Wednesday, 27 July 2016

The new Opinion on Data Retention: Does it protect the right to privacy?

Matthew White, Ph.D candidate, Sheffield Hallam University


Has an Advocate General (AG) in the Court of Justice of the EU unleashed the power of the European Convention on Human Rights (ECHR)? On 19 July 2016, the AG gave his Opinion in the joined cases of C203/15 and C698/15 Watson and Tele2. The AG felt that general data retention obligations  imposed by Member States may be compatible with fundamental rights enshrined in EU law, provided that there are robust safeguards (para 7). This post briefly outlines the background (for a more detailed background, see Professor Lorna Woods’s take) to this case whilst highlighting aspects relating to the ECHR and that some of the AG’s conclusions become self defeating for requiring EU law to be no less stringent than the ECHR.


Case C203/15

A day after Digital Rights Ireland (where the Court of Justice of the European Union (CJEU) ruled that the EU’s Data Retention Directive (DRD) was invalid for being incompatible with the Charter of Fundamental Rights (CFR)), the first claimant, Tele2, notified the Swedish Post and Telecommunications Authority (PTS) of its decision to cease retaining data in Chapter 6 of the LEK (the relevant Swedish law) with the aim of deleting (para 50). The National Police Board (RPS) complained to the PTS about Tele2’s actions as having serious consequences for law enforcement activities (para 51). PTS ordered Tele2 to resume retention in accordance with Chapter 6 (para 52), to which Tele2 appealed to the Stockholm Administrative Court (SAC) but lost (para 53).  Tele2 then sought to appeal against the SAC (para 54), but the Stockholm Administrative Court of Appeal (SACA) felt making a preliminary reference to the CJEU would be more appropriate where it asked:

·         Is a general obligation to retain all traffic data indiscriminately compatible with Art.15(1) of the (ePrivacy) Directive (Directive) and Articles 7, 8 and 52(1) of the CFR?

·         If no, is such an obligation nevertheless permitted  where:
§  access by national authorities was governed in a specified manner, and
§  the protection and security of data are regulated in a specified manner, and
§  all relevant data is retained for six months?

Case C698/15 

I previously blogged on the situation in the UK, but will make a quick summary for the purposes of this post (or alternatively see paras 56-60 of the Opinion). The UK responded to Digital Rights Ireland by introducing the Data Retention Investigatory Powers Act 2014 (DRIPA 2014). This was successfully challenged in the High Court by Tom Watson MP and David Davis MP. But the success was short lived when the Court of Appeal disagreed with the High Court, but made a preliminary reference to the CJEU asking:

·         Did the CJEU in Digital Rights Ireland intend to lay down mandatory requirements of EU law with which the national legislation of Member States must comply?
·         Did the CJEU in Digital Rights Ireland intend to expand the effect of Articles 7 and/or 8, EU Charter beyond the effect of Article 8 ECHR as established in the jurisprudence of the ECtHR?

AG’s Opinion

Asking the wrong question?

The AG initially dealt with the question regarding whether Digital Rights Ireland extended the scope of Article 7 and/or Article 8 of the CFR beyond that of Article 8 of the ECHR. The AG considered this question inadmissible (para 70 and 83) because that possibility was not directly relevant to the resolution of the current dispute (para 75). The AG admitted that the first sentence of Article 52(3) (which lays down rules of interpretation) of the CFR makes clear that any corresponding rights must be the same in meaning and scope to that of the ECHR (para 77). But highlighted, the second sentence of Article 52(3), can permit CJEU to extend the scope of the CFR beyond that of the ECHR (para 78). The ECHR has always been a minimum benchmark as in Trucl and Others v Slovenia it was noted that ‘rights guaranteed by the Convention represented minimum standards’ (para 115). Thus if the EU did acceded to the ECHR (and even if it did not), with or without the second sentence Article 52(3), the CJEU would be free to extend the scope CFR as it saw fit. Therefore in agreement with the AG, the Court of Appeal asked the wrong question.

Lack of corresponding right means rules of interpretation does not apply?

Another important aspect was  pointed out by the AG, who maintained that Article 8 of the CFR has no ECHR corresponding right and therefore the rules of interpretation laid out in the first sentence of Article 52(3) does not apply (para 79). However, there is cause for slight disagreement on this interpretation of Article 52(3). While the High Court admitted that protection of personal data fell within the ambit of Article 8 of the ECHR, they felt Article 8 of the CFR went beyond this because it was more specific and the ECHR had no counterpart (para 80). However, the High Court did so without actually considering Article 8 ECHR case law, therefore their conclusions did not appear to based on anything but mere conjecture and the wording of Article 8 CFR.  This was also questioned by Stalla-Bourdillon because it appeared the High Court followed this interpretation based on there not being an ECHR counterpart. But on closer inspection, as Stalla-Bourdillon highlighted, there is extensive Article 8 case law on the protection of personal data, which is suggested, does in fact correspond with Article 8 CFR. Therefore, both the High Court and AG has fallen prey to only considering the provisions of the ECHR and not the European Court of Human Rights’s (ECtHR) interpretation of those provisions, thus substance over form seemingly prevailed.

It is suggested because there is such extensive case law on the protection of personal data in light of Article 8 ECHR, it is only right that it should be used as a guide when considering Article 8 CFR. Article 52(3) notes that ‘the meaning and scope of those rights shall be the same as those laid down by the said Convention.’ In PPU J McB v LE the CJEU held that not only does the rights set out in ECHR are to correspond, but also the meaning given through the ECtHR’s jurisprudence (para 53) (see also). In Schecke the CJEU held that:

[T]he right to respect for private life with regard to the processing of personal data, recognised by Articles 7 and 8 of the Charter, concerns any information relating to an identified or identifiable individual...and the limitations which may lawfully be imposed on the right to the protection of personal data correspond to those tolerated in relation to Article 8 of the Convention.’(para 52). 

It has been maintained that such an interpretation can be problematic because the CJEU has allowed Article 8 CFR to be absorbed by Article 7. However, this does not and would not weaken the stance that Article 8 CFR as a standalone right should be interpreted (where possible) in accordance with principles of data protection embedded within the ECtHR’s jurisprudence. Read as a whole, Article 52(3) would therefore be properly adhered to, and would also allow the CJEU to deviate, if need be, to offer a higher standard of protection.

