Threats to public security in EU immigration law: Finding the right discretion

Recently, the CJEU has handed down two judgments in E.P. and G.S & V.G. on the interpretation of the Schengen Borders Code (SBC) and the Family Reunification Directive respectively. On that occasion, the Court was asked to clarify the conditions pursuant to which a third-country national could be considered a ‘threat to public security’. Does it suffice that immigration authorities regard a third-country national with suspicion, or are they conversely required to reproduce the exigent standards developed in free movement law? The latter would require national authorities to demonstrate that a third-country national’s personal misconduct amounted to ‘a genuine, present and sufficiently serious threat affecting one of the fundamental interests of society’. The Court’s responses therefore affect the degree of discretion awarded to national immigration authorities in the determination of threats to public security in immigration law. Whereas these questions are indubitably of practical importance to national immigration authorities, they equally lend themselves to a reflection on national administrative discretion in EU (immigration) law more generally: When does the CJEU acknowledge a wide margin of appreciation to the benefit of national authorities, and for what reasons so?

The facts of the cases

The rulings in E.P. and G.S & V.G. arose against the background of negative immigration decisions by the Dutch Staatssecretaris van Justitie en Veiligheid. The decisions affected three individuals. First, E.P., an Albanian citizen who claimed to stay as a tourist in the Netherlands, where he was arrested during a police raid which led to the impounding of large amounts of drugs. The police assumed that he was indeed caught red handed. On the suspicion that he constituted a threat to public security, the Staatssecretaris ordered E.P. to leave the EU territory within 28 days, arguing that he no longer fulfilled the conditions for stay pursuant to the SBC. The second individual was G.S., an Indian national who had been sentenced to four years and three months imprisonment in Switzerland and whose extension of a residence permit in the Netherlands for the purpose of family reunification was rejected for that reason. Third, the application for family reunification by V.G., an Armenian who had been sentenced repeatedly for drug offences, was refused on the same grounds.

The referring court was uncertain whether Union law required the Staatssecretaris to base each of those decisions on the fact that the third-country nationals constituted ‘a genuine, present and sufficiently serious threat affecting one of the fundamental interests of society’. The Staatssecretaris firmly disputed that. In his view, national authorities would be under no obligation to apply the same standards as those developed in the context of free movement law affecting EU citizens as well as their family members. Rather, competent authorities should have wide margin of appreciation to determine whether third-country nationals should be considered threats to public security.


The CJEU agreed with the Staatssecretaris’ view. It found that mere reference to a threat to public security does not automatically reverberate the Court’s jurisprudence in free movement law. Rather, an award of discretion to national immigration authorities follows from the legal provision that gives rise to it, which calls, in return, for an interpretation in light of its wording, context and objective. In this vein, the judgments in E.P. and G.S & V.G. confirm a line of jurisdiction developed in Koushkaki and Fahimian, highlighting that the specific legislative context is key for determining the standards guiding administrative discretion.

Regarding the SBC in E.P., the Court confirmed that national authorities’ wide discretion by reason of the complexity surrounding the verification of entry conditions. Since conditions for issuing a visa are subject to complex assessments, the Court considered that a similar degree of complexity would logically permeate Schengen entry conditions as well. The Court consolidated that functional argument by producing indications for a deliberate choice taken by the EU legislature to leave the threat to public security unqualified. To that end, it first pointed to the fact that Article 6 of the SBC did not stick to the Court’s free movement jurisprudence for the definition of a ‘threat to public security’, even though this was the case elsewhere in the Code. In addition, the CJEU referred to the SIS II Regulation to highlight that the perceptions of a threat to public security are ‘appreciably different’ in the context of the Schengen acquis and in free movement law. By virtue of Schengen law, national administrations are awarded a wider margin of discretion to establish whether a third-country national poses a threat to public security. That discretion remains, nonetheless, bound to the principle of proportionality.

