Thursday 5 June 2014

Double Jeopardy and EU Law: Court clarifications and Commission reports

Steve Peers

It’s just like buses. You wait almost four years for a CJEU judgment on the Schengen double jeopardy rules, and then two judgments come along almost at once. What’s more, the Commission has also just released two reports on the application of related EU legislation. The judgments have provided a number of useful clarifications of the rules, which the reports paint the familiar disappointing story of tardy implementation of Member States’ legal obligations.


Double jeopardy (or ne bis in idem) in criminal cases, ie trying the same person for the same crime more than once, is banned by many national constitutions and by the Seventh Protocol to the European Convention on Human Rights (although a few Member States have not ratified this Protocol).  However, these rules generally only apply the rule to trials within the same State.

EU law goes further than this, by establishing a cross-border rule. Article 54 of the Schengen Convention, which applies in all Member States except Ireland, as well as the Schengen associates (Norway, Iceland, Switzerland and Liechtenstein), states that ‘A person whose trial has been finally disposed of in one Contracting Party may not be prosecuted in another Contracting Party for the same acts provided that, if a penalty has been imposed, it has been enforced, is actually in the process of being enforced or can no longer be enforced under the laws of the sentencing Contracting Party.’

Article 55 of the Convention allows for a limited number of exceptions to this rule, and Article 56 provides that if there is a second prosecution, any period of deprivation of liberty arising from the same acts in the first Member State must be deducted from any sentence imposed in the second Member State. Article 57 sets up a process of consultation between Member States to implement the rules, while Article 58 allows Member States to apply more generous rules between themselves.

Despite the current limits on the CJEU’s jurisdiction over EU criminal law measures adopted before the Treaty of Lisbon, there have been a considerable number of judgments on these rules over the years, since about two-thirds of Member States have opted into the Court’s jurisdiction. We might expect more cases being sent to the CJEU on these issues after December 1 this year, when those limits will be removed. Cases could even be sent from the UK courts, since the UK intends to opt back in to (among other things) these provisions of the Schengen Convention when it opts out of a lot of pre-Lisbon EU criminal law measures on that date.

EU law does not generally regulate the application of the double jeopardy principle within the same Member State, although it might be relevant where the proceedings concern a substantive criminal law issue linked to EU law (see, for instance, the Fransson judgment).

While national, international and EU law on substantive criminal law often provides for extraterritorial jurisdiction, an investigation into the same person for the same acts in multiple Member States is in prlnciple a waste of time and money, since the person concerned cannot usually be tried twice in both Member States. So back in 2009, the EU adopted a Framework Decision on conflicts of jurisdiction, setting up a process of information and consultation between national authorities if there was an indication that there multiple proceedings concerning the same acts committed by the same person.

On the other hand, if the same person has committed different crimes, national criminal law systems usually impose some form of added penalty, in the form (for instance) of harsher bail conditions or a longer sentence in the event of a conviction. It follows from the cross-border nature of EU Justice and Home Affairs cooperation that this principle should apply even if a prior conviction was handed down by another Member State. So another Framework Decision, adopted in 2008, sets out the details of how this rule applies.

The judgments

The first recent judgment, in Spasic, raised two points. First of all, is the rule limiting the double jeopardy rule to cases where the penalty (if there is one) has been enforced, can no longer be enforced or is being enforced (the ‘execution condition’), valid in light of the EU Charter of Rights? Secondly, how does that condition apply where a criminal conviction has resulted in both a custodial penalty and a fine, where the fine has been paid but the custodial penalty has not been served?

These questions were relevant in this case because Mr. Spasic had paid a fine of 800 euros imposed as a sentence by an Italian court for passing counterfeit currency, but had not served the one-year sentence which was imposed for the same acts. So could he still be prosecuted in Germany for them, given that the double jeopardy right in Article 50 of the Charter does not mention any execution condition?

In the Court’s view, the execution condition was valid in light of Article 52(1) of the Charter, which sets out the rules for limiting Charter rights. Limitations must be provided for by law, respect the essence of the rights and freedoms and be necessary and proportionate in pursuit of an objective of general interest.