A general obligation to retain:

The AG then considered whether Article 15(1) of the Directive allowed Member States to impose a general data retention obligation (para 84) by establishing whether such an obligation fell within the scope of the Directive (para 86). The Czech, French, Polish and United Kingdom Governments all contended that data retention was excluded by Article 1(3) (which excludes matters such as public security, defence, State security from the scope of the Directive) (para 88). However, the AG rejected this by highlighting that:

·         Article 15(1) governed precisely that (retention of data) (para 90),
·         Provisions of access falling within Article 1(3) does not preclude retention from falling within Article 1(3) (para 92-94),
·         The approach taken by the CJEU in Ireland v Parliament and Council meant that general data obligations were not within the scope of criminal law (para 95).

When it came to the issue of whether the Directive applied the AG referred to the Member States ‘entitlement’ under Article 15(1) i.e. Member States have a choice (para 106). The AG then referred to Recital 11 of the Directive which did not alter the balance between an individual’s right to privacy and the possibility of Member States to take measures necessary for the protection of public security etc (para 107). Moreover, the AG highlighted that the Directive did not alter the ability of Member States to carry out lawful interception of electronic communications, or take other measures, if necessary for any of these purposes and in accordance with the ECHR (para 107). The AG opined that general data retention obligations were consistent with the Directive and therefore Member States were entitled to avail themselves of that possibility under Article 15(1), subject not only to its requirements, but that of the CFR in light of Digital Rights Ireland (para 116). Although the AG felt that general obligations of data retention were permissible under EU law (subject to restrictions), an avenue was created for testing the general obligations itself under the ECHR.

In accordance with  the law? But does this not defeat the AG’s premise?

When the AG considered the requirement for legal basis in national law, he invited that CJEU to confirm that the interpretation of ‘provided for by law’ in Article 52(1) CFR accorded with that of the ECtHR’s jurisprudence on a measure being ‘in accordance with the law’(para 134-137). The AG highlighted that the ECtHR has developed a substantial body of jurisprudence on the matter which could be summarised as follows:

·         A legal basis that is adequately accessible and foreseeable i.e. the law is formulated with sufficient precision to enable the individual — if need be with appropriate advice — to regulate their conduct,
·         This legal basis must provide adequate protection against arbitrary interference, and
·         Must define with sufficient clarity the scope and manner of exercise of the power conferred on the competent authorities (para 139).

The AG was of the opinion that ‘provided for by law’ in Article 52(1) CFR needs to be the same as that ascribed to it in connection with the ECHR (para 140). The AG’s reasoning was as follows:

·         Article 53 CFR explains that its provisions must never be inferior to what is guaranteed by the ECHR and therefore the CFR must at least be as stringent as the ECHR (para 141),
·         It would be inappropriate to impose different criteria on the Member States depending on which of those two instruments was under consideration (para 142).

The AG felt that general data retention obligations must be founded on a legal basis that is adequately accessible and foreseeable and provides adequate protection against arbitrary interference (para 143). This would solve the problem of the CJEU falling into ‘the trap of tautologically regarding a legal norm, the validity of which is being questioned, as being allegedly in accordance with the law because it is a law.

This then raises the interesting issue, if this is the preferred interpretation, how could a general obligation to retain data not amount to arbitrary interference? The AG later admits that the disadvantage of this general obligation arises ‘from the fact that the vast majority of the data retained will relate to persons who will never be connected in any way with serious crime’ (para 252). If the vast majority of data retained is of individuals who are unrelated to any serious crime, how could this even be suggested to not be arbitrary?

If in line with the ECtHR’s jurisprudence, that for a measure to be in accordance with the law, a measure must be sufficiently precise so individuals can regulate their conduct, how could this square with general obligations to retain data which occurs irrespective of conduct? The ECtHR’s Grand Chamber in Zakharov v Russia maintained that the ‘automatic storage for six months of clearly irrelevant data cannot be considered justified under Article 8’ (para 255). As the AG indicated, most data retained will have no relation to the fight against serious crime and therefore, in line with Zakharov, cannot be justified under Article 8. Member States would then have to justify why most data unrelated to serious crime is relevant to the fight against serious crime. In stressing that Article 52(1) should reflect the ECtHR’s jurisprudence the AG may have undermined his own position when believing that general obligations to retain data were permissible under EU law by unleashing the ECHR in terms of Recital 11 and the interpretation of ‘provided for by law.’

Data retention does not adversely affect the essence of the right, or does it, or should it?

The AG listed six requirements a general data retention obligation must meet to be justified, one of such is that it ‘must observe the essence of the rights enshrined in the Charter’ (para 132). The AG recalled that Article 52(1) CFR provides that any limitation to the rights enshrined must respect the essence of those rights and freedoms (para 155). The AG referred to para 39 of Digital Rights Ireland where the CJEU held that the DRD did not adversely affect Article 7 CFR since it did not permit the acquisition of knowledge of the content of the electronic communications as such (para 156). The AG felt this also applied to the current case (para 157) and this was equally the case for Article 8 CFR (paras 158-9) but ultimately left it for the CJEU to decide (para 160).

However, the AG later contradicts his own opinion when considering the disadvantages of data retention. The AG accepted that ‘a general data retention obligation will facilitate equally serious interference as targeted surveillance measures, including those which intercept the content of communications’ (para 254). The AG stopped short of referring to data retention as mass surveillance, but instead referred to it as mass interference (para 255) and that it affected a substantial portion, if not all of the relevant population (para 256). The AG even went further by describing with the example of an individual who access retained data (instead of analysing content) to screen out those within the Member State who have a psychological disorder or any field specialist medicine (para 257). The AG continues, this same person who sought to find out who opposed government policies, could do so with the possibility of identifying individuals taking part in public demonstrations against the government (para 258). 

The AG agreed with the position of several civil society groups, the Law Society and United Nations High Commissioner for Human Rights that the ‘risks associated with access to communications data (or ‘metadata’) may be as great or even greater than those arising from access to the content of communications’ (para 259). The AG further added that the examples given demonstrate that ‘metadata’ can facilitate ‘the almost instantaneous cataloguing of entire populations, something which the content of communications does not’ (para 259). The AG also added that there is was nothing theoretical about the risks of abuse or illegal access to retained data (based on the number of requests by Swedish and UK authorities) and that such risk of ‘illegal access on the part of any person, is as substantial as the existence of computerised databases is extensive’ (para 260).