Similarly, with a view to the Family Reunification Directive, the CJEU acknowledged in G.S & V.G. that the intention of the legislature was to leave the national administrative authorities some leeway to determine threats to public security. The Court did so by looking at the Directive’s drafting history, finding that the Council had deliberately refrained from reproducing the standard developed in free movement law. In this context, the judges highlighted that other standards guiding national administrative discretion must be taken into account; particularly, the severity or type of offence as recital 14 of the Family Reunification Directive spells out; the need for an individualised assessment as ordained by Article 17 thereof as well as a strict interpretation of the proportionality principle. Since the Council had intentionally awarded national authorities a margin of discretion, however, the Court accepted that national immigration authorities could rely solely on the criminal conviction of the person to consider him a threat to public security, given that the negative immigration decision is proportionate.

All this demonstrates that the CJEU closely examines the legislation at play to verify the intention of the EU legislature to award national immigration authorities discretion. In doing so, it has recourse to both functional as well as doctrinal arguments. Nevertheless, the confirmation whether a provision of Union law provides for a (wide) margin of appreciation is not a foregone conclusion. Rather, in the absence of explicit legislative indications, the court picks from a bouquet of arguments that may cater to one end or another. This is, after all, not a bad thing. The Court’s approach may provide as much guidance as is needed to reduce – as far as possible – a capricious, excessive interpretation of ‘threats to public security’. However, the reasons for that approach should be spelled out clearly. This may be illustrated with regard to the arguments of complexity and severity that the Court addresses in E.P. and G.S & V.G..


It is widely accepted that administrative authorities may be awarded a wide margin of discretion if they are to carry out complex assessments. This may often be the case if the competent authorities wish to predict an individual’s future conduct. The appraisal of risks then reasonably calls for complex assessments. In the case of E.P., however, the Court seems to paper over that the situation is less characterised by complexity, but rather by a lack of information. E.P. is an Albanian citizen who was exempted from visa obligation, staying in the EU for a short-period of time. In such a situation, as some of the intervening national governments had claimed, competent authorities operate on the basis of limited information.

As a matter of fact, this may render it very difficult for national authorities to demonstrate ‘a genuine, present and sufficiently serious threat affecting one of the fundamental interests of society’. If such an assessment had to be based on the punctual information accessible, the CJEU would arguably require immigration authorities to write checks they cannot cash. The Court could have turned that finding into an argument of practicality: Since the EU legislature had not provided national authorities with the means for doing so, it did not have the intention to require them to demonstrate ‘a genuine, present and sufficiently serious threat affecting one of the fundamental interests of society’. As such, the situation in E.P. differs from the one in Koushkaki, in which the competent authorities were able to base complex assessments on solid information. The Court could have spelled out this argument more clearly, looking the argument of ‘complexity’ in more detail.


With regard to the determination of threats to public security, it must be noted that the Court does not draw a categorical distinction between free movement law, on the one hand, and the legal framework concerning third-country nationals, on the other hand. A cursory glance over the Court’s jurisprudence in the context of asylum law (see the judgments in N. and T.) confirms this. Even though the rulings concerned third-country nationals, the CJEU insisted that an individual must represent ‘a genuine, present and sufficiently serious threat, affecting a fundamental interest of society’ to be considered a threat to public security. As such, it reproduced the free movement standard in the context of asylum law. The judgments in E.P. and G.S. & V.G., to the contrary, reject the automatic reproduction of that standard. How can that contextual divergence be explained?

The judgments in E.P. and G.S & V.G. may allow for a glimpse behind the curtain in that regard. It may be noted that the strictness of standards imposed upon national administrative discretion correlates with the severity with which a negative immigration decision may affect the person concerned. The revocation of a refugee’s residence permit may yield very harsh repercussions, whereas the untimely end of a tourist visit – unpleasant as it may be – does not. Accordingly, and insofar as the wording of the legislation at question provides for it, the reasoning of the court falls neatly into place with considerations of severity.

This is not to say that it is intuitively clear which repercussions are considered severe in the eyes of the Court. But for that reason alone, the CJEU should put this card on the table: Just as it motivates its reasoning with regard to the complexity of assessment, it may equally limit national administrative discretion on the basis of the severity of its implications. Simply put, where consequences are dire, the Court should constrain national administrative discretion in determining threats to public security. Where they are not, it may acknowledge a wide margin of discretion. As a functional argument in favour or against national administrative discretion, severity could provide some indications which situations call for wide discretion and which do not.