European Union: gradual integration, centralisation and the limits of national sovereignty

Gradualism as method: from Schuman to “ever closer union”

The point of the gradualism thesis is not that every individual step must be harmful. The question is different: can small changes, each politically acceptable on its own, create over decades an institutional order that is fundamentally different from the starting point?

In the EU this can be tested quite directly. We can compare what common European institutions were empowered to do in 1951, after the Treaties of Rome, the Single European Act, Maastricht, Amsterdam, Nice and Lisbon — and how the space for national vetoes, national legislation, monetary policy and constitutional control changed along the way.

The May 1950 Schuman Declaration was both narrowly focused and long-term in ambition. It proposed a common High Authority for coal and steel, but also described concrete achievements as foundations for a European federation. The Treaty of Rome later referred to the foundations of an “ever closer union among the peoples of Europe”, language that survived in later Treaty law.

This matters because it corrects the idea that political integration was continuously concealed from the public. Different actors held different views of the end state, but the aspiration toward deeper integration was openly expressed in key documents. It is therefore more defensible to investigate the method and cumulative effect of integration than to claim that the direction itself was never stated.

From coal and steel to the common market

The European Coal and Steel Community was narrow in scope: it was not a European government and did not regulate most domestic policy. Yet it contained an important innovation — a common High Authority whose decisions were binding on participating states in the agreed field.

The first element of gradualism is already visible here. The initial step did not transfer general state power, but it established a practical precedent for exercising part of national decision-making through a shared supranational institution.

The 1957 Treaties of Rome expanded integration far beyond coal and steel. The European Economic Community pursued a common market, customs union and common policies in selected fields. The Treaty language also retained the goal of an “ever closer union”.

This was still not today’s EU. National states retained the overwhelming majority of their powers and the European Parliament did not yet possess its current legislative role. But the direction was already broader than the original sectoral arrangement.

From the national veto to Maastricht: institutional deepening

The Single European Act was central to completing the internal market and expanded the use of qualified-majority voting. Institutionally, this matters: under unanimity each state can block a decision; under qualified majority a state can be outvoted and still be bound by a validly adopted measure.

Supporters justified wider majority voting through efficiency and the need to prevent blockages in building the single market. Critics view the same mechanism as a concrete move away from intergovernmental cooperation, in which every state retains the final word, toward shared political decision-making.

The 1992 Maastricht Treaty created the European Union and marked a new stage of integration. It introduced EU citizenship, broadened cooperation and established the legal path and institutional framework for economic and monetary union, from which the euro later emerged.

A common simplification is to say that Maastricht “gave us the euro”. More precisely, Maastricht laid down the path, criteria and institutional structure for the single currency; the euro was introduced later. But the underlying point about a shift in authority is sound: monetary policy for euro-area states eventually moved from the national to the common level.

Amsterdam and Nice are less recognisable than Maastricht or Lisbon, but they continued the same institutional development: extending fields of joint decision-making, strengthening the European Parliament, adapting institutions to enlargement and integrating additional policy areas into the common framework.

This is precisely the type of process highlighted by the gradualism thesis: historical transformation does not require one dramatic rupture. Much institutional change comes through technical revisions whose full significance becomes visible only across a longer time horizon.

Referendums, the European Constitution and the Lisbon Treaty

The Treaty establishing a Constitution for Europe was signed in 2004 but required ratification by all Member States. French voters rejected it on 29 May 2005, and Dutch voters did the same on 1 June 2005. The ratification process stalled.

This event is central to arguments about the direction of integration, but it must be stated precisely. The result shows that a major proposal for deeper institutional reform lost referendum approval in two states. It does not by itself establish why every voter rejected it or that every individual reform contained in the text was separately rejected.

After a two-year “period of reflection”, governments agreed the Lisbon Treaty in 2007. EUR-Lex states that it was largely inspired by the rejected Constitutional Treaty and incorporated the majority of the institutional and policy reforms envisaged by it. The European Parliament similarly states that Lisbon preserved most of the constitutional project’s substantive achievements.

The two instruments were nevertheless not identical. Lisbon did not replace the founding Treaties with a single constitutional text, omitted certain state-like symbols and formal constitutional terminology, and included specific opt-outs and modifications. Saying that “the same constitution was simply renamed” is therefore too strong; saying that there was no major continuity is equally misleading.

This leaves a legitimate political question about method. Critics argue that substantial content rejected in constitutional form was later adopted in treaty form without new referendums in France and the Netherlands. Supporters respond that Lisbon was a different compromise ratified under the constitutional procedures of every Member State. Both positions can be set against the same documented chronology.

Ireland was the only country to hold a referendum on the Lisbon Treaty. It rejected the Treaty in June 2008. The European Council and Irish government then developed legal and political guarantees addressing concerns including taxation, military neutrality, family and social issues and retention of one Commissioner per Member State. A second referendum in October 2009 approved the Treaty with 67.1% voting in favour.

This too supports two interpretations. Critics see the repeat vote as pressure to obtain the desired outcome. Another interpretation emphasises that the political and legal context changed after additional guarantees and that the second referendum was itself a democratic vote. The article therefore documents the sequence without treating either side’s motive as established fact.

What becomes centralised and what does subsidiarity limit?

The cumulative effect can be measured without using the label “superstate”. Today the EU can adopt binding law on certain subjects; in shared competences, exercise of EU competence limits national room for action; in many fields the Council can decide by qualified majority; EU law has primacy under the doctrine of the Court of Justice; and monetary policy is common for euro-area states.

These are genuine shifts in political and legal autonomy. At the same time, extensive national competences remain, Treaty changes require Member-State ratification under their constitutional procedures, national security remains their responsibility, and Article 50 permits withdrawal. “Centralisation” therefore describes a direction in certain fields, not the complete abolition of statehood.

The Treaties themselves recognise the tension between common decision-making and keeping decisions close to citizens. Subsidiarity requires the EU, outside its exclusive competences, to act only when objectives cannot be sufficiently achieved by Member States and can be better achieved at Union level.

National parliaments have an early-warning or “yellow card” mechanism, but the threshold has been reached only rarely. This can be read in two ways: either most proposals do not generate sufficiently broad national objections, or the thresholds and time limits are too weak for effective political control. The documents themselves do not resolve the interpretation.

Democratic legitimacy and Slovenia’s decision

The EU has a dual democratic basis: citizens directly elect the European Parliament, while national governments participate in the Council and European Council. Member States ratified the Treaties under their constitutional procedures. This matters as a response to claims that European institutions operate without any democratic source.

But consent to the system is not the same as consent to every later decision. A state can be outvoted in the Council, a national parliament can lose the ability to change a rule unilaterally, and the European Commission has the central role in formal legislative initiative across most policy fields. This is where the question of democratic distance arises — a concern often emphasised more strongly in critical accounts than in standard institutional descriptions of the EU.

Slovenia’s experience shows both sides. Article 3a of the Constitution explicitly created the possibility of transferring the exercise of part of sovereign rights, while the referendum of 23 March 2003 approved EU membership with 89.64% of valid votes. Accession therefore did not occur without domestic political and constitutional legitimacy.

At the same time, the referendum was not a vote on every future regulation, judgment, extension of majority voting or later institutional interpretation. That is not unique to the EU — representative democracies do not hold referendums on every law — but it matters to the gradualism thesis because the long-term system always becomes broader than the original decision to join.

Irreversibility of integration and the question of a “superstate”

The claim that the system cannot be left is legally wrong. Article 50 expressly permits withdrawal, and the United Kingdom demonstrated in 2020 that it can be used. Treaty amendments still require ratification by Member States according to their constitutional rules.

Yet decades of integration create dependencies, shared legal regimes, supply chains, currency arrangements, regulatory standards and political costs of changing course. Legal reversibility is therefore not the same as practical ease. This is one of the strongest elements of the gradualism thesis: later choices are made in a different environment because earlier steps have already altered the cost of alternatives.

As a legal label it is too crude. The EU lacks the general competence of a state; Member States remain international legal persons, retain their own constitutions, armed forces and tax systems, and can withdraw. The Treaties formally limit the EU to conferred competences.

On the other hand, the EU has legal personality, a directly elected parliament, a court, a common currency for much of its membership, binding law, a budget and exclusive competences in selected fields. Political debate about federalisation or state-like characteristics is therefore not invented. A more precise description is a hybrid supranational system that has deepened over time, rather than a declaration in advance of its final constitutional status.

What remains after verification: cooperation, centralisation and the political boundary

The strongest part of the original thesis survives scrutiny: European integration has been gradual, each major Treaty altered the balance between national and common decision-making, and the cumulative result is a far more integrated legal-political system than existed in the 1950s.

It is also legitimate to raise the democratic question of whether citizens encounter deep institutional change mainly through individual Treaties, elections and indirect parliamentary legitimacy, while the overall direction becomes visible only across decades. The Constitutional Treaty, Lisbon and the Irish referendums are therefore important and should not be omitted.

The weaker part of the original thesis is the leap from documented direction to proven intention: the fact that integration proceeds gradually does not establish that every step is controlled by a single covert centre or that every argument for cooperation is merely a cover. The documents in fact show that federal and integrative ambitions were partly expressed in public from a very early stage.

The most useful question is therefore not “is the EU good or bad?”, but: what will today’s small transfer or common solution mean after ten more steps? The question is legitimate precisely because EU history shows that institutional weight really does change gradually.

The reverse question is equally legitimate: what does a state lose if it insists on solving genuinely cross-border problems alone? A common market, environmental standards, trade policy or currency can increase collective capacity. The price is that an individual state no longer retains complete unilateral autonomy in parts of those fields.

Documents can show direction, mechanisms and consequences. They cannot determine on behalf of citizens where the boundary should lie. After verification, the most important distinction is: integration is not identical to unity, and cooperation is not identical to centralisation — but the boundary between them has to be chosen consciously before a long-term process chooses it by default.