Considering the incredible detail the AG went to describe the risks posed by the retention of data, it makes little sense to have the opinion that a general data retention obligation does not adversely affect the essence of the right. The AG and CJEU in Digital Rights Ireland premise of this was based on the idea that communications data would not permit acquisition of knowledge of the content of the electronic communications. Yet the AG described in great detail the amount of knowledge that could be gained from communications data. And it is this acquisition of knowledge that is the important factor, the AG described the example of the ability of gaining sensitive knowledge without analysing the content. And so the AG, like the CJEU has created an arbitrary distinction that although the same knowledge can be gained from communications data or content, it is only access to content that could adversely affect the essence of the right (para 94). If it is acknowledged that similar knowledge can be gained from both measures, the CJEU and indeed the AG has not sufficient explained this differential treatment. Furthermore, by only considering that access to content adversely affects the essence of the right, this would promote the use of retention and access to communications data to a greater degree which as the AG admits, can provide far richer information than content.

Indiscriminate data retention maybe EU compliant, but not ECHR compliant

The AG highlighted that the CJEU in Digital Rights Ireland pointed out that the DRD covered all users and all traffic data without differentiation or limitation (para 197). The AG described what the CJEU considered the practical implications of the absence of differentiation i.e. concerning those with no link to serious crime, no relationship between retention and threat to public security, and no temporal, geographical and associate based restriction (para 198). The AG concluded that the CJEU did not hold that the absence of differentiation in itself went beyond what was strictly necessary (para 199).

The AG justified this one four grounds, firstly, the CJEU ruled the DRD as invalid because of the cumulative effects of generalised data retention and the lack of safeguards which sought to limit what strictly necessary for the interference with Article 7 and 8 CFR (paras 201-202). Secondly, in light of Schrems (para 93) the AG inferred again that only general data retention obligations accompanied by sufficient safeguards would be EU law compatible (para 205). Thirdly, the AG felt national measures should be scrutinised at a national level, where the national courts should rigorously verify whether general data retention obligations are the most effective at fighting serious crime i.e. whether there are other less intrusive alternatives (paras 209-210). Fourthly, the AG agreed with the Estonian Government that limiting data retention to a particular geographical area may cause a geographical shift in criminal activity (para 214).

Considering indiscriminate data retention as permissible under EU law if there is a sufficiently robust safeguard mechanism creates problems with the ECHR. In the case of S and Marper v United Kingdom the issue at hand was the retention of finger print and DNA records. In finding the retention regime incompatible with Article 8 (para 126) the ECtHR was struck by blanket and indiscriminate nature of the power because:

[119] material may be retained irrespective of the nature or gravity of the offence with which the individual was originally suspected or of the age of the suspected offender...
[122] Of particular concern in the present context is the risk of stigmatisation, stemming from the fact that persons in the position of the applicants, who have not been convicted of any offence and are entitled to the presumption of innocence, are treated in the same way as convicted persons...
[125] In conclusion, the Court finds that the blanket and indiscriminate nature of the powers of retention of the fingerprints, cellular samples and DNA profiles of persons suspected but not convicted of offences, as applied in the case of the present applicants, fails to strike a fair balance between the competing public and private interests and that the respondent State has overstepped any acceptable margin of appreciation in this regard. Accordingly, the retention at issue constitutes a disproportionate interference with the applicants’ right to respect for private life and cannot be regarded as necessary in a democratic society. This conclusion obviates the need for the Court to consider the applicants’ criticism regarding the adequacy of certain particular safeguards, such as too broad an access to the personal data concerned and insufficient protection against the misuse or abuse of such data.

 S and Marper’s significance has been linked to data retention, and therefore it is important to apply the principles to the AG’s Opinion. The ECtHR criticised the UK regime for not distinguishing between those who had been suspected and those who had committed offences. Regarding data retention obligations, this indiscriminate power is more profound because suspicion would not be a necessary component for the justification of retention. As the AG highlighted, most data retained is of individuals who bare no relation to serious crime and therefore creates issue with the presumption of innocence to an unacceptable level. The most important aspect of the ECtHR’s reasoning in S and Marper was that the retention itself was contrary to the Convention without having to consider the safeguards that may have been in place. This is direct contrast with Digital Rights Ireland and the AG’s Opinion.

Regarding the fourth point, it is submitted that the Estonian Government and the AG misunderstood how data retention and location data works in practice. It is not the physical area that is the important factor, but the location of the device in question at a particular time. This was apparent in Uzun v Germany when the ECtHR described Global Positioning System (GPS) as allowing ‘continuous location, without lapse of time, of objects equipped with a GPS receiver anywhere on earth, with a maximum tolerance of 50 metres at the time’ (para 12-13). This is all the more relevant as location data is becoming more and more sophisticated. Therefore applying a data retention obligation in a specific geographical area creates a false premise as the obligation on the service provider is to keep record of the location data of a device when it’s service is used (which will indicate where an individual might be) irrespective of geographical area. Furthermore a targeted data retention approach would not be confined to a geographical as such, but to criminal activity (based on individual use of device and service) within a particular area.      
Six months retention is reasonable?

The issue of retention period was also considered by the AG (para 242) who felt that according to Zakharov a period of six months would be reasonable provided irrelevant data was immediately destroyed (para 243). However, by making this connection, the AG created a false analogy of what the ECtHR held. Zakharov concerned judicially authorised interception and monitoring of communications data of individuals for six months (para 44-48). Therefore the analogy with targeted measures and that of general data retention begins to falter, as in the AG’s own words ‘metadata’ facilitates the almost instantaneous cataloguing of entire populations, something which the content of communications (via interception) does not (para 259).


Although most of the finer details, in the Opinion of the AG should be left to national courts (para 263) the issue of data retention as a challenge to fundamental rights persists. The AG, by placing great significance on the ECHR and the ECtHR’s jurisprudence unwittingly undermined some of his own key points because they do not accord with the ECHR. It is unlikely that the CJEU are going to rule per se that a general obligation to retain communications data is incompatible with EU law, and therefore maybe an issue for the ECtHR to decide themselves. In light of S and Marper it is possible that the ECtHR would produce a ruling that is in contrast to the CJEU. The United Nations General Assembly has affirmed that same rights that people have offline must also be protected online. The late Caspar Bowden once described data retention as akin to having CCTV inside your head. And so the question becomes, would the AG/CJEU consider that CCTV inside every home would be compatible with EU law provided that access to that footage would be circumscribed by adequate safeguards?

