Removing persons from a community because of a crime they committed is a common phenomenon in law. In medieval England, sources show that men ‘of particular ill-repute’ or presented for serious crimes were forced to ‘abjure the realm’, sometimes even if they passed the procedure of an ordeal (J Hudson, The Formation of the Common Law, Longman, London 1996, 177). More modern international law renders such a mixture between an immigration measure and criminal punishment somewhat more complex: A state can no longer simply expel its own nationals. Still, for foreigners the question continues to arise whether a crime they committed should exclusively be tackled with the tools of criminal law or whether that crime should be seen as a rupture of the bonds of integration between the foreigner and society, resulting in the foreigner’s expulsion.
This thorny question is raised by the case of P.I. Mr I has lived in Germany since 1987. From 1990 onwards, he comitted acts of sexual coercion, sexual assault and rape on his former partner’s daughter who was 8 years old when the offences began. His acts were only discovered later, because he continuously threatened and isolated his victim. In 2006, he was eventually convicted to a term of imprisonment of seven years. By a decision of 2008, Mr I was ordered to leave the territory and lost the right to enter and reside in Germany.
In the appeal to this decision and the subsequent preliminary reference to the CJEU, the question arose whether the long period of residence should prevent an expulsion or whether the nature and context of the crime Mr I committed called for a different solution. Directive 2004/38 on the right of EU citizens to move and reside freely creates a system substantially based on an ever ‘greater degree of protection against expulsion’, the ‘greater the degree of integration of Union citizens’ becomes (recital 24). For Union citizens who have resided for ‘many years in the territory of the host Member State’, an expulsion measure should only be taken ‘where there are imperative grounds of public security’ (ibid.). Putting these objectives into practice, Article 28 of the Directive requires in its first paragraph that before taking an expulsion decision based on ‘public policy or public security’, factors to be taken into account by a Member State are the period of residence, age, state of health, family and economic situation, social and cultural integration into the host Member State and the extent of links with the country of origin of the EU citizen. The second paragraph raises the bar, requiring ‘serious’ grounds of public policy or public security for those Union citizens having gained the right of permanent residence. Finally, the third paragraph provides that in cases where a Union citizen has resided in the host Member State for the previous 10 years, ‘imperative grounds of public security’ must be brought forward to justify an expulsion decision.
In Tsakouridis, the Court had already interpreted the notion of ‘imperative grounds of public security’ in the case of a person dealing with narcotics as part of an organized group. The Court accepted that an expulsion measure could be taken on such grounds, noting (para 47):
Since drug addiction represents a serious evil for the individual and is fraught with social and economic danger to mankind (see, to that effect, inter alia, Case 221/81 Wolf  ECR 3681, paragraph 9, and Eur. Court H.R., Aoulmi v. France, no. 50278/99, § 86, ECHR 2006‑I), trafficking in narcotics as part of an organised group could reach a level of intensity that might directly threaten the calm and physical security of the population as a whole or a large part of it.
The Court thus emphasized in particular the more general economic and social consequences of the crime for society as a reason to accept that an expulsion measure against the individual involved in such activities could legitimately be based on ‘imperative grounds of public security’.
In his opinion on P.I., Advocate General Bot chose a different path. As a starting point, he found that the present crime did not raise security concerns falling within the ambit of such ‘imperative grounds’, as the mere risk of reoffending was not sufficient for this purpose (para 40) and P.I.’s case also had to be distinguished from that of ‘sexual predators’ who indeed created a threat for society at large (para 44). He then suggested, however, that Article 28 (3) of the Directive merely provided for a ‘presumption of integration’ after 10 years of residence (para 55-56), which could be rebutted. Integration in the Directive, in his view, consisted of territorial and temporal, but also ‘qualitative’ elements: Mr I’s crimes showed a ‘lack of desire’ to integrate, and if they had been discovered earlier, he would never have been able to complete a 10 years period of residence (para 60). Essentially, Mr I is a fraud in the Advocate General’s eyes.
The problem with this approach, as Dimitry Kochenov has pointed out succinctly in his post on the opinion, is that it effectively introduces a new exception to the system of Article 28 of the Directive. Foreigners who commit a crime that does not amount to the threshold justifying expulsion on ‘imperative grounds of public security’ can thus still be denied protection from expulsion despite 10 years of residence. A slippery slope opens up in deciding what crimes will be considered as sufficiently grave to rebut the ‘presumption of integration’.
