A Last-Minute Postscript: the CJEU finally dares to find that the NHS is a provider of ‘comprehensive sickness insurance’

 

Dr Sylvia de Mars, Newcastle University*

* With many thanks to Prof Charlotte O’Brien for all the idea-swapping that preceded this blog post.

 

Background

Does the NHS provide ‘comprehensive sickness insurance’ (CSI) cover? It is a small question, but one that has had significant consequences for a wide range of EU nationals who resided in the UK between at least 2011 and today.

The background to the question is something I have written about for over a decade at this point (‘Economically Inactive EU Migrants and the NHS: Unreasonable Burdens without Real Links?’ (2014) 39(6) ELRev 770; ‘Managing Misconceptions about EU Citizens’ Access to Domestic Public Healthcare: an EU-Level Response?’ (2019) 25(4) European Public Law 709): economically inactive EU citizens, under the rules for residency set out in Article 7(1)(b) Directive 2004/38 (the ‘Citizenship Directive’, or ‘CD’), only have a right to reside for longer than three months in a different Member State if:

‘They have sufficient resources for themselves and their family members not to become a burden on the social assistance system of the host Member State during their period of residence and have comprehensive sickness insurance cover in the host Member State’.

A similar rule, set out in Article 7(1)(c), informs students that they must have comprehensive sickness insurance in order to live in a host Member State for more than 3 months—although it does not require the same for their family members.

What is comprehensive sickness insurance? Like many key EU law terms, it is a generic catch-all intended to encompass all sorts of ‘healthcare’ provision in all the Member States, and it is not further defined anywhere.  The CJEU’s case law, clarifying most aspects of unhelpfully general EU law, has never addressed the concept directly – and so it has fallen to national courts to interpret it instead.

The UK here faced a particular question: under national law, everyone ordinarily resident in the UK is entitled to use the NHS for free. ‘Ordinary residence’ does not have a specific meaning in legislation in the UK, but according to settled case law (Shah v London Borough of Barnet [1983] 1 All ER 226), effectively just refers to the country where someone lives for a ‘settled purpose’—and this can be forward-looking, in the sense that if you move to the UK with the intention of settling, you are also ordinarily resident. In practice, this means that any EU national who moves to the UK with the intention of staying for more than three months will be ‘ordinarily resident’ and will be entitled to access the NHS.

Here is where the legal problem that the CJEU only tackled in 2022 arises. In the UK, taking what is in essence a purposive approach to Directive 2004/38, the courts have long insisted that CSI cover cannot mean being permitted to access the NHS. Sullivan LJ’s comments in Kamau (Kenya) [2010] EWCA Civ 1302 are illustrative:

‘A person who has to rely on the United Kingdom's National Health Service is no more self-sufficient than a person whose resources are inadequate so that he may become a burden on the United Kingdom's social assistance system."

This was also Home Office policy from 2008 onwards: applicants for documentation concerning the right to reside in the UK under EU law consequently had to either be employed, or had to demonstrate they had some sort of ‘comprehensive sickness insurance’ in the UK. The problem these applicants faced was that there is no ‘comprehensive sickness insurance’ cover in the UK; private insurance generally only covers secondary care, and frequently can only apply once someone has already lived in the UK for over 12 months. An EU national student moving to the UK in 2010, in other words, could as a matter of practical fact not actually prove to the Home Office that they had EU rights to reside – although in practice, the Home Office started accepting unusable and only somewhat comprehensive private health insurance policies as ‘evidence’ of CSI.

EU nationals themselves very often did not actually encounter the sting in the tail of this policy for years, as they themselves were not required to hold documentation confirming they had a right to reside in the UK: their passport did that for them. However, there were two situations where EU nationals would have concrete reasons to prove their residency status:

If they had a family member from outside of the EU, whose residency rights were dependent on their own satisfying of Article 7 of the Citizenship Directive; or

If they wanted to apply for UK nationality after holding permanent residency for one year, where the Home Office also required them to demonstrate that they had comprehensive sickness insurance during any periods of time when they were not employed.

Our EU student who started living in the UK in 2010, as such, might not have found they did not hold the documentation the Home Office required them to until 2016 – and their new ‘counter’ for lawful residence under the Directive, as interpreted by the Home Office, would start in 2016, if and when they got private health insurance.

Did nobody ever complain about this policy?