Barnard & Peers: chapter 9
JHA4: chapter II:7

Photo credit:

Monday, 25 July 2016

Domestic violence and free movement of EU citizens: a shameful CJEU ruling

Steve Peers

EU laws on the free movement of EU citizens don’t give non-EU citizens rights in their own name. Rather, non-EU citizens can only gain rights under those laws if they have a family link with an EU citizen. This creates an obvious problem in cases of domestic violence committed by an EU citizen against a non-EU citizen family member. If the non-EU family member breaks the family link in order to flee the violence, there could be a risk of expulsion. So the victims might stay with their abusers due to a fear of removal from the country, which might include separation from their children.

However, there are provisions of EU law that mitigate this risk. It has long been the position (since the CJEU judgment in Diatta) that non-EU citizens can remain in the same country as their EU citizen spouse following a separation. Their position only changes after divorce.

Upon divorce, the EU citizens’ Directive (which sets out most of the rules governing EU citizens who moved to another Member State) provides specific protection. As a general rule, they can remain if the marriage has lasted three years, including one year in the host State. They can also stay if they have custody of the children, or access to them in the host State. Finally, they can also stay if ‘this is warranted by particularly difficult circumstances, such as having been a victim of domestic violence while the marriage…was subsisting’. In any of these cases, they can later get permanent residence status.

The recent judgment in NA was the first time the Court of Justice has interpreted the specific rule on domestic violence cases. It follows last year’s judgment in Singh, where the Court first interpreted the general rule about divorce. Both cases raised the same underlying issue: what happens if the EU citizen leaves the host State before the divorce is finalised? Does that departure immediately end the non-EU citizen’s status under the citizens’ Directive, trumping the retention of their residence that would otherwise apply during their separation and (probably) their subsequent divorce from their spouse?  

According to the Court of Justice, it does. This reasoning was an unconvincing interpretation of the general rule in Singh, and it is a particularly unconvincing interpretation of the special rule on domestic violence victims, taking insufficient account of the social context of this rule. While the victim in NA was able to rely on other provisions of EU law, not all victims will be able to. The Court of Justice should therefore rethink its position if the case arises – particularly in light of the EU’s planned signature of the Istanbul Convention on violence against women.

The judgment

The case concerned the status of a Pakistani woman who moved to the UK with her German husband. (The judgment will no longer be relevant to the UK if the country leaves the EU without an agreement on the continued free movement of people. However, it will still be relevant to other Member States). Her husband worked in the UK, and the couple had two daughters, both of whom are German citizens. She left the household due to domestic violence, and her husband left the country shortly after that, before any divorce proceedings began. She subsequently sought permanent residence status in the UK.

Did she retain rights under the citizens’ Directive? As noted above, the Court said she did not. It simply followed Singh, ruling that rights for the non-EU family member ended the moment the EU citizen left the country, unless divorce proceedings had begun beforehand. It gave three reasons. First, the wording of the relevant clause referred only to divorce. Second, the context was the ‘exceptional’ case where a non-EU citizen retained a right to stay despite no longer being a ‘family member’ of an EU citizen living in the same Member State. Third, the Court referred to the ‘aims’ of the law. In its view, the EU legislature had declined to make provision for cases where the EU citizen had departed the Member State, and the original proposal referred to possible ‘blackmail accompanied by threats of divorce’, with ‘safeguards’ only ‘necessary…in the event of final divorce’, as the right of residence is ‘not at all affected’ by a ‘de facto separation’.  

So she had no rights under the citizens’ Directive. However, the Court then examined two other arguments for her stay. First, the Regulation on free movement of workers says that the children of EU workers are entitled to access education. The Court had previously ruled that this entailed a right for the children to stay even if the worker had left, along with a corollary right for the parent caring for them to stay as well. The UK court questioned whether this rule applied even if (as in this case) the children only started school after the worker had left the country; the Court of Justice confirmed prior case law that it did. It was sufficient that the children were resident at some point while one parent was working in that Member State.

Finally, the Court examined a two-fold argument for NA’s right to stay on the basis of the EU Treaties. The first leg of this argument invoked Ruiz Zambrano, the well-known 2011 CJEU judgment which said that non-EU family members of EU citizen children who had not moved within the EU could derive a right to stay based on their children’s rights as EU citizens not to be deprived of the benefits of their EU citizenship. But the Court said that Zambrano was irrelevant to NA, since it only applied as a default, where no EU legislation could protect the legal status of the person concerned.

The second leg invoked the general right of free movement of EU citizens set out in the Treaties. But the Court ruled that the Treaty right was subject to secondary legislation. Applying the citizens’ Directive, NA’s EU citizen children could stay if they had ‘sufficient resources’, which could be derived from a parent. There was then a corollary right for a parent to stay with them – reaffirming case law going back to the Court’s well-known Chen and Zhu judgment.


In this case, Ms NA got to stay in the country – but that wasn’t actually the issue. She was seeking rather the right to permanent residence, but at first sight she will not obtain that. That right applies to a family member who retains a right of residence under the citizens’ Directive following divorce (among other cases). But it’s not clear if it applies to those who are resident only as a corollary to their children under the Directive. And it certainly does not apply to those who are only resident on the basis of the Regulation on free movement of workers, rather than the Directive: the Court said as much in its Alarape judgment.

The bigger problem with this judgment is the scope it opens to Member States to remove the victims of domestic violence from their territory. Ms NA only got the right to stay on the basis of her care for the children, not as a victim of domestic violence. So a victim without children would not have such protection. Also, rather arbitrarily, a victim whose husband had been self-employed, rather than a worker, would not benefit from the free movement of workers Regulation (see the Czop judgment). Moreover, the Court skipped over the point that her second child had not yet been born at the time when her husband left the country. The judgment would not help those victims whose children had been abducted by the husband when he returned to his home State, or who did not have parental responsibility for the children.

The logic of the Court’s analysis is deeply flawed. First, it isn’t self-evident that the rule on divorce is trumped in the event of separation: the EU legislature simply didn’t explain which rule takes precedence in that case. Secondly, the numbers of non-EU citizens with the right to stay will still remain ‘exceptional’, even if it is extended to cover also the (hopefully) small numbers of cases where an EU citizen perpetrates domestic violence against a non-EU citizen and then leaves the country before divorce proceedings start.