The Court in P.I. does not take up the Advocate General’s suggestion, but also deviates from its approach in Tsakouridis. Emphasizing that the Member States are to define ‘imperative grounds of public security’, the Court nonetheless insists on the ‘control by the institutions of the European Union’ over the discretion of Member States in that matter (paras 22-23). Examining the offences Mr I has committed, the Court refers to sexual exploitation of children as ‘one of the areas of particularly serious crime with a cross-border dimension’ under Article 83 (1) TFEU (para 25), which led the EU legislature to adopt Directive 2011/93 (para 26). In the light of this state of law, Member States may (paras 28-29)
regard criminal offences such as those referred to in the second subparagraph of Article 83(1) TFEU as constituting a particularly serious threat to one of the fundamental interests of society, which might pose a direct threat to the calm and physical security of the population and thus be covered by the concept of ‘imperative grounds of public security’ [...], as long as the manner in which such offences were committed discloses particularly serious characteristics, which is a matter for the referring court to determine [...].
Should the referring court find that, according to the particular values of the legal order of the Member State in which it has jurisdiction, offences such as those committed by Mr I. pose a direct threat to the calm and physical security of the population, that should not necessarily lead to the expulsion of the person concerned.
As the Court underlines, one must still ascertain whether the personal conduct of the individual represents a ‘genuine, present threat affecting one of the fundamental interests of society or of the host Member State’, which means that there must in general be a ‘propensity to act in the same way in the future’ (para 30).
However, remarkable emphasis is put on the ‘particular values of the legal order of the Member State’ to judge whether a ‘threat to the calm and physical security of the population’ is at stake (para 29); this can be contrasted with the economic and social danger for society at large referred to in Tsakouridis. The condition of a ‘genuine, present threat affecting one of the fundamental interests of society or of the host Member State’ is nothing but a reiteration of current case law (as are the subsequent paragraphs of the Court’s decision).
While the Court has thus not taken up the suggestion of the Advocate General, the judgment is still somewhat unclear on the scope of the concept of ‘imperative grounds of public security’. The structure of Article 28 of the Directive suggests with its Article 28 (3) that after a period of 10 years residence, the simple fact of this period of residence requires a specific treatment of EU citizens, preventing their expulsion from a host Member State unless there are exceptional circumstances (see also the Court in para 19 and in para 40 of Tsakouridis). The interpretation of ‘imperative grounds of public security’ is thus key to establish what these exceptional circumstances are. The Court seems to mainly render the concept of ‘imperative grounds of public security’ applicable here due to the particularly heinous nature of the crimes committed by Mr I, opting to leave it ‘open’ to Member States to regard crimes as particularly threatening based on the ‘particular values’ of their respective legal orders.
In principle, one can agree with the fact that some latitude is given to Member States by the Court to judge which ‘particular values of the legal order of the Member State’ become the benchmark for a threat to the ‘calm and physical security of the population’. As the Court puts it, no ‘uniform scale of values’ should be imposed by European Union law (para 21). A problem arises, however, if this discretion waters down the difference between expulsion under Article 28 (2) and (3). There must necessarily be a difference between ‘serious grounds of public policy or public security’ and ‘imperative grounds of public security’ to give effect to the text and structure of Article 28, as also the Court emphasizes (para 19). The examination of whether an individual constitutes a ‘genuine, present threat’ offers some leeway to assess public security issues, but is a standard test to be met also under the other paragraphs of Article 28. There must thus be more to a case to meet the threshold of ‘imperative grounds of public security’. A threat to a particularly important value in a Member States’ legal order on its own seems insufficient in this regard. The most pertinent part of the judgment seems to be here the mandate given to the national court by the CJEU to examine the ‘particularly serious characteristics’ disclosed by the manner in which the offences at issue were committed. But the Court fails to give any further guidance as to the nature of these characteristics.
What Article 28(3) with its omission of public policy grounds seems to suggest in my view is that – to a certain extent – the mere fact of a 10 years residence has created a link between the EU citizen and the host Member State that is similar to the link between a national and its state; as a consequence expulsion can only be a valid means if this link is deliberately destroyed by the EU citizen; this would be the case of a serious security threat, i.e. an individual determined to engage in actions that jeopardize the security of the host Member State’s society at large, which could indeed be the case of organized crime. It is thus not only the importance of the value under threat, but also the scope of the threat that should matter. In Tsakouridis, a plausible argument of a threat to public security could therefore be made based on the socioeconomic consequences of organized drug crime. A similar argument could perhaps be made if the case of Mr I would raise comparable elements of organized crime. But that does not seem to be the case here, which is perhaps the reason why the Court did not give more detailed guidance. It remains unfortunate that the Court missed an opportunity to set out in more detail how the national court’s assessment under Article 28 (3) of the Directive should differ from that under the other paragraphs of Article 28.