Investment Arbitration and State Aid Law: Here We Go Again

The relationship between EU State aid law and investment arbitration has become an increasingly important source of doctrinal uncertainty. What was once largely treated as a conflict between EU constitutional autonomy and international investment protection now raises new and difficult questions for the basic concepts of State aid law itself.

Several pending cases may soon give the Court of Justice of the European Union an opportunity to clarify this tension. The Commission’s Antin decision is currently being challenged before the General Court. Meanwhile, a German court has recently referred a number of questions to the European Court of Justice on the recognition and enforcement of arbitral awards and their compatibility with State aid law in RWE Renewables. And let us not forget Micula, which is still making its way through the courts and is now before the ECJ for the second time.

This blogpost sheds some light on a few questions raised in these cases. If you want to read an in-depth analysis, have a look at our recent paper, available open-access here.

Background: An Unusual Mix

Investment arbitration and EU State aid law are not often discussed in the same breath. One belongs to the world of international investment protection, where investors sue States before arbitral tribunals. The other belongs to the large world of EU competition law, where the European Commission controls advantages granted by Member States to undertakings. Yet, in recent years, these two legal worlds have collided.

The background is the long-running tension between EU law and intra-EU investment arbitration. After the Court of Justice decided Achmea, Komstroy, and PL Holdings, intra-EU investor-State arbitration was treated as incompatible with EU law because it threatens the autonomy of the EU legal order. In simple terms, the Court did not want disputes involving EU law to be decided by arbitral tribunals outside the EU judicial system.

But the problem did not disappear. Many investors had already brought claims under investment treaties, and some had already won large awards. This is where the current cases come in. In Antin, for example, the case concerned investors in Spanish solar energy projects. Spain had introduced a renewable energy support scheme in 2007 but later changed and repealed parts of that regime. The investors argued that Spain had breached its obligations under the Energy Charter Treaty, especially the fair and equitable treatment standard. An ICSID tribunal agreed and awarded the investors around EUR 101 million plus interest. The European Commission then stepped in and declared that the award, or at least its implementation, constituted unlawful State aid incompatible with the internal market. Spain was told not to pay and to prevent recognition or enforcement of the award, even outside the EU. This decision is now challenged in court.

Stretching State Aid Notions

But can these actions amount to unlawful State aid? The question matters because the answer could change the way arbitral awards against EU Member States are treated. If an award, or its payment, is State aid, then the Commission can indeed prohibit payment, require recovery, and ask national courts to prevent enforcement. That turns State aid law into a powerful weapon against investment arbitration. This move is far from straightforward and could stretch State aid law close to its conceptual breaking point.

First off, everything depends on what the State aid “measure” actually is. State aid law always requires a measure to be assessed. Is the measure the arbitral award itself? Is it a Member State’s payment of the award? Is it the recognition or enforcement of the award by a court? Or is it a country’s earlier decision to sign and ratify treaties such as the Energy Charter Treaty and the ICSID Convention?

The Commission’s decision in Antin, for example, seems to move between these options, speaking of the award “and/or” its implementation. This is a serious problem. A legal analysis cannot shift its object depending on which State aid criterion is being discussed. The exact measure matters for imputability, advantage, and enforcement.

The second key issue is imputability. In State aid law, a measure must be attributable to the State. If a country voluntarily pays an award, that payment is clearly State conduct. But what about the arbitral award itself? This is much more difficult. An arbitral tribunal is not a Member State. A foreign court recognising the award is not a Member State either. If a US, UK, or Australian court enforces an ICSID award under its own legal system, it is hard to say that the resulting act is attributable to a Member State.

The Commission’s reasoning in Antin, for example, depends heavily on the idea that Spain originally consented to arbitration by entering the relevant treaties. But does State aid law prohibit “original sins”? Does the fact that Spain once accepted the treaty framework mean that every later arbitral or enforcement step should be treated as flowing from, and being tainted by, that original decision?