Thirdly, the Court’s analysis of the aims of the EU legislation is clearly absurd. True, the right of residence is ‘not at all affected’ by a ‘de facto separation’ – as long as the EU citizen spouse remains in the country. If the EU citizen spouse leaves, according to the Court’s own interpretation, the right of residence isn’t just ‘affected’ – it instantly vanishes entirely. So in that case ‘safeguards’ would be ‘necessary’. Otherwise there could be ‘blackmail accompanied by threats’ of departure, rather than divorce. Did the EU legislature really intend to make that fine distinction: it’s terrible to threaten a non-EU spouse in one case, but perfectly acceptable in the other? Does the humanity of our response to domestic violence rest on that technicality?

This analysis is shared by the Advocate-General’s opinion, which notes that the loss of status ‘could be used as a means of exerting pressure…to wear the victim down psychologically and, in any event, to engender fear of the perpetrator’. The Court’s interpretation could also complicate criminal proceedings, and deprive the victim’s derived right of its effectiveness. Overall that interpretation is ‘manifestly contrary to the objective of legal protection pursued by’ the citizens’ Directive.

Two final points on the broader context. First, the Istanbul Convention on violence against women, which the Commission has proposed that the EU sign, provides for a right of residence in domestic violence cases (see Article 59). There’s no reference to any distinction based on whether the perpetrator has left the country or not. Neither is there any such reference in the explanatory memorandum to the Convention. And why the hell should there be? Who cares where the perpetrator is, in this context? The sole purpose of the Convention – like the relevant clause in the citizens’ Directive – is obviously to help the victim.

Secondly, let’s examine the law from the perspective of the actual victim in this case. She could have preserved her position by bringing divorce proceedings before her husband left the country. But she had just fled her home, five months pregnant with an eleven-month old child. She may well have faced problems relating to work, benefits or accommodation. And on top of all this, the Court of Justice gives the nod to the Home Office to question her immigration status.

With all due respect, this is one of the most shameful judgments in the Court’s long history. It should be revisited at the earliest opportunity, in particular if the EU has concluded the Istanbul Convention in the meantime.

Barnard & Peers: chapter 13
JHA4: chapter I:6

Photo image:

Friday, 22 July 2016

H v Council: Another Court breakthrough in the Common Foreign and Security Policy

Graham Butler, PhD Fellow, Centre for Comparative and European Constitutional Studies, Faculty of Law, University of Copenhagen, Denmark

This summer alone, the Court of Justice (‘the Court’) has issued two important decisions that will further shape the legal dimension of the Common Foreign and Security Policy (CFSP). Despite this largely intergovernmental sphere of law (the former Second Pillar) being merged into the unified ‘EU’ at the Treaty of Lisbon, the pillar’s shadow still lives on. Lasting evidence of CFSP as a separate but integrated sphere of law allow for it to be titled ‘CFSP law’, with judgments of the Court arising from interinstitutional and direct action litigation, permitting its legal development. The two judgments, Tanzania (Case C-263/14) and H v. Council (Case C-455/14 P) address different questions, and with a third, Rosneft (Case C-72/15), being delivered later in the year. This sequence of judgments demonstrates the fluidity of CFSP dynamics. In this blog post, analysis will focus on the H v. Council judgment, and specifically, given its peculiarity, the jurisdiction of the Court of Justice in CFSP.

From a simple reading of the Treaties, it would be assumed that three individual articles pose problems for the Court of Justice’s jurisdiction is matters pertaining to CFSP.

Article 24(1) TEU states, ‘…The Court of Justice of the European Union shall not have jurisdiction with respect to these provisions, with the exception of its jurisdiction to monitor compliance with Article 40 of this Treaty and to review the legality of certain decisions as provided for by the second paragraph of Article 275 of the Treaty on the Functioning of the European Union.’

Article 40 TEU states, ‘The implementation of the common foreign and security policy shall not affect the application of the procedures and the extent of the powers of the institutions laid down by the Treaties for the exercise of the Union competences referred to in Articles 3 to 6 of the Treaty on the Functioning of the European Union. Similarly, the implementation of the policies listed in those Articles shall not affect the application of the procedures and the extent of the powers of the institutions laid down by the Treaties for the exercise of the Union competences under this Chapter.’

Article 275 TFEU states, ‘The Court of Justice of the European Union shall not have jurisdiction with respect to the provisions relating to the common foreign and security policy nor with respect to acts adopted on the basis of those provisions. However, the Court shall have jurisdiction to monitor compliance with Article 40 of the Treaty on European Union and to rule on proceedings, brought in accordance with the conditions laid down in the fourth paragraph of Article 263 of this Treaty, reviewing the legality of decisions providing for restrictive measures against natural or legal persons adopted by the Council on the basis of Chapter 2 of Title V of the Treaty on European Union.’

Generally therefore, it can be assumed that the there is no basis for the Court to possess jurisdiction in matters relating to CFSP, set out between Articles 21-46 TEU (‘Title V’). All relatively straightforward it would seem? Not necessarily so. In fact, the complexity of the rules, and the derogations set out are subject to varying interpretations which inevitably end up before the Court such as this case in point.

H v. Council (Case C-455/14 P) was an appeal of an Order by the General Court, concerning jurisdiction to rule on a staff dispute raised by an official serving the European Union Police Mission (EUPM) in Bosnia and Herzegovina (BiH). Initially, upon receipt of the case, the General Court made an Order on 10 July 2014 (Case T-271/10, H v. Council, ECLI:EU:T:2014:702) stating that the General Court had no jurisdiction on the matter given it was CFSP, thereby interpreting the Treaties provisions on lack of jurisdiction, Articles 24(1) TEU and Article 275 TFEU in a broad manner. H appealed the Order of the General Court to the Court, as H was of the view that the staffing issue within CFSP was an administrative act, and cannot be construed as to entail non-jurisdiction of the Court, notwithstanding the fact that the EUPM was formed on a CFSP legal basis. The case presented itself as another opportunity for the Court to refine what the extent of the jurisdiction of the Court is within the delicate field of CFSP.

In the General Court, both the Council and the Commission said the matter was CFSP, and therefore, pursuant to the second paragraph of Article 24(1) and the first paragraph of Article 275 TFEU, that the General Court did not possess any jurisdiction over it. On this point, the General Court agreed. H disagreed, and in this appeal, set out two pleadings. The Court of Justice only dealt with the second plea – the jurisdiction of the EU judiciary in matters pertaining to CFSP. Given its findings on the second ground of appeal on jurisdiction, it said there was no need to examine the first.