Immigration practitioners have been furious about the Home Office interpretation of the rules for as long as they have existed, both because they are impossible to satisfy, and because there is nothing in Directive 2004/38 that actually precludes NHS coverage from being CSI. And they took their concerns to the Commission—who in April 2012 confirmed that they were taking infringement proceedings under Article 258 TFEU against the UK by sending them a reasoned opinion about their implementation of the Citizenship Directive…but never followed through on that reasoned opinion, let alone with a reference to the CJEU.

I have long had suspicions as to why. It is a matter of fact that the Commission does not have to enforce breaches of EU law, and the CJEU has ruled out the idea that its use of Article 258 TFEU could be judicially challenged if it was clearly politically motivated… as I suspect it was here. The idea that the NHS was a ‘free for all’ for newly arrived EU nationals would have been politically toxic in the UK, and the idea of pushing such a claim forward with a Member State with rising Euroscepticism would have undoubtedly brought exactly those types of ‘benefit tourism’ headlines out in the UK. And while the Commission hesitated, or paused, the UK voted to leave the European Union—at which point, other matters (like ensuring EU nationals in the UK retained their rights) became a much higher priority than stopping the Home Office from demanding economically inactive EU nationals demonstrate they have health insurance that doesn’t exist in the UK.

Fast-Forward to 2022: VI v HMRC

The EU has avoided ruling on what the NHS ‘is’ for a long time, and not only with respect to comprehensive sickness insurance. Only in 2022 did it finally take an opportunity to make an assessment of how the NHS, as a service, should be treated, although the preliminary reference sent by the Social Security Appeal Tribunal (NI) would have let it avoid commenting on the NHS as ‘comprehensive sickness insurance’ (eg, see paragraph 31: a question as to whether the Common Travel Area (CTA) provides reciprocal healthcare coverage as suggested in Ahmad can be answered without actually addressing what the NHS ‘is’.).

The case in which a definition of ‘CSI’ came to the fore is an odd one, not least of all because it is on the basis of the known facts actually a case in which CSI should not have figured at all. The case concerned VI, a Pakistani national, who lived in Northern Ireland with her husband (also Pakistani) and her four children. At least one of the children has Irish nationality on account of having been born in Northern Ireland in 2004.

The parties were agreed that VI had ‘sufficient resources’ and thus met that requirement of Article 7 CD, and they also agreed that between August 2006 and August 2014, VI and her entire family had comprehensive sickness insurance.  As that is a period longer than 5 years, the Irish national son was also agreed to have attained permanent residence (presumably by 2011).

However, HMRC found that VI was not entitled to receive Child Tax Credit or Child Benefit for the periods in which she requested it and in which she was not covered by comprehensive sickness insurance. This, in total, was approximately 3 months in 2006 and 1 month in 2014. HMRC also appears to have conceded that she does not have to repay the benefits she received in those months: it was their error, in that she did not misrepresent her status and was clear about when she did and did not hold ‘comprehensive sickness’ insurance.

Jurisdiction

Before it addressed the referred questions, the CJEU clarified its jurisdiction in the case: it could answer this preliminary reference request despite the fact that it was submitted on 7 April 2020.  First, Article 127 of the withdrawal agreement (WA) makes clear that during the so-called ‘transition period’, EU law would continue to operate in the UK as it did when it was a Member State; and secondly, Article 86(2) WA made clear that the CJEU would retain the jurisdiction to answer preliminary references submitted before the end of the transition period. This was not contentious – but probably worth stressing all the same, given the timing of the ruling.

Referred Questions

The Social Security Appeal Tribunal (NI) sent three questions to the CJEU.

First, it was asked if permanently resident children have to maintain CSI in order to maintain a right to reside. The CJEU was quick on this, pointing out that Article 16(1) CD makes clear that the conditions in Article 7 do not apply to EU national who has permanent residence [54]; and Article 16(2) CD and Article 21 TFEU as interpreted by the CJEU extends the same to ‘family members’ of EU nationals [56], even where they are not ‘dependent family members’ in the ascending line [57-58]. From 2011 onwards, in other words, neither VI nor her son would have needed comprehensive sickness insurance in order to maintain their right to reside in Northern Ireland.