This distinction is important because not every event in the life of an award involves the same degree of State action. Treaty ratification is State action. Voluntary payment is State action. But an arbitral tribunal’s decision, judicial recognition, and coercive execution against assets are different. These are adjudicative or enforcement acts, often carried out by independent courts or authorities. The Commission’s imputability analysis works only if one blurs these different stages together. Once they are separated, the reasoning becomes much weaker.

The third major issue is advantage. State aid exists only if an undertaking receives an economic benefit it would not normally have obtained. Can an award be treated as an advantage because, the arbitral route was incompatible with EU law and the compensation should not have been granted?

This seems to be a shortcut. Compensation for loss is not normally an advantage. If a State commits a wrong and pays damages, the injured party is not necessarily being enriched; it is being restored to the position it would have been in without the wrong. One has to draw a line between compensation and a State aid advantage. A State aid advantage makes the recipient better off. Compensation makes the recipient whole.

Of course, things become more complicated if the compensation merely restores an earlier unlawful aid advantage. That was the logic in Micula, where the award compensated investors for the withdrawal of incentives that had been linked to unlawful aid. But in Antin, for example, the situation is not the same. Spain’s 2007 renewable energy scheme had not previously been fully and properly classified as unlawful aid. The Commission’s later attempt to characterise it as such appears underdeveloped and possibly reverse-engineered to support the conclusion that the award itself was aid.

Here, the Commission seems to move from the conclusion backwards. It wants the award to be State aid, so it treats the underlying Spanish renewable energy scheme as unlawful aid. But if that underlying characterisation is weak, then the whole advantage analysis becomes unstable. The Commission still has to do the normal State aid work: identify the advantage, define the relevant benchmark, and show that the recipient is better off than it would be under normal market conditions. The mere fact that the award comes from an arbitral process disliked by EU law does not automatically make it State aid.

The Enforcement Dilemma and National Courts

The final and perhaps most practical issue is enforcement. The Commission decision in Antin, for example, does not merely say that Spain should not pay. It also requires Spain to prevent recognition, enforcement, or execution of the award, including in third countries. And here seems to be a major mismatch between EU State aid law and the ICSID enforcement system. ICSID awards are designed to travel. Courts in contracting States are generally expected to recognise and enforce them as if they were final domestic judgments. EU (State aid) law may tell Spain not to pay, but it cannot simply command US, UK, or Australian courts to stop applying their own enforcement rules.

This creates a real-world problem. The Commission’s approach is driven by effectiveness: if unlawful aid must not be paid, then Member States should prevent enforcement. But this logic becomes unstable once it reaches beyond the EU. Article 108(3) TFEU may justify restraining a Member State’s own conduct. It may also justify interim measures in national courts where there is a strong EU connection. But it does not easily become a global anti-enforcement tool capable of stopping foreign ICSID proceedings. The more the Commission tries to project EU State aid law outward, the more it exposes the limits of EU legal authority.

What to Make of it All?

Yes, EU law has rejected intra-EU arbitration. Yes, the Commission has a strong interest in preventing unlawful aid. But those points do not mean that every investment award, every recognition judgment, or every enforcement step can automatically be treated as State aid. Legal categories still matter. Imputability must be shown. Advantage must be proven. The measure must be defined consistently. Enforcement powers must have limits.

Decisions, such as Antin, are innovative but fragile. They are innovative because they use State aid law to respond to the continuing problem of intra-EU investment arbitration as an overall constitutional problem. They are fragile because it asks State aid law to do more than it may be conceptually able to do: managing the afterlife of investment arbitration awards. That may be understandable from the perspective of EU constitutional autonomy, but it creates serious doctrinal and practical problems.

Investment arbitration awards should not be treated as State aid simply because they are inconvenient for EU law. Some payments may raise State aid concerns, especially where they reinstate an unlawful aid advantage. But the analysis must be careful, act-specific, and grounded in the ordinary requirements of Article 107(1) TFEU. Otherwise, State aid law risks becoming a general anti-arbitration instrument rather than a disciplined tool for controlling distortive economic advantages.

For now, much depends on whether the Court of Justice will clarify where the line should be drawn between protecting the autonomy of EU law and stretching State aid rules beyond their proper limits.