Following the Opinion of Advocate-General Wahl delivered in April 2016, the Court of Justice delivered its judgment in July 2016. Whilst the Advocate-General said the General Court was right by saying the Court had no jurisdiction, the Court took a contrasting outlook to them both. Meeting in Grand Chamber, the Court acknowledged that, principally, it can be assumed that the jurisdiction of the Court does not extend to the CFSP provisions of the Treaties, known as Title V, or Articles 21-46 TEU. Whilst this was the case, the Court also said the general exclusion of the Court cannot extend to all aspects of CFSP. Based on this assertion, it can be assumed that acts adopted on a CFSP legal basis may come within the Court’s jurisdiction. In this case, this was despite the EUPM being grounded on a CFSP legal basis through Article 28 TEU and Article 43(2) TEU. The issue was therefore, should administrative decisions falling within the ‘day-to-day’ sphere of operations on the ground in the EUPM constitute non-jurisdiction of the Court?

Decision 2000/906/CFSP amongst other things, set outs the staffing arrangements for the EUPM. What is clear from the Decision is that staff on the ground in EUPM are all subject to the rules and the direction of the Civilian Operation Commander. This is sensical, given that the idea of senior officials not being in central control of all staff would be an operational and logistical nightmare. On closer inspection of the Decision however, the legal positions of positions are in fact distinct, in that some are seconded from national public bodies, and others are seconded from various EU institutions, agencies, and bodies. Despite this difference, the Decision allows for the coordination of day-to-day operations to cover ‘all’ staff. This is the tool that the Court uses to prise open jurisdiction for the matter at hand. Given that acts of staff management occurs in all EU public bodies, the Court noted that the CFSP Decision on staff arrangements within EUPM is similar to those exercised in EU institutions (paragraph 54). As a result, the Court believes that the derogations imposed on the Court’s jurisdiction in both Article 24(1) TEU and Article 275 TFEU cannot prevent the Court from exercising review over staff management in EUPM, notwithstanding the fact that EUPM is provided for by a CFSP legal basis.

This interpretation by the Court of Justice is not without further justification. Reliance is placed upon a Decision governing the statute, seat and operational rules of the European Defence Agency (EDA). The Council’s Decision on the EDA, 2015/1835/CFSP, confers jurisdiction upon the Court to adjudicate on matters relating to seconded national experts. Furthermore on judicial review, the Court said, ‘…the very existence of effective judicial review [is] designed to ensure compliance with provisions of EU Law…’ (paragraph 41). This not the first time that the Court has used ‘rule of law’ considerations in justifying allowances for judicial review. In addition, the Court stated that the issue in this case was redeployment, and not secondment itself, which it says the General Court mistook. Conclusively, the ‘no jurisdiction of the Court’ articles in the Treaties for CFSP does not mean everything in a CFSP mission is beyond the Court’s reach. The Court therefore concluded that the General Court erred in taking a broad view the Court’s principally excluded position.

The Court’s judgment in H v. Council appears to be sound. To interpret all matters relating to CFSP missions, including the administrative, procedural, and operational issues as being ‘CFSP acts’ to escape judicial oversight of the EU judicial body would have been over-interpretation of the restrictions on the Court which have been set down by the Treaties. What construes a ‘CFSP act’ has gotten smaller as a result, as the Court took a narrow construal of what a CFSP act is, and the derogations imposed on the restrictive judicial review arrangements. The Treaties distinguish between acts of foreign policy, and implementing acts. Article 40 TEU states, ‘The implementation of…[CFSP] shall not affect the application of the procedures and the extent of the powers of the institutions laid down by the Treaties for the exercise of the Union competences…’. Even though implementing acts of foreign policy, it would be erroneous to construe all decisions of various importance within CFSP as ‘CFSP acts’, thereby excluding judicial review.

Notwithstanding the environment in which missions like EUPM operate, the Council’s argument that an operational issue in the context of security and defence should fall outside the EU judicature is not particularly strong. The Council has basically shot itself in the foot as a result of the EDA Decision last year. Perhaps without realising the full ramifications of Council Decision 2015/1835/CFSP in 2015, it is now having a spill-over effect. The Council itself trampled over its own arguments by granting the Court jurisdiction within one CFSP Decision on the EDA in 2015, but then trying to claim that it does not have the same adjudication powers within another early CFSP Decision on the EUPM. This argument wasn’t sustainable, which the Council would have realised when the Court likely queried the matter during oral hearing.

H v. Council is another case in a series of breakthroughs for the Court of Justice in CFSP. Just because a measure is concluded on a CFSP legal basis does not de-facto exclude the Court. The Court can use secondary Union law, a CFSP Decision, to prise open the jurisdictional bounds imposed on the former Second Pillar. The Mauritius (Case C-658/11) and Tanzania (Case C-263/14) cases have demonstrated that the Court has been strong on institutional procedure, and it is notable that the Court has opened up its jurisdiction in CFSP once again, without making use of Article 40 TEU – its ‘border policing’ provisions between CFSP and non-CFSP. The outcome of this case meant firstly, the General Court’s order finding of no jurisdiction has been set aside. The Statute of the Court (Article 61) permits it to send back cases to the General Court, for which it will be bound now on points of law that have been issued by the Court. Hence, the Court has bounced the issue back to the General Court to decide the case on matters of substance, now that its jurisdiction has been affirmed.

The forthcoming Rosneft judgment (Case C-72/15, Rosneft Oil Company OJSC v. Her Majesty’s Treasury and Others), a preliminary reference dealing with, amongst other things, the Court’s jurisdiction in CFSP, will be decided by the Court later this year, and will be keenly watched. (See discussion of the opinion in that case here). As time goes on, the two differing legal regimes of CFSP and non-CFSP on EU external action will continue, unless ‘splitting’ begins to become more commonplace, where EU measures are decided upon a dual legal basis. Until such a time when these issues of EU constitutional law are ironed out, the Court will continue to be asked detailed intricate questions about its jurisdiction in CFSP.

Barnard & Peers: chapter 10, chapter 24

Photo credit:

Wednesday, 20 July 2016

The new Blue Card proposal: Will it attract more highly skilled workers to the EU?

Jean-Baptiste Farcy, Research Assistant, Universite Catholique de Louvain


Following the failure to adopt a horizontal Directive, proposed in 2001, on the conditions of entry and residence of third-country nationals (TCN) for the purpose of employment in the European Union (EU), EU labour migration policy is characterised by its fragmentation and sectoral approach. Given the sensitive nature of immigration policies and Member States’ attachment to their sovereignty, the only way forward was to adopt a limited number of legal instruments addressing the conditions of admission for few selected categories of economic migrants.