Here, the limitation was clearly provided for by (EU) law. It respected the essence of the right, since it only limited the double jeopardy rule where the person concerned had escaped punishment. It aimed at securing objectives of general interest, since the Treaty rules on JHA objectives implicitly aimed to avoid situations of impunity. 

As for proportionality, while there were other relevant EU measures, the Framework Decision on conflicts of jurisdiction did not ensure that the person would be punished, and the Framework Decision on transfer of prisoners was subject to the conditions of consent of the person concerned and of the sentencing State. However, the CJEU accepted that it would not be proportionate to bring a second set of proceedings in Germany, if Italy wanted to enforce its sentence (Mr. Spasic had not served that sentence simply because he was imprisoned in Austria for different offences at the time).

On the second point, the Court ruled that custodial and non-custodial penalties were severable for the purpose of applying the execution condition. So paying an 800-euro fine did not equate to partial satisfaction of a one-year jail term, and so did not exempt the person concerned from being prosecuted in a second Member State.

The second case, M, concerned allegations of child abuse against an Italian man living in Belgium. The Belgian authorities investigated these, but a pre-trial chamber gave a ‘non-lieu’ ruling, meaning that the case could not proceed to trial due to insufficient evidence. This ruling was eventually upheld on appeal. In the meantime, the Italian courts had begun their own investigation, and wanted to know if the Belgian ruling was a ‘final judgment’ which prevented them from proceeding to trial.

In the Court’s view, it was. A final judgment could include a pre-trial measure (like the plea bargain accepted by prosecutors in the earlier Gozutok and Brugge case), as long as it finally determined the case on the merits according to national law (applying the test in Turansky). That was the case here, once the appeal was decided. Referring to the Zolukhtunin judgment of the European Court of Human Rights, the CJEU ruled that a possibility of an extraordinary remedy did not preclude a trial from being final. Nor did the possibility, referred to in the ECHR Protocol and Belgian law, of bringing a fresh prosecution, as a derogation from the double jeopardy rule, in the event of new or newly discovered evidence. The Court confirmed that in that case, only the first Member State, not the second, could bring a fresh prosecution.

The reports

For the Framework Decision on conflicts of jurisdiction, the Commission report notes first of all that only about half of the Member States have implemented it (the deadline was June 2012). This makes sense for the UK, since it does not seek to opt back in to this measure. It should be noted that infringement proceedings against Member States will also be possible when the 1 December deadline soon expires.

Overall, while the Commission refers to this measure as a ‘first step’, it states that it cannot draw ‘general conclusions’ about the quality of implementation, because few Member States have practical experience in its implementation. But it does note that some Member States have not applied the key rules on the information to be transferred between authorities. Most Member States permit parallel investigations to take place, even after the consultation process, although they are a waste of time and money. Only Croatia has an express rule (corresponding to the M judgment) that a final judgment in one Member State must terminate a pending prosecution in another.

As for the report on the Framework Decision on the consequences of convictions, six Member States have still not applied it (the deadline was back in 2010). This time, the UK, which seeks to opt back in to this measure, has applied it.

The Commission is rightly concerned that Member States limit the mutual recognition obligation to cases of a final conviction, for otherwise they would be infringing the presumption of innocence. Overall, the Commission expresses dissatisfaction mainly with a group of nine Member States which have claimed to implement the Framework Decision without giving further detail. Obviously it would be wise for those Member States to provide that further detail over the next six months, before infringement actions can be launched. The Commission does not assess the practical application of the European Criminal Records Information system (ECRIS), established by separate EU legislation, which is the main way of ensuring that the relevant information is transferred between Member States in the first place.


The Court’s ruling that the ‘execution condition’ in the Schengen rules was valid is convincing in light of the great public interest in preventing impunity for those convicted of criminal offences. For the same reasons the Court was right to reject the argument that payment of an 800-euro fine was equivalent to serving a year in jail. That would have amounted to de facto impunity.

