This piece was originally posted on the Verfassungsblog and is reposted here with thanks.
The Czech Constitutional Court’s recent decision to declare a previous CJEU judgment ultra vires has raised considerable concern in the EU legal community. Jan Komarek on this blog accused the CCC of “playing with matches”. Arthur Dyevre analyzes the situation from a game theory and international relations perspective:
On the face of things, the CCC’s judgment, by declaring an EU act, namely a ruling of the Court of Justice, ultra vires, constitutes a momentous and unprecedented display of judicial defiance. To my knowledge, no domestic court has ever taken this step before in a final judgment on the merits; and certainly not in so explicit a manner. In recounting the background of the case and the sequence of events that led to the CCC’s decision, Jan Komarek points to a number of intriguing aspects of the case. One, which has already been highlighted in connection with the Melki case, is the difficulty for the CJEU to avoid alienating domestic judges when it is dragged into domestic judicial politics by way of the preliminary ruling mechanism. Here, however, my primary interest is in the significance and implications of the CCC’s decision for the EU multi-level legal system. In a non-hierarchical court system, where courts at the upper echelon do not have the power to strike down the decisions of courts at the lower level, judicial cooperation appears to be essential to the effectiveness of the higher-level law. So, by defying the authority of the Court of Justice in such blatant fashion, the CCC’s judgment may be viewed as striking a terrible blow to the authority of EU law. Doomsayers may see it as the first event in a chain reaction that will ultimately bring down the whole constitutional edifice of EU law. From now on, every domestic judge will assume that she can safely ignore EU law whenever she sees fit. Still, while there is no gainsaying that judicial defiance at domestic level may potentially raise major problems, I would nonetheless suggest, borrowing insights from game theory and international relations, that this judgment is more likely to remain an isolated event. An all-out war with the CJEU is not in the long-term strategic interest of any domestic court. Hence there is a fairly good chance that, one way or another, the CCC will soon come to its senses and will repudiate a decision that seems to be driven by anger rather than by reason. If it wants to remain a player in the multi-level judicial game, the CCC should take a closer look at the German Federal Constitutional Court (GFCC), which has so far proved a more thoughtful strategic player in its relations with the CJEU.
Of Hawks, Doves and Chicken
Of the classical game forms found in the game-theoretic literature, the one that seems to best approximate the conflictual relations between the CJEU and national courts such as the GFCC and the CCC is the game of Chicken. Most people, presumably, know it from the classical presentation where two drivers head for the same single lane bridge from opposite ends, with the last to swerve counting as the winner. The same game, with the same payoff structure, is also known as the Hawk-Dove game. There two players compete over a resource and have to decide whether to fight (play Hawk) or to acquiesce (play Dove). For each player, the best outcome is, of course, when she plays Hawk and the other player plays Dove, while the worst outcome sees both players choosing to play Hawk. This game form has been widely used in international relations to analyse crisis situations. What strategy should a country adopt, knowing that it will incur severe losses if it acquiesces to the demands of its neighbour but knowing at the same time that a war would be even more costly? I see several reasons why this approach also constitutes an appropriate way to model judicial interactions in the EU legal system. First, jurisdiction is a rivalrous good. Assuming that courts want to expand – or at least preserve – their jurisdiction, courts on opposite sides of a jurisdictional dispute, like the two players in a Hawk-Dove game, have opposite institutional interests. When one court expands its jurisdiction, it normally does so at the other’s expense. (More powerful domestic courts have obviously more to lose, which would explain why the mighty GFCC has been at the forefront of the judicial resistance to integration.)
Second, analogous to the players in a Hawk-Dove game, the courts perceive, or should perceive, the disastrous consequences that would ensue, should both pursue a strategy of defiance. For the CJEU, a single case of overt non-compliance by an influential domestic court may set a dangerous precedent, damaging its authority as well as the effectiveness of EU law. But putting a threat of non-compliance to execution may attract problems to domestic courts, too. A ruling that comes to be regarded as detrimental to the country’s interests and membership in the EU may trigger adverse political reactions. Legislators may decide to punish the unruly court by rolling back its jurisdiction, changing its rules of procedure, appointing new judges, etc. On that score, it is worth remembering that a group of respected German academic lawyers reacted to the GFCC’s ruling on the Lisbon Treaty – which stopped short of holding the Treaty unconstitutional but was nonetheless regarded as articulating a strongly Eurosceptic position – by calling on legislators to amend the Federal Constitutional Court Act (Bundesverfassunggerichtsgesetz). The proposed amendment would have required that the GFCC send a reference for a preliminary to the Court of Justice before entering any judgment on the ultra vires character of an EU act. Had it become law, the amendment would certainly have dealt a severe blow to the GFCC’s institutional standing. More generally, despite the growing anti-EU sentiment among their voters, government parties in the Member States, even the more Eurosceptic ones, usually agree about the fact of EU membership. This entails that domestic courts can ill-afford to make decisions that would imperil their country’s full membership in the supranational club. In my view, this fact places an upper limit on the level of defiance of domestic judges. The doctrines of direct effect and supremacy are now part – though not necessarily in the form expounded by the CJEU in its jurisprudence – of the “acquis communautaire”. Thus, unless the government parties wish to leave the EU, a court that blatantly defies it will face a political backlash. This is why the decision of the Czech constitutional judges looks daft. As the GFCC understood early on, going to war should always be the ultima ratio.
Bluff and Brinkmanship in the Judicial Cold War
In the situation we are concerned with, the courts’ strategic choice mirrors the one faced by the players in a Hawk-Dove game. Assuming that a constitutional crisis is a worse outcome than a jurisdictional loss, a domestic court’s best response to a dovish CJEU is hawkish judicial expansion or reassertion, but its best response to a hawkish Court of Justice is judicial restraint. Hence it is easy to see to that each court would ideally be the Hawk and have the other be the Dove. Yet strategic decision-making – again, as in the standard formulation of the Hawk-Dove game – is rendered difficult by the fact that the courts do not have a dominant strategy – i.e. a strategy that remains the best whatever strategy the other court happens to choose. This difficulty is further compounded by the iterative character of judicial interactions in the EU court system. We’re not dealing with a one-shot game but with repeated interactions.
Now, when we repeat a game indefinitely many different equilibriums are possible. But if the players are allowed to communicate about their future choices, they may be able to use communication so as to induce an equilibrium more favourable to their interests. A country’s leader may thus want to signal hawkishness and announce he is ready to go to war. Similarly, a court may hint that it is ready to risk a constitutional crisis to force the other to acquiesce to its jurisdictional demands. The resolve of the judges just as that of the leader may be impossible to establish with certainty. But credible enough the signal may well work. This I would argue is the game the GFCC has been playing ever since its first Solange decision. On numerous occasions, the German Court threatened to disapply EU legislation if found to be ultra-vires or to violate basic human rights. Yet it has never put its threat to execution. To many legal scholars, this is proof that the GFCC is a dog that barks but never bites. But in fact this may be a sign of its success in countering the CJEU’s activist impulses. Without ever setting aside a single EU act, the Karlsruhe court may have managed to set limits on the European Court’s jurisdictional expansion. Perhaps it was bluff all along (the judicial Hawk was in reality a judicial Dove). But if bluff it was, it seems to have worked, at least some of the time. The Court of Justice’s human rights jurisprudence is often presented as a response to Solange. The GFCC’s decision on the Lisbon treaty, meanwhile, invites comparison with the Cuban missile crisis. By designating its most Eurosceptic judge, Udo di Fabio, as rapporteur and by issuing an opinion with strong sovereigntist overtones, the GFCC may have successfully emulated President Kennedy’s cautious firmness with the Soviet Union. Having made its voice heard, the Honeywell ruling was then similar to the Kennedy’s decision to withdraw nuclear warheads from Turkey: a face-saving exit for the CJEU that would ease and bring tensions back to a more manageable level.
I do not mean to say that the GFCC is always a force for good in the EU legal system. Its role in the ongoing debt crisis, where it seems to serve as pre-commitment device for the German government in negotiations with other Member States (“We can’t accept this because our constitutional court will say no”), is, for my money, highly objectionable on normative grounds. So too is the declaration of its President, Andreas Vosskuhle, that the German Constitution “hardly admits of more integration”, whose subtext seems to be “We, the Court do not want to see more integration”. But this is not my point. Rather my point is that thoughtful domestic judges, even if they take a sceptic view of integration, should first seek negotiations with the judges in Luxembourg before even thinking about pressing the big red button. Those who fear a judicial Armageddon will find some comfort in experiments that have shown that the iterated Hawk-Dove game (in its Snowdrift variant) leads to consistently higher levels of human cooperation than other iterated games such as the iterated Prisoner’s Dilemma. As with superpowers during the Cold War, the threat of mutually assured destruction seems to provide human beings with a strong incentive to cooperate. Let’s hope judges are human beings too.
Arthur Dyevre is Senior Research Fellow at the Max Planck Institute for Comparative and International Public Law in Heidelberg.