One of these categories is highly qualified workers. As they are deemed to be beneficial from an economic perspective, there is increasing competition among industrialised States to attract them. To that end, the European Commission proposed in 2007 to facilitate the admission of highly qualified workers and to grant them attractive residence conditions, as well as to create a common fast-track procedure. Two years later, the proposal led to the adoption of Directive 2009/50, known as the “Blue Card” Directive.

However, as discussed in the Commission’s 2014 report on the application of the Blue Card Directive, this Directive has not proven to be very effective, as shown by the limited number of permits (blue cards) delivered which is below expectations. While this may be explained by the subsistence of national schemes and the lack of publicity of the Blue Card, the conditions of entry and residence laid down in the Directive are arguably too restrictive.

For this reason, Jean-Claude Juncker declared his intention to review the Directive in order to enhance its attractiveness and overcome its intrinsic weaknesses. The Commission followed the desire of its President and the reform of the Blue Card Directive was part of the European agenda on migration of May 2015. Following four months’ public consultation on the future of the Blue Card Directive (results can be consulted here), the reform proposal was made public on 7 June 2016.

This blog post assesses the main developments included in the proposal and analyses whether it could be more effective in attracting talents and skills to Europe. First, a short review of the current Blue Card Directive is necessary to understand the extent of the proposed reform. Given the limited scope of this commentary, the 2009 Directive cannot be described at length but I will focus on its main characteristics.

It should be noted that the UK, Ireland and Denmark have opted out of the Directive. However, if the UK leaves the EU in future without any special arrangements on the movement of persons with the EU, the Directive will paradoxically become relevant to the UK nonetheless – since it would then regulate the admission of highly qualified British citizens to the remaining European Union.

The current Blue Card Directive and its limits

Although the Blue Card Directive aims at offering favourable admission and residence conditions to highly qualified workers, numerous intrinsic weaknesses have hindered its attractiveness. Labour migration being a sensitive issue, such weaknesses are to a large extent the result of Member States’ reluctance and dissension.

This is first exemplified by the determination of who qualifies for a Blue Card. According to Article 3 of the Directive, a highly qualified worker is someone who occupies a highly qualified employment, which is considered as requiring, either the successful completion of a post-secondary higher education programme lasting at least three years or, when provided by national law, at least five years of relevant professional experience. As a result, the very definition of who is a highly qualified worker is not uniform and may vary from one Member State to another.

Restrictive conditions of admission have made the Blue Card unpopular as they limit the number of potential applicants. For a Blue Card to be delivered, the TCN must have a valid work contract or a binding job offer if allowed by national law, and the prospective salary has to be at least 1.5 times the average gross national salary (meaning at least 51.466€ in Belgium). While the first condition means that a job must be secured from abroad (in-country application may be accepted in accordance with national law), the second criteria benefits large companies and senior positions.

In line with most Member States’ labour migration policy, the Blue Card Directive is based on a demand-driven entry system. As a result, it is no surprise that a TCN must have a valid work contract in order to apply for a Blue Card and the Directive does not provide for job-seeking permits. Also, the Directive allows Member States to conduct a labour market test which is a ground for refusal to deliver or renew a Blue Card during the first two years of employment (Article 8). This employer-led approach also justifies the fact that unemployment exceeding three consecutive months or occurring more than once during the period of validity of the Blue Card is a cause of withdrawal of the Blue Card (Article 13).

Furthermore, for the first two years of employment, Blue Card holders have a limited access to the labour market in the Member State concerned. Changing job is subject to prior authorisation, the new job must be highly qualified employment, and the salary condition applies (Article 12). Equal treatment with nationals, yet limited to access to highly quailed employment, may be granted after two years. As part of an approach based on the needs of Member States, the current Blue Card Directive suffers from significant shortcomings.

In contradiction with the internal (labour) market logic, the current Blue Card Directive provides for limited facilitation for intra-EU mobility, as a result of Member States’ dissension. TCN can only move to another Member State after 18 months and the Blue Card holder does not have a right to work in that second Member State. Because the TCN must apply for a Blue Card in that Member State, which may be lengthy (90 days at most), intra-EU mobility is subject to the fulfilment of the conditions imposed for first admission. For these reasons, intra-EU mobility is severely restricted even though TCN and highly qualified workers are usually more mobile compared to nationals and low-skilled workers.

While all these elements undoubtedly contribute to the unpopularity of the Blue Card, the most significant reason beyond its lack of success is the subsistence of national schemes for admitting the same category of highly qualified workers. This has resulted in parallel rules, conditions and procedures which precludes an EU-wide usage of the Blue Card system and limits its publicity. The limited success of the Blue Card does not mean that few highly qualified people have been admitted in Europe, the majority of them have been allowed under national schemes (24,922 out of 38,774 in 2014).

Even though the Blue Card Directive is an important instrument of the EU labour migration policy, it has had limited harmonisation effect because it only sets minimum standards and Member States retain a significant margin of discretion. Also, although the Directive grants a number of rights to highly qualified TCN, various restrictions are limiting the attractiveness of the Blue Card scheme, thus failing to supersede national schemes.

Now that the shortcomings of the current Directive have been exposed, let us examine how the Commission proposal intends to overcome them in order to meet the objectives that the Directive was meant to achieve.

The Commission proposal

Among the various options considered, the Commission chose to ease the admission conditions and make the Blue Card accessible to a wider group of highly skilled workers, while not extending the scope beyond highly skilled TCN. The proposal also intends to improve the rights associated with the Blue Card.

Firstly, the concept of “highly qualified employment” is replaced by that of “highly skilled employment” in order to include individuals who have completed the equivalent of a bachelor degree as well as those who have at least three years of relevant professional experience. The Commission also proposes to extend the scope of the Directive in order to include highly skilled beneficiaries of international protection. Recognised refugees already have access to the domestic labour market, but being a Blue Card holder would grant them rights associated with the Blue Card, including greater intra-EU mobility.

Secondly, the proposal clearly states that “Member States shall not issue any other permit than an EU Blue Card to third-country nationals for the purpose of highly skilled employment”. In hope to develop the Blue Card into a truly EU-wide scheme, all parallel domestic rules and procedures would be abandoned. Potential highly skilled TCN would have no choice but to apply for an EU Blue Card, if they wished to work in the EU. The EU would then have a genuinely EU-wide scheme but, as such, this would not make the EU more attractive. While this would be important in terms of visibility and clarity, it is likely to attract reluctance from Member States.