The second question referred asked if the CSI requirement actually applied to a child covered by Article 7 CD, and their third country parent that is taking care of them. The CJEU here indicates that non-English versions of the Directive make it clear that the requirement for comprehensive sickness insurance applies to both the EU national and their family members, as defined by the Directive and by CJEU case law on parents who are caretakers for their EU national children [such as Chen]. That requirement, per the Directive, is there for the entirety of the 5 year residence period prior to an EU national attaining permanent residence—and the CJEU makes it clear that it does not matter if the CSI is held by the child, covering the parent, or held by the parent, and covering the child [67].

Here is where the CJEU, without being expressly prompted, comments on the NHS:

68      In the present case, it is apparent from the documents before the Court that VI and her son were affiliated during the period in question, namely from 1 May 2006 to 20 August 2006, to the United Kingdom’s public sickness insurance system offered free of charge by the National Health Service.

69      In that regard, it must be recalled that, although the host Member State may, subject to compliance with the principle of proportionality, make affiliation to its public sickness insurance system of an economically inactive Union citizen, residing in its territory on the basis of Article 7(1)(b) of Directive 2004/38, subject to conditions intended to ensure that that citizen does not become an unreasonable burden on the public finances of that Member State, such as the conclusion or maintaining, by that citizen, of comprehensive private sickness insurance enabling the reimbursement to that Member State of the health expenses it has incurred for that citizen’s benefit, or the payment, by that citizen, of a contribution to that Member State’s public sickness insurance system (judgment of 15 July 2021, A (Public health care), C‑535/19, EU:C:2021:595, paragraph 59), the fact remains that, once a Union citizen is affiliated to such a public sickness insurance system in the host Member State, he or she has comprehensive sickness insurance within the meaning of Article 7(1)(b).

70      Furthermore, in a situation, such as that in the main proceedings, in which the economically inactive Union citizen at issue is a child, one of whose parents, a third-country national, has worked and was subject to tax in the host State during the period at issue, it would be disproportionate to deny that child and the parent who is his or her primary carer a right of residence, under Article 7(1)(b) of Directive 2004/38, on the sole ground that, during that period, they were affiliated free of charge to the public sickness insurance system of that State. It cannot be considered that that affiliation free of charge constitutes, in such circumstances, an unreasonable burden on the public finances of that State.

The CJEU then rules out that VI and her son find themselves in a Texeira situation, where EU nationals parents retain a right to reside after having been workers because their children are education—and confirms that as a consequence, VI and her son both needed to be covered by comprehensive sickness insurance.

Finally, the CJEU addresses a particular ‘twist’ on CSI cover that the UK courts have ruled on, which is that if the UK has reciprocal healthcare arrangements with a Member State, and so the UK can ‘reclaim’ healthcare costs from that Member State if one of its nationals uses the NHS, that satisfies the requirement for CSI. In practice, this UK means that anyone receiving treatment under the European Health Insurance Card (EHIC), intended to cover only temporary stays abroad, is deemed by the Home Office to have comprehensive sickness insurance. The CJEU declined to decide if the UK’s arrangements with Ireland under the Common Travel Area made for a ‘reciprocal arrangement’ that amounted to CSI, simply because the referred questions did not actually clarify what those arrangements were [73-74].

This, of course, would not matter greatly to the referring tribunal, because the CJEU’s interpretation of comprehensive sickness insurance is enough to decide VI’s case: if NHS cover is comprehensive sickness insurance, VI was entitled to Child Tax Credit and Child Benefit throughout her residence in the UK, regardless of whether she also held private insurance.

Comments

General

A first observation is that it is doubtful that VI is a case that actually should have produced a definition of CSI in relation to NHS entitlement. This is nothing to do with Brexit, but rather to do with the facts of the case itself. The point at which HMRC effectively conceded that they were not going to reclaim the benefits paid out, and the error was theirs, it is not clear what ‘dispute’ remained between VI and HMRC. The CJEU has interpreted the references as being primarily about VI’s ‘right to reside’ during the relevant periods where she did not have CSI (see, eg, para 29, where the CJEU notes that the parties disagree on VI’s entitlement to Child Tax Credit and Child Benefit – but rewords that into a question on the ‘subject matter’ of ‘VI’s right of residence in the UK’), but obviously the actual dispute between VI and HMRC was about the benefits she was given during those periods; a dispute about VI’s right to reside in the absence of the benefits, after all, would have been with the Home Office, not HMRC.

It thus looks like the case was constructed very carefully to avoid falling into the realm of ‘hypothetical’ rulings that the CJEU has consistently refused to consider because the dispute that resulted in the referred question had been resolved by the time of its ruling (Case C-225/02 García Blanco). Add to this the fact that the preliminary reference questions sent did not actually require the CJEU to offer an opinion on whether NHS entitlement equated to CSI, and we can only conclude that the CJEU was determined to make this case ‘work’ for it.

Having said that, there are a number of reasons to wonder if VI is actually going to mean anything for anyone—at least in the UK.

The highlighted passages in VI about CSI may have an impact in other Member States that operate ‘public health services’ like the NHS: to the best of my knowledge, these do not currently require economically inactive EU nationals to contribute in order to gain access, but the VI judgment clearly enables them to set up something akin to the Immigration Health Surcharge that the UK charges immigrants for economically inactive EU nationals. It reemphasizes the distinctions in terms of EU law-based rights and duties between those who ‘integrate’ and those who do not, as the CJEU has been doing ever since its Dano judgment—though it does also stress, in paragraph 70, that any requirement for ‘paying into’ a national health service has to be proportionate. In the VI case, VI’s husband was a full-time employee who paid taxes much as any employed UK national, so ‘double-charging’ would have been disproportionate. A family like VI’s clearly does fit the ‘integration’ demands the CJEU has recently placed on EU nationals who move to other Member States: in the CJEU’s own words, they do not form an unreasonable burden. Given the requirement for ‘proportionate’ charges to EU nationals for use of a public healthcare system, the impact in other Member States may be quite limited in practice; it is not immediately obvious that the cost of administering such a system is outweighed by the actual charges that would be recovered.

VI and Brexit?

But what of in the UK? As far as VI herself goes, apparently HMRC are not asking for the money back—so regardless of what the CJEU said, it appears that her dispute with HMRC is for all intents and purposes done. Her main victory here is that a court that the UK wanted rid of at all costs as part of the Brexit process has, in essence, told her she was right.

What is far less clear is what happens with this CJEU ruling in practice in the UK, where problems with demands for CSI have affected hundreds of thousands of EU nationals since at least 2011. If the UK were still an EU Member State, the follow-up is obvious: anyone whose application for permanent residency or—the far more costly—UK citizenship was rejected because of the absence of CSI when they were entitled to NHS care would have a cause for damages. The effects of a status being declined for some will have meant that they took out (also expensive) private insurance they could not use and waited six years—but for others will have meant separation from family members, insecurity in terms of access to benefits because they did not have permanent residency in the UK, and far greater expenses. These appear to be losses caused by what as of 10 March 2022 we know are breaches of EU law: applicants would have been entitled to permanent residency as a matter of EU law, and the UK’s failure to grant them evidence of this status, and the attached benefits, caused identifiable losses. Were we still a Member State, then, a Francovich claim for state liability seems like the logical follow-up to VI.

But we are not, and Schedule 1(4) of the EU (Withdrawal) Act 2018 generally rules out the possibility of Francovich damages via domestic court proceedings after the transition period in the withdrawal agreement:

There is no right in domestic law on or after [F1IP completion day] to damages in accordance with the rule in Francovich.

However, hidden in paragraph 39(7) of Schedule 8 of the EU (Withdrawal) Act 2018 is an exception that covers rights held and violated before Brexit:

Paragraph 4 of Schedule 1 does not apply in relation to any proceedings begun within the period of two years beginning with exit day so far as the proceedings relate to anything which occurred before exit day.

VI falls within this exception, and so Francovich damages for historic mistreatment of EU nationals seeking confirmation of permanent residency and being told they could not have it because they were reliant on the NHS are a definitive possibility, provided they are started before the end of 2022. Historic wrongs here can thus still be righted, provided claims are made quickly enough.

What happens if they are not, and EU nationals want to make a claim after 2022?

Schedule 1 of the EU (Withdrawal) Agreement 2018 is itself very clear – but what is not clear is whether this domestic law provision survives the UK’s obligations under the Withdrawal Agreement itself as implemented in UK law. Section 5(7) of the EU (Withdrawal) Act 2018 thus indicates that Schedule 1 is “subject” to so-called “relevant separation agreement law”, and section 7(C) of the same Act makes clear that “relevant separation agreement law” includes the Withdrawal Agreement. If state liability is an EU law doctrine that forms part of the Withdrawal Agreement, then, it would take precedence over the UK’s abolition of the doctrine as a matter of domestic law.

Is Francovich part of the Withdrawal Agreement? On a purely practical level, it’s not clear what the point of Articles 86-89 WA is in the absence of an ability to provide remedies for breaches of EU law. Article 86 WA gives the CJEU the jurisdiction to hear this reference post-transition, and Article 89 WA makes clear that the ruling in VI is still binding on the entirety of the UK:

1.Judgments and orders of the Court of Justice of the European Union handed down before the end of the transition period, as well as such judgments and orders handed down after the end of the transition period in proceedings referred to in Articles 86 and 87, shall have binding force in their entirety on and in the United Kingdom.

2.If, in a judgment referred to in paragraph 1, the Court of Justice of the European Union finds that the United Kingdom has failed to fulfil an obligation under the Treaties or this Agreement, the United Kingdom shall take the necessary measures to comply with that judgment.

If a CJEU judgment would not have the same effects in the UK (where it is binding in its entirety) as it would in the Member States, it is difficult to see how the ability for the CJEU to rule on matters after 2022 has much meaning: an obligation to comply with CJEU rulings on how to apply EU law applicable in the UK when it was a Member State is effectively moot, as the same EU law is not applicable in the UK anymore now. For the CJEU’s declaratory judgments to have any practical effects, they would need to be able to result in remedies to be obtainable for those who experienced losses because of historic, pre-Brexit breaches of EU law.

Important here is that that Article 4(1) WA declares the entire Withdrawal Agreement to be directly effective where it is clear, precise and unconditional, and Article 4(3) and 4(4) WA requires that “Union law or concepts or provisions thereof” have to be interpreted and applied in accordance with both methods, general principles, and pre-Brexit CJEU case law.

In 2023, therefore, an individual could argue before a UK court that Article 89 WA is directly effective—and that part of the UK taking the ‘necessary measures’ to comply with the judgment means that UK courts should be able to consider claims for state liability, as that is part of the EU law’s normal enforcement repertoire (as established by CJEU case law). Whether a UK court would accept this logic, however, is debatable; while the Withdrawal Agreement as a whole is supreme to any conflicting domestic law, and the EU (Withdrawal) Act 2018 acknowledges as much for domestic law purposes, it is a genuine matter of interpretation if Schedule 1(4) of the EU (Withdrawal) Agreement 2018 is in conflict with the Withdrawal Agreement. What would happen with any attempt to ‘Francovich’ the VI ruling after 2022, in other words, is anyone’s guess.

In the absence of an EU law remedy, anyone affected by what is over a decade of misapplication of Article 7 CD but unable to act on that via a case before 2023 looks out of luck. Judicial review applications have a time limit of 3 months maximum, and even the most generous application of that in the case of VI would have the clock starting on 10 March 2022. An argument that links VI to the withdrawal agreement's general 'EU law' effects might thus look like a long shot, but by 2023, it will be the only possible shot. 

In terms of ongoing UK policy that VI affects, we also have to look at what the withdrawal agreement applies, and requires, in terms of UK interpretation and application of that EU law. The UK has left the EU, of course, but the Citizenship Directive remains a key component of the Withdrawal Agreement’s provisions on Citizens’ Rights. All EU nationals resident in the UK on the basis of the Withdrawal Agreement are in effect still here under the rules set out in the Citizenship Directive. A key difference between now and then, however, is that as a policy choice, the UK government chose not to require economically inactive EU nationals who wished to apply for ‘Settled Status’ under the Withdrawal Agreement to hold comprehensive sickness insurance: five years of residency in the UK while having sufficient resources was all that was required.

If this seems generous, the sting is once more in the tail, because these EU nationals with ‘settled status’ nonetheless need to demonstrate they held CSI during any periods they resided in the UK under EU law (rather than the ‘Settled Status’ scheme) – so, any time before they applied for either ‘Settled Status’ or ‘Pre-Settled Status’. There are thus inevitably a number of EU nationals in the UK who have lived here for far longer than six years, but who nonetheless are not able to apply for UK citizenship because of an administrative requirement covering their pre-Brexit residency that the CJEU has now declared incompatible with EU law.

Can VI help them? Opinions are bound to differ here as well. One possible argument for the UK government to make here is that citizenship law is outside of EU competences in a very general sense, and so the UK asking EU nationals to have CSI is simply not something that the Withdrawal Agreement addresses. It might as well ask EU nationals to dance the hokey cokey: that, too, is outside of what the EU can control.

However, it is also arguable that ‘CSI cover’, as the UK demands it of EU nationals in order to naturalise, is an EU law concept that is covered by the Withdrawal Agreement, and therefore must be interpreted in line with the CJEU’s case law—bringing this particular dimension of the administrative dimension to UK nationality law within the scope of EU law. If the UK government were insistent that EU nationals could only naturalise if they had held some sort of non-NHS health insurance, the fix would be simple: just call it something else but require that EU nationals held it anyway, as retrospective (or ex post facto) legislative activity is perfectly possible under UK law. This would mean that those wanting UK citizenship would simply have to get whatever non-NHS comprehensive sickness insurance cover is, as so many before them did, and wait several more years in order to put in their application.

VI and Northern Ireland?

One further final ‘Brexit’ dimension that has to be considered in light of VI is how the judgment will work not in the UK generally, but specifically in Northern Ireland. As has been set out in a great many pieces of academic writing and commentary since 2015, the ‘Brexit’ that Northern Ireland is experiencing is distinct from the one that Great Britain has achieved, and there are particular dimensions of the Protocol on Ireland/Northern Ireland (Protocol) that remain unclear but are worth exploring in light of this judgment.

Key is Article 2 of the Protocol, of which the relevant parts read:

The United Kingdom shall ensure that no diminution of rights, safeguards or equality of opportunity, as set out in that part of the 1998 Agreement entitled Rights, Safeguards and Equality of Opportunity results from its withdrawal from the Union, including in the area of protection against discrimination, as enshrined in the provisions of Union law listed in Annex 1 to this Protocol, and shall implement this paragraph through dedicated mechanisms.

There has been significant debate as to the scope of this ‘no diminution’ commitment. In terms of concrete ‘rights’ protected, the evidence to the Lords Committee on the Protocol has suggested that there are no clear limitations to the GFA concepts of ‘right, safeguard or equality of opportunity’— the GFA at best sets out a non-exhaustive list of examples.

Of course, Article 2 of the Protocol only applies to a diminution of rights that followed from Brexit, and so the likely way in which any cause for action on the basis of VI is linked to Article 2 is via remedies. Given that the EU (Withdrawal) Act 2018 will end all possibility of state liability claims in relation to pre-Brexit breaches of EU made law by the UK at the end of 2022, it appears clear that there will be a diminution of rights here (eg, right to an effective remedy) and that that diminution has been a direct consequence of Brexit (many thanks to Dr. Eleni Frantziou for pointing this out).  

This consequently would suggest that while EU nationals living in Great Britain who were denied, for example, benefits or citizenship because of the CSI requirement, may fall between the cracks created by Brexit, in the absence of a ‘willing’ judiciary, EU nationals living in Northern Ireland look like they might have a further shot at a successful claim even if they do not 'get it in' before the end of 2022. Their argument would be that a violation of EU law pre-Brexit should result in undiminished state liability for the UK post-Brexit because of Article 2 of the Protocol. But there, too, a judiciary looking to give Article 2 of the Protocol a broad meaning is a prerequisite for a result, and so it is far from obvious that anyone in the UK will experience the effects that VI would have had if the UK were still a Member State unless they get in a claim for damages before the end of this year.

Conclusion

The Commission was too worried to pursue the matter in 2012, and the CJEU seemed to feel confident enough to rule on what the NHS actually does only in 2022, after the UK left the EU, and when the ruling would probably not cause a shockwave anymore. That assessment might be slightly misplaced, however. If the above interpretation of the effects of VI in the post-Brexit UK legal landscape is correct,  the CJEU may have done just enough to still give EU nationals a surefire remedy for the wrongs they experienced for the wrongs they experienced if they leap on this judgment - and at least several routes by which they can try to attain a remedy beyond 2022. Given the general politics surrounding Brexit even now that it is supposedly 'done', one can only imagine how this will go down in Westminster. 

Photo credit: Matt Brown, via Wikimedia Commons