comes to institutional reform. It must fix the Council of Ministers voting rules.


1 R. Dehousse, F. Deloche-Gaudex and Dumamel, 2007, “Elargissement, Comment l’Europe s’adapte", Science Po, Paris.

2 S. Hagemann & Julia De Clerck-Sachsse, "Old rules, new game: Decision making in the council of ministers after the 2004 enlargement," CEPS Special Report, March 2007.

3 See R. Baldwin and M. Widgren, 2007, “Does the EU need a new Treaty?” CEPR Policy Insight, No. 2


315 Reads