Thirdly, the salary condition would be lowered in order to be less restrictive and make the Blue Card more accessible. The salary threshold remains relative given the wide disparity among Member States and shall be in between 1.0 and 1.4 time the average gross salary in the Member State concerned. The maximum threshold would then be less than the current minimum. The proposal also provides for two exceptions for which the salary threshold shall be lower (80% of the above threshold). This would apply for professions suffering from shortage occupations as well as for young graduates.

Since the current salary condition is relatively high, the proposed threshold is likely to enhance the effectiveness of the Blue Card as it would be more inclusive. In particular, the exception in favour of young graduates, combined with the new Directive 2016/801 allowing students/researchers to stay at least nine months after the completion of their studies/research in order to seek employment (as discussed here), reinforces the attractiveness of the EU and make it easier for young graduates, who cannot claim high salaries, to apply for a Blue Card.  

Fourthly, the possibility to conduct a labour market test would be limited to exceptional circumstances such as a high level of unemployment in a given occupation or sector and justification is to be given to the Commission. As States’ oversight of the labour market is severely limited, a highly skilled TCN who meets the admission conditions, including a valid work contract, could not normally be refused access on the ground that another worker on the labour market is available. As a result, the Blue Card system would be more effective in attracting highly skilled TCN as it moves further away from a labour market adjustment rationale.  

Fifthly, labour market access would be significantly increased as the Blue Card holder is to be granted full access to highly skilled employment. The TCN would be allowed to freely change employer as long as it still qualifies as a highly skilled employment, even during the first two years of employment. However, this would not affect the possibility for Member States to withdraw or refuse to renew a Blue Card where conditions are not fulfilled, notably the salary criteria. The proposal also allows Blue Card holders to engage in self-employed activity, yet in parallel only. Since this goes towards more autonomy for TCN and greater equality with nationals, the attractiveness of the Blue Card is likely to be enhanced. 

Sixthly, Blue Card holders would benefit from facilitated access to the long-term resident status. If adopted as proposed, the new Blue Card Directive would derogate from Directive 2003/109 by granting long-term resident status after three years (not five) of legal and continuous residence within the territory of the Member States concerned. However, if the TCN becomes unemployed and does not have sufficient resources to maintain him/herself, the long-term resident status may be withdrawn before the usual five years’ time-limit is reached. Again, this is welcome as it enhances the TCN’s prospects of integration, which may be an important consideration when deciding on a country of destination.

Finally, the proposal wishes to reinforce the attractiveness of the EU by facilitating intra-EU mobility, in line with the desire to make the Blue Card a genuinely EU-wide scheme. The minimum residence period required before a Blue Card holder can move to another Member States is thus shortened to 12 months. While the TCN still needs to apply for a Blue Card in that second Member State, he or she would be allowed to work immediately after submitting an application (this would no longer be a possibility to be defined by national law). Also, Member States’ discretion is limited. Most notably, a labour market test would only be allowed if also in place for first entry applications, and no quotas would be allowed, contrary to the current situation.

Indeed, Member States’ right under Article 79(5) TFEU to determine the volumes of TCNs coming for the purpose of work is limited to TCNs coming directly from third-countries and does not apply in case of intra-EU mobility. As a result, intra-EU mobility would clearly be enhanced and TCNs would enjoy facilitated access to the labour market of other Member States. While this reinforces the impression that there is a single EU labour market, which is far from true, it remains to be seen whether highly skilled TCN actually move across the EU. Since the vast majority (around 90%) of Blue Cards are currently delivered by one Member State (Germany), figures on the mobility of highly qualified TCN is difficult to obtain.


Concerned about the underperformance of the Blue Card system launched in 2009, the Commission proposed a complete overhaul of this flagship policy in order to catch up in the competition among industrialised states to attract highly workers.

Overall, for the reasons explained above, the proposal appears to be relatively ambitious. Less restrictive admission criteria would make the Blue Card more inclusive and Member States’ leeway would be reduced, thus furthering harmonisation. Also, the limited possibility to undertake a labour market test means that labour migration is to be more than a labour market adjustment channel.

The system remains demand-driven, as potential candidates still need a work contract, but highly skilled labour is also praised as a source of human capital. This is illustrated by the fact that the nature of labour migration is meant to be less temporary than in the past. The Commission proposal intends to give Blue Card holders facilitated and quicker access to the long-term residence status. While this is arguably an element of attraction for potential migrants, this may be linked to long-term population objectives given Europe’s demographic trends and needs for human capital.

Despite these positive elements, the Commission proposal may prove too ambitious for Member States to approve, yet insufficient to effectively attract a significant number of highly skilled workers to the EU.

Although the Blue Card Directive needed to be reformed given its limited added value, the Commission proposal is arguably untimely. As we have witnessed with the current asylum crisis, there is increasing political resistance to developing common European rules related to migration, particularly when such rules imply a loss of sovereignty and control over entry rules. The Commission high level of ambition is therefore likely to attract resistance, especially since national schemes for highly skilled workers would no longer be allowed. As a result, the legislative process may prove to be long and difficult, despite the increasing recognition that skilled labour migration is beneficial to economic competitiveness.

Unlike countries such as Canada or Australia, the EU Member States do not face a high number of applications. The goal of the Commission is therefore to increase the attractiveness of the EU through migration policy. While a harmonised EU-wide scheme would enhance clarity and predictability for the benefit of both employers and potential candidates, the Blue Card system is only one element of attraction among others. As the public consultation tells us, the quality of life (including welfare and health care systems, wages, safety and the environment) makes the EU attractive, yet difficulties of getting a permit and the lack of integration perspective (openness to immigration, language, integration assistance,…) are unappealing factors.

The Commission proposal would arguably ease the issuance of a permit as the Blue Card would be accessible to a wider group of highly skilled workers, including young graduates. However, it is doubtful whether the goal of attracting more skilled labour to the EU would be met (the estimate of the Commission that at least 32,000 additional permits would be delivered under the new scheme seems quite optimistic). For instance, the liberal Swedish immigration system has not resulted in a sharp increase in the number of highly skilled workers. Therefore, while the Commission proposal is more inclusive than the current Directive, it is uncertain whether the goal of attracting more skilled labour would be reached without accompanying policies.

Barnard & Peers: chapter 26
JHA4: chapter I:6

Photo credit: