EU law supremacy and national courts: who has the last word?

In 1962 an Italian lawyer named Flaminio Costa refused to pay an electricity bill of under two thousand lire. The sum was trivial. The case was not, because Costa argued that the law nationalising Italy's electricity industry breached the EEC Treaty. When the case reached the Court of Justice, the judges used it to state a principle that has structured European law ever since: the law that comes from the treaties cannot be overridden by domestic legislation.

The principle is called primacy or supremacy. It is a foundation of the EU legal order and one of its most argued-over, because it asks who has the last word in a legal system with more than one source of authority. This explainer outlines the doctrine, where national courts have pushed back and how it applies to Czechia. The Czech side is treated in more detail in Czech Constitutional Court and EU law, and the question of how EU acts arrive in national law in EU regulations versus directives.

The doctrine from the Court of Justice

The Court of Justice built the principle in stages.

Van Gend en Loos (1963) held that the Community was a new legal order of international law, for whose benefit states had limited their sovereign rights, and that EU law could create rights individuals could enforce before national courts. That is direct effect.

Costa v ENEL (1964) went further. The treaty had created its own legal system, which became an integral part of the legal systems of the member states, and states could not put a later unilateral national measure above a system they had accepted on a basis of reciprocity. If national law could override it, the Community's legal basis would be called into question.

Internationale Handelsgesellschaft (1970) added that an EU measure's validity could not be challenged on the basis of fundamental rights as formulated in a national constitution. The Court insisted that rights protection be found within EU law itself, and it has built that protection up over time, now codified in the Charter of Fundamental Rights.

Simmenthal (1978) drew the practical conclusion for judges: a national court applying EU law must apply it in full and set aside any conflicting national provision, without waiting for the legislature or the constitutional court to remove it.

The principle is not written into the main text of the treaties. The member states did record it in Declaration 17, annexed to the Lisbon Treaty, which notes that the treaties and EU law adopted on their basis have primacy over the law of member states, in accordance with the Court's case law. A declaration is not a treaty article, and that is one reason critics call the doctrine a creation of the Court rather than of the states.

Working with national courts

The Court of Justice does not decide national cases. It interprets EU law, and national courts apply that interpretation. The link is the preliminary reference procedure under Article 267 TFEU. A national court that has a question about the meaning or validity of EU law can refer it to the Court of Justice, and a court against whose decisions there is no judicial remedy is, in principle, obliged to do so, subject to exceptions set out in the CILFIT case of 1982, such as where the answer is clear.

Because of this design, primacy rarely looks like a confrontation. A national judge applies a rule, checks it against EU law and, if they conflict, applies EU law. The cases of resistance are few, but they show where the limits lie.

Where national courts have drawn lines

The most developed national case law comes from Germany. The Federal Constitutional Court said in the Solange I decision of 1974 that it would review EU measures against German fundamental rights as long as the Community lacked comparable protection, and in Solange II of 1986 that it would refrain from doing so as long as that protection was in place. In later rulings on the Maastricht and Lisbon treaties, it reserved the power to check whether EU acts exceed the competences transferred (an ultra vires review) and whether they respect the constitutional identity of Germany. In May 2020, in the case on the European Central Bank's bond-purchasing programme, it held for the first time that a ruling of the Court of Justice was itself outside the powers conferred, prompting a sharp response from the Court of Justice and the Commission.

Other courts have taken their own positions. The Danish Supreme Court in the Ajos case of 2016 accepted the principle in general but reserved the possibility of refusing to apply EU law that would conflict with Danish constitutional identity. In October 2021, the Polish Constitutional Tribunal declared parts of the EU treaties incompatible with the Polish constitution, a decision that the Commission and the Court of Justice rejected and that was widely seen as a rule-of-law problem, because the independence of that tribunal itself was contested.

The Court of Justice, for its part, has also given some ground on the question of fundamental rights. In the Melloni case of 2013 it held that a national court could not apply a higher national standard of rights protection where doing so would undermine the primacy, unity and effectiveness of EU law. But it also accepts that, where EU law leaves room for discretion, national standards may apply.

What it means in Czechia

Czechia accepted the EU legal order on accession in 2004, through a constitutional provision allowing certain powers to be transferred to an international organisation. The Constitutional Court accepts primacy in practice, while reserving the right to review whether the transfer has gone beyond what the constitution permits and whether EU law respects the essential features of a democratic state governed by the rule of law. It has used that reserve sparingly. In the widely discussed Slovak Pensions ruling (the Landtová case, judgment of 31 January 2012, Pl. ÚS 5/12) it found that a judgment of the Court of Justice was ultra vires. These cases are explained in Czech Constitutional Court and EU law, and the institution itself in the Czech Constitutional Court explained.

For ordinary courts in Czechia, the working rule is that, when a provision of Czech law cannot be read in line with directly effective EU law, the court should apply the EU rule. Administrative authorities are under the same duty. Lawyers who practise in areas such as consumer law, competition, data protection and public procurement raise EU law arguments as a matter of routine.

The debate going on

The argument is part legal, part political. Supporters of the Court's approach say the EU order works only if applied uniformly, since otherwise a state could escape its obligations by pointing to its own law. Critics say the final word on constitutional questions must rest with national courts, because the member states remain the "masters of the treaties" and the source of democratic legitimacy. Both sides agree that the system works best when courts talk to each other and avoid open confrontation.

To follow the case law, search the Court's database (curia.europa.eu) by case number. Many national constitutional courts publish decisions with English summaries.

Frequently asked questions

Where is the principle of primacy written down?

It comes from the case law of the Court of Justice, starting with Costa v ENEL in 1964, and is noted in Declaration 17 annexed to the Lisbon Treaty.

Can a national court refuse to apply EU law?

As a rule no. Some constitutional courts reserve a power to review whether the EU acted outside its powers or breached constitutional identity, but they use it rarely.

What is a preliminary reference?

A question from a national court to the Court of Justice about the interpretation or validity of EU law, under Article 267 TFEU.

Does primacy mean EU law overrides the constitution?

That is exactly the point of dispute. The Court of Justice says yes in matters within EU law. Many national courts say the constitution keeps a final reserve.