Crucially, the Court rightly applied the principle of proportionality to rule that on the facts of the case, a second prosecution could nonetheless not take place if it was still possible in practice to serve the initial sentence instead. This rule ensures that the person concerned does not end up serving two separate sentences for the same crime (recalling that any time served in one Member State would have to be deducted from the time served in another one), and could prevent the cost and waste of time of a second prosecution if it proves to be unnecessary after all.

It should be noted that the Court did not rule here on whether the national derogations permitted by Article 55 of the Schengen Convention are valid. Unlike the execution condition, these rules do not aim to avoid impunity, but rather allow a second prosecution to be brought where: the act took place on a Member State’s territory; the act concerned national security or ‘equally essential’ interests; or the person concerned was an official of the State concerned. In all these cases, justice would best be served by giving the Member State concerned the first opportunity to bring a prosecution, and the Framework Decision on conflicts of jurisdiction should facilitate that. But it may be questioned whether those interests are strong enough to justify the power to bring a second prosecution (although again it must be recognised that any time served in the first Member State must be deducted from any sentence in the second).  

As for the M case, it answers a number of important issues. It is now clear that the double jeopardy rule is triggered only on disposition of a final appeal in the first Member State, and (implicitly) that the rule not only prevents the start of proceedings in a second Member State, but also terminates proceedings that are already underway. Again, the relevance of the Framework Decision on conflicts of jurisdiction becomes clear: whichever of the Member States concerned lost the ‘race’ to a final judgment (stereotypically, this was Italy) would end up having wasted its time and money.

The clarification of what happens in the event of new evidence is also important. It should be noted that, as the Advocate-General pointed out, if new evidence surfaces in Italy (or any other Member State), EU rules can facilitate its transfer to Belgium for the authorities to consider it. Also, the Court makes a general statement that only a ruling on the ‘merits’ counts as a final judgment. This sits oddly with its judgment in Gasparini, where it ruled that a proceeding which was time-barred in one Member State counted as a final judgment, so that case must be seen now as an anomaly.

More broadly, the two new judgments go a long way to reconcile the Schengen double jeopardy rules with human rights law. Oddly, the Court had never before mentioned Article 50 of the Charter in the context of the Schengen rules, until the Spasic case forced it to. Afterward, in the M case, the Court now refers to interpreting the Schengen rules in light of the Charter. As for the ECHR, the CJEU positively dismissed its relevance back in the Van Esbroeck judgment, but since the Strasbourg Court aligned itself with the CJEU's interpretation of the principle in Zolukhtunin, the CJEU is now happy to rely upon the other Court's rulings as a source of interpretation.   

What about the EU legislation? The Court itself, in the Spasic case, said in effect that the Framework Decision on conflicts of jurisdiction was too weak to affect its interpretation of the double jeopardy rules, since it did not ensure that the person concerned would be punished (although actually, that isn’t its main purpose). The M case shows, as noted above, how an early decision on conflicts of jurisdiction is necessary to avoid wasting time and money, and to reduce complications for suspects, particularly given that (in the view of the Belgian courts) there was insufficient evidence to prosecute in that case.

Overall, there seems to be plenty of reason to fear that, as I predicted some time ago, the Framework Decision on conflicts of jurisdiction has done nothing much to avoid multiple prosecutions, in particular because it lacks an obligation in principle to centralise prosecutions, never mind a list of criteria to apply to determine where the prosecution should take place. It has obviously had no effect in the half of Member States which have not bothered to transpose it at all. Since the Commission regards this measure as a ‘first step’, it should have considered measures to amend it to this end at the earliest opportunity.

As for the Framework Decision on the consequences of prior convictions, the Commission assesses its implementation on paper, but not its implementation on the ground. The real question is how it works in conjunction with the ECRIS system, and in particular whether there are any problems as regards the classification of convictions rendered in other Member States.

These reports are released as justice ministries begin their last lazy summer before the prospect of infringement proceedings to make them to live up to their legal obligations (as we have seen previously on this blog, these failures extend also to legislation on prisoners and probation, and to hate crime law too). Time will soon tell whether the Commission will step up to the plate and assume its role as the guardian of EU law in this field.

Barnard & Peers: chapter 9, chapter 25

No comments:

Post a Comment