The intended right and the breached obligation are not necessarily one and the same

Our first point concerns the idea – which is easily and often overlooked – that the ‘right’ which provides the basis for an individual’s claim under Francovich is not necessarily the same as the ‘obligation’ whose breach provides the basis for the Member State’s eventual liability. To be fair, in most cases, the standard formulation used by the Court clearly suggests that the relevant ‘right’ and ‘obligation’ are merely two sides of the same coin. As originally expressed in Brasserie de Pêcheur and repeated countless times since, the rule of law infringed must be intended to confer rights.Footnote ⁵ And true enough: in most cases, there is a perfect concurrence of identity between the right relied upon by the claimant and the obligation breached by the Member State. Consider (for example) a Francovich action based upon a straightforward infringement of the right to free movement of goods or services though the erection of an unlawful barrier to trade by the national legislative or executive authorities.Footnote ⁶

But the situation is not always just so. In certain cases, there is a distinction between (on the one hand) the substantive right which the individual was intended to enjoy under EU law and which entitles them to seek reparation under Francovich; and (on the other hand) the substantive obligation which the Member State has been found to breach and which leads them to incur liability under EU law. Indeed, it is quite possible that the intended right which acts as the initial trigger for seeking reparation is not addressed to the Member State at all, while the obligation whose infringement serves to bring the Member State’s liability to fruition need not be directly concerned with the protection of any individual.

The point is most obvious when it comes to non-contractual liability claims against the Member State based upon the latter’s failure to transpose a directive correctly or indeed at all. Consider the situation in Francovich itself.Footnote ⁷ The substantive right intended under EU law was for the employees of an insolvent undertaking to be paid certain of their outstanding wages through a guarantee fund created pursuant to Directive 80/987.Footnote ⁸ But the actual obligation breached by Italy consisted in the duty under Article 288 TFEU to implement Directive 80/987 into national law within the applicable deadline. The latter obligation is addressed solely to the Member States and can hardly be described as intended to confer rights on individuals in any direct or specific way; at any rate, the Member State’s obligation to transpose under Article 288 TFEU is entirely distinct from the right to payment of outstanding wages as provided for under the substantive provisions of Directive 80/987.

This distinction between ‘right’ and ‘obligation’ was in fact recognised in the original Francovich formulation of the three conditions required to incur liability: the result prescribed by the relevant directive should entail the grant of rights to individuals; it should be possible to identify the content of those rights on the basis of the provisions of the relevant directive; and finally, there should be a causal link between breach of the Member State’s obligation to transpose under Article 288 TFEU (on the one hand) and the harm suffered by the claimant (on the other hand).Footnote ⁹ It was only in the subsequent Brasserie ruling that the Court conflated the idea of the rule of law which had been infringed by the Member State, with the rule of law which was intended to confer rights upon the claimant.Footnote ¹⁰ That was perhaps unsurprising, given the factual and legal situation at stake in Brasserie itself.Footnote ¹¹ It is interesting to note that in Dillenkofer, where the Court’s primary concern was to reconcile the apparent differences which had emerged between the Francovich and Brasserie formulae into a single unified test for incurring liability, the judgment was still careful to separate explicitly the right which EU law had intended to confer upon the claimant, from the Member State’s obligation to transpose the relevant directive into national law.Footnote ¹² But even if the more nuanced and accurate approach from Francovich and Dillenkofer still emerges occasionally in a few other judgments,Footnote ¹³ the case law has largely adopted the Brasserie terminology as a matter of course, even in disputes based upon the non- or incorrect transposition of EU directives into national law.Footnote ¹⁴

However, it is not only in disputes involving the non- or incorrect transposition of EU directives where it is possible to separate the claimant’s intended rights from the Member State’s breached obligations. Consider also a case like Köbler.Footnote ¹⁵ As is well known, the dispute centred around the conduct of a national court of last instance which had failed to comply with its obligation to make (or rather, to maintain) a preliminary reference to the European Court of Justice under Article 267 TFEU. Austria objected to the possibility of liability on the grounds that that provision was an instrument of judicial cooperation between the domestic and EU courts and was not intended to confer any rights upon individuals.Footnote ¹⁶ In order to counter that argument, the Court held that the substantive right intended under EU law, which provided the trigger for the claimant’s action, actually consisted in the right of migrant workers to equal treatment with home workers as regards their terms and conditions of employment in accordance with Article 45 TFEU. When it came to the obligation which the Member State was alleged to have infringed, the Court likewise focused on the Austrian Supreme Court’s failure to uphold the claimant’s right to equal treatment on grounds of nationality as regards their terms and conditions of university employment. But for the purpose of determining whether that infringement could be considered ‘sufficiently serious’, the main factor to be taken into consideration was the contravention by the Austrian Supreme Court of its obligations relating to the preliminary reference procedure under Article 267 TFEU.

Viewed from a perspective which treats the relevant ‘right’ and ‘obligation’ as interchangeable, one can sympathise with Court’s dilemma in Köbler. Article 267 TFEU was undoubtedly central to Austria’s infringement of EU law in this dispute, for the purposes of identifying a ‘sufficiently serious breach’. But to have then also identified Article 267 TFEU as the basis for the claimant’s entitlement to sue, for the purposes of the ‘intention to confer rights’, would have risked either ruling out any possibility of compensation from the outset (because of the consistent case law insisting that Article 267 TFEU is merely an instrument of judicial cooperation); or instead transforming the very nature of Article 267 TFEU within the EU legal order (i.e. precisely into an explicit source of individual rights). The Court’s solution to that dilemma was to identify equal treatment on grounds of nationality as both the intended right for the claimant and the incriminating obligation of the Member State.

Yet viewed from a perspective which separates the qualifying ‘right’ from the relevant ‘obligation’, this solution appears both artificial and unnecessary. It feels artificial because the substantive right to equal treatment, and the corresponding obligation to provide it, was addressed primarily to the authorities responsible for determining the terms and conditions of employment for university staff; that right/obligation was addressed to the national judiciary only at such a high level of abstraction as one could say the same of any legal right capable of being vindicated through judicial process. But more importantly, the Court’s preferred solution feels unnecessary because it would have been entirely proper in Köbler to identify the right to equal treatment under Article 45 TFEU as the claimant’s intended right for the purposes of Francovich; then to treat the national court’s duty to make (maintain) a preliminary reference under Article 267 TFEU as the Member State’s relevant obligation for the purposes of assessing the ‘sufficiently serious breach’ requirement. Just as with the obligation to transpose directives into national law under Article 288 TFEU, so too it is immaterial whether the duty to request preliminary guidance from the Court under Article 267 TFEU is itself intended to confer any rights on individuals.

Be that as it may, and even taking the Court’s reasoning on its own terms, Köbler remains a useful illustration of our underlying point of principle: within the legal structures of Francovich, the right which initially qualifies the claimant to seek compensation, and the obligations which are relevant to assessing the Member State’s eventual culpability, are not necessarily one and the same.

When it comes to the basic legal structures of the Francovich action, our discussion so far therefore suggests that liability arises from the combination of two distinct yet closely inter-related criteria: the existence of an intended right for the claimant; and the determination of a culpable breach of its obligations by the Member State. Those two conditions will often refer to precisely the same provisions of EU law, but that need not necessarily or at least entirely be the case. It is more important that the two requirements stand in a sufficiently proximate relationship to each other – an assessment which can for present purposes be adequately expressed through a ‘but for’ test: but for Italy’s breach of its obligation correctly to implement Directive 80/987 within the applicable deadline, the claimants in Francovich may have enjoyed their otherwise distinct right to unpaid wages; but for Austria’s breach of its obligation to request preliminary guidance under Article 267 TFEU, the claimant in Köbler may have benefited from his otherwise distinct right to equal treatment.Footnote ¹⁷

It is that interaction between two conceptually separable criteria which provides the basis for something greater than the sum of their individual parts: a new and autonomous right to reparation under EU law, derived from but independent of both the original intended right and the original breached obligation.Footnote ¹⁸ Moreover, this methodology is essential if Francovich is to fulfil its function of compensating individuals for the losses they suffer as a result of the Member State’s infringement of its Treaty obligations. After all, if the Court were to insist that the intended right and the breached obligation were indeed the direct converse of each other, that would lead to an excessively restrictive liability system – whereby claimants in a position analogous to Francovich or Köbler could have no chance of claiming reparation in respect of their losses, despite a clear line of responsibility connecting their intended rights to the Member State’s default.

The prominent role allotted to national procedural autonomy within the Francovich action

The second important point to note about the basic legal framework of the Francovich system is that the autonomous core of the right to reparation, generated through the combination of an intended right and a culpable breach, is then supplemented by rules which are heavily conditioned by the influence of national law.Footnote ¹⁹ Consider the requirement of a direct causal link between the Member State’s breach and the harm suffered by the claimant – the third criterion laid down directly by EU law in order to establish the Member State’s liability to compensate. The case law has proven ambiguous about whether the national court was merely expected to apply a concept created and defined at the level of EU law, albeit that the detailed content of that concept still remained to be spelt out more fully by the Court in future rulings; or whether instead the national courts were free to apply their own domestic rules on causation, subject only to scrutiny by the Court itself to ensure that those domestic rules met certain minimum criteria expected under EU law.Footnote ²⁰ However, the Court in Danfoss explicitly stated that the applicable rules of causation are indeed determined in the first instance by national law, subject to compliance with the principles of equivalence and effectiveness.Footnote ²¹

The same approach – a presumption of national autonomy, rebuttable on the basis of the principles of equivalence and effectiveness, as well as other mandatory requirements imposed by EU law, such as the right to effective judicial protection under Article 47 of the CharterFootnote ²² – applies to all other consequential aspects of the right to reparation once the substantive conditions required to establish liability have been confirmed. That includes: the extent of recoverable losses and, in particular, the precise heads of damage available;Footnote ²³ the very nature of the reparation offered by the Member State (for example, through retroactive performance of its obligations, with damages provided only in respect of supplementary losses);Footnote ²⁴ the imposition of a duty to mitigate one’s own losses, including the requirement to exhaust adequate alternative remedies;Footnote ²⁵ the internal distribution of liability to make compensation between the various public authorities of each Member State;Footnote ²⁶ and more technical procedural issues such as the limitation periods which apply to exercising the right to reparation,Footnote ²⁷ or the availability of legal aid for Francovich claimants.Footnote ²⁸

The ‘intention to confer rights’ criterion as the natural forum for analysing protective scope issues

One might naturally assume that the ‘intention to confer rights’ requirement furnishes the primary forum for exploring those protective scope issues within the Francovich right to reparation. After all, the ‘intention to confer rights’ acts as the initial gateway into liability, through an explicit concern to identify the entitlement of any given claimant to seek redress under EU law – allowing the ‘sufficiently serious breach’ criterion then to concentrate on the extent of the Member State’s culpability in breaching its EU law obligations (which as we know might well be distinct from the claimant’s putative rights) and leaving all other issues, relevant especially to exercise of the right to reparation, to be governed by the presumption of national procedural autonomy (subject to the oversight of EU law).

True enough, various judgments do indeed engage directly with the ‘intention to confer rights’ criterion – allowing us to understand how the sorts of protective scope issues raised by the Francovich action have in practice been handled by the Court. For example, when it comes to the question of whether the claimant enjoys any EU law right which deserves in principle to be protected, the case law usefully reveals some of the methodological tools at the Court’s disposal. Consider cases like Dillenkofer and Rechberger: the Court specifically enquired whether the relevant consumer protection directive entails the grant of rights to individuals, by examining its preamble and objectives, as well as the nature of the duties the directive seeks to impose upon undertakings – treating the implicit creation of specific consumer rights as a direct corollary of the formulation under EU law of various precise commercial obligations.Footnote ²⁹ Consider also the leading case of Peter Paul: through a detailed examination of the relevant directives, the Court concluded that EU law was not intended to confer rights upon individual depositors in the event of the defective supervision of credit institutions by the responsible national authorities – taking into account, for example, the lack of any express provision in EU law conferring rights on depositors; the fact that such rights were not necessary to achieve the underlying aim of EU harmonisation, which was limited to securing mutual recognition as regards the authorisation and supervision of credit institutions; the obligation of the national authorities to take into account a plurality of interests, including the overall stability of the financial system; and the fact that EU law had created a dedicated deposit guarantee scheme which offered a degree of protection for individuals, including as regards problems arising from defective regulatory supervision.Footnote ³⁰

Or again: when it comes to the question of defining the precise content of the claimant’s EU law right, the case law illustrates the Court’s capacity to identify and address pertinent protective scope issues.Footnote ³¹ Consider rulings that deal with the personal scope of an EU law right, that is, whether the claimant falls within an identifiable class of intended beneficiaries: disputes such as Dillenkofer show the Court referring to the definition of ‘consumers’ under the relevant EU directive, for the purposes of verifying entitlement of seek Francovich damages against the Member State for non-transposition.Footnote ³² Consider also judgments that concern the material scope of an EU law right. In particular, the Court has on several occasions explored whether it is possible to identify with sufficient precision the minimum content of an intended right from the relevant provisions of EU law; if so, the fact that there may be some uncertainty about other aspects of the individual’s entitlements, arising from the conferral of limited discretionary choices upon the Member State, will not be fatal to a Francovich claim.Footnote ³³

Not only should we see the ‘intention to confer rights’ as a natural forum for the initial discussion of protective scope issues. One might also assume that how the Court chooses at the outset to define the nature and extent of the claimant’s ‘intended right’ should play a potentially crucial role in shaping the subsequent character of the entire Francovich assessment: for example, influencing the range of factors that may be relevant to whether the breach was ‘sufficiently serious’; determining the circumstances to be taken into account when calculating the direct causal link; or conditioning the proper range of losses that should be considered potentially recoverable.

True enough, there is again evidence in the case law of the Court considering how the subsequent legal structures of the Francovich action should be informed by and adjusted to the sorts of protective scope issues dealt with under the initial ‘intention to confer rights’ criterion. A prime example is the ruling in Leth, which we will return to consider in greater detail below.Footnote ³⁴ Suffice for now to observe how the Court here decided that the Environmental Impact Assessment Directive is intended to generate an individual right to have the environmental effects of qualifying projects assessed by the competent authorities and to protect against pecuniary damage which is the direct economic consequence of such effects.Footnote ³⁵ But what the Court gave with one hand, it promptly took away with the other: for the subsequent purpose of verifying the direct causal link necessary under a successful Francovich claim, the Court stressed that the Member State’s failure to conduct an obligatory assessment could not of itself be held responsible for any decrease in the value of the claimant’s property arising from the environmental effects of the disputed project, since the purpose of the Directive is only to require an assessment rather than to determine its outcome. In other words: the specific nature of the claimant’s intended right as defined by the Court had a decisive impact upon the nature of the causal link which the national courts were entitled to expect in order to vest a fully-fledged right to reparation under EU law. The same sort of joined-up thinking is evident in other contexts. For example, if the Member State choses to make reparation through the retroactive application of the relevant EU legislation, that can include the exercise of any discretionary powers that entitle the Member State to limit the scope or content of the claimant’s intended rights, though it is without prejudice to payment of compensation in respect of supplementary losses incurred by the claimant which are specifically protected under EU law but not remedied by retroactive compliance alone.Footnote ³⁶

The Court’s tickbox approach to the ‘intention to confer rights’ criterion and its potential implications

Notwithstanding such useful insights into the Court’s understanding of and approach towards the ‘intention to confer rights’ requirement, it is evident that – in the great majority of Francovich cases referred to Luxembourg – this criterion is treated as a relatively straightforward hurdle for the claimant to surmount. Indeed, when it comes to most of the decided cases on Francovich, protective scope issues are not the subject of any more specific or detailed judicial discussion at the stage of the ‘intention to confer rights’ assessment.Footnote ³⁷ Is this because the ‘intention to confer rights’ and with it the protective scope issues relevant to a Francovich claim are genuinely unproblematic in most disputes? Or is it possible that the Court’s treatment of the ‘intention to confer rights’ is not always so probing of potential protective scope issues as one might expect?

To some degree, the answer may well lie in a simple change of judicial linguistics. After all, the original formulation of the criteria for establishing liability, as set down in Francovich itself, were relatively explicit about asking whether the claimant could establish the existence of an EU law right in principle, then proceeding to query the more precise personal and material content of that EU law right in practice. By contrast, the alternative formulation which emerged in Brasserie and became generalised after Dillenkofer, is much less explicit or demanding – at least on its face – than the original Francovich approach: the Court seems to ask quite simply, does EU law intend to confer rights upon the claimant? Issues such as how to define the precise personal and material scope of protection under EU law are not explicitly articulated or indeed separately identified. In such circumstances, there is a risk that the ‘intention to confer rights’ becomes excessively abstract – a requirement to be ticked off on a checklist, rather than the crucial gateway that will define the parameters of Member State liability.

If that risk materialises, it opens up the route to two main subsequent possibilities. The first is that the Court’s neglect to explore the ‘intention to confer rights’ in fuller detail will lead to some important protective scope issues simply falling from view right at the outset and not being recaptured during the later stages of the Francovich analysis. If so, that could have some significant consequences. To begin with, it makes it more likely that subsequent steps in the Francovich analysis might proceed on an unclear basis. For example, it surely becomes more difficult to identify the direct causal link between the claimant’s loss and the Member State’s breach, if we have not clarified from the outset the exact nature of the losses the claimant is even entitled to be protected against; or rather, it may well become much easier to establish the required causal link, if the claimant’s intended rights have been articulated so generally that the Member State’s breach can almost automatically be associated with the claimant’s losses.

Furthermore, paying lip service to the ‘intention to confer rights’ and otherwise neglecting the protective scope issues relevant to Francovich implies a subtle but important shift in the emphasis and indeed purpose of the right to reparation under EU law. The more the Court effectively glosses over the ‘intention to confer rights’ criterion, the less effort it takes to articulate the legally-protected expectations of the individual; and instead, the more the Court’s attention will naturally focus upon the ‘sufficiently serious breach’ requirement, and thus upon the degree of culpability that should be attributed to the defaulting Member State. The central attention of Francovich liability moves away from ensuring the effective judicial protection of individuals in respect of damage done to their legally-defined personal interests; towards acting rather as a vehicle primarily for admonishing the Member State for having breached its obligations under the Treaties.

Such a prospect appears problematic for several reasons. To treat Francovich less as an action for reparation in respect of individual losses unjustly suffered, and more as a vehicle for imposing punitive and deterrent fines in respect of wrongs done, obviously runs contrary to the Court’s basic statement of principle in A.G.M.-COS.MET. But more fundamentally, to gloss over any discussion of the appropriate protective scope of EU law is also to avoid engaging with potentially contested legal and moral questions about the nature and extent of the rights and interests which might warrant protection through the EU’s system of non-contractual liability. In extreme cases, to treat Francovich primarily as a means of securing retribution against the defaulting Member State, divorced from any clear and compelling understanding of the basis upon which any given claimant should be entitled to challenge the relevant infringement of EU law, might even facilitate the unjust enrichment of particular individuals. Such concerns are surely magnified in those situations where (as we have seen) the obligations which the Member State is alleged to have breached are even legally distinct from the intended rights which are supposed to provide the basis for the individual’s call for reparation: the claimant might well be compensated, less on the basis of some superficial assessment of his/her own entitlement to legal protection, and more on account of the Member State’s behaviour in respect of an entirely separate duty imposed under the Treaties. Such an approach perhaps also sits uneasily with the scheme for punishment and deterrence envisaged by the Treaties themselves: after all, the imposition of financial penalties upon Member States for infringing their EU law obligations is meant to be governed by the particular legal and institutional provisions set out in Article 260 TFEU.

The second main possibility is that, far from being entirely neglected, certain protective scope issues will instead be picked up and addressed in contexts other than the ‘intention to confer rights’ criterion – be it the direct causal link, or some consequential dimension to the right to reparation – so that the subsequent legal elements of the Francovich action are not merely reflecting the findings about protective scope made at the ‘intention to confer rights’ stage, but actively seeking to perform some of the basic protective scope functions which were not conducted by the Court at the outset. Such an outcome is not inherently problematic (even if it sounds intuitively capable of aggravating concerns about maintaining coherence and consistency across the Francovich analysis). After all, this might well be seen as a choice made by the Court (to a greater or lesser extent deliberate) about where best to locate and address certain of the key protective scope issues posed by the Member State’s non-contractual liability under EU law within the particular legal structures of the Francovich action.

Nevertheless, that choice does have certain consequences – not least since, in most cases, it will imply that the relevant protective scope questions will be negotiated through a very different legal framework than that provided by the ‘intention to confer rights’ criterion. Instead of being treated as an integral part of defining the EU law right that forms the very basis of the claimant’s action, such protective scope issues would be understood rather as matters of subsidiary concern and thus suitable for regulation by domestic law in the first instance (subject primarily to the principles of equivalence and effectiveness). As such, the Court’s role is not to lay down common standards for the protective scope of Francovich liability in the exercise of its own hermeneutic monopoly; but rather to scrutinise ex post the Member States’ various choices about the nature and extent of the legal protection that might have been intended under EU law itself. Even if one accepts that national procedural autonomy is the natural point of departure for addressing supplementary issues such as the applicable limitation periods or the exhaustion of alternative remedies, it is perhaps less obvious that the Court should surrender its own responsibility to define directly the content and limits of the individual rights envisaged under the Treaties. That could risk compromising, at least to some degree, the unity and integrity of EU law itself.

Exploring further the potential implications of the Court’s tickbox approach to the ‘intention to confer rights’

How far might such risks and possibilities go beyond the purely linguistic and theoretical, so as actually to impact upon the nature and quality of the Francovich action in practice? We might usefully begin by considering how the Court responds to one of the most common and apparently most straightforward type of Francovich claims: those based upon an alleged breach of the Treaty free movement provisions. True enough, there is evidence here of some important protective scope questions being overlooked during the Court’s relatively superficial assessment of the ‘intention to confer rights’, instead to be picked up (if at all) only further down the line within other structurally distinct elements of the Francovich analysis.

After all, the Court’s usual response to whether the Treaty free movement provisions are intended to confer rights upon individuals rarely goes beyond the equivalent of a simple ‘yes, of course they are’.Footnote ³⁸ There is seldom any further investigation, at this stage, as to the precise nature or content of the rights conferred upon individuals by the Treaty free movement provisions. For example, consider entitlement to reparation in respect of pure economic loss. On one view, this question constitutes an integral part of defining the protective scope of the relevant EU law right, since it is directly concerned with identifying the latter’s basic material content, i.e. the extent of the legal interests in respect of which the claimant is entitled to legal protection through the medium of compensation. As such, one might expect this issue to be located squarely within the Court’s analysis of the ‘intention to confer rights’ criterion. But instead, the question of pure economic loss is generally overlooked by the Court, throughout its analysis of the substantive conditions that vest the right to reparation in principle, until we reach subsequent consideration of the recoverable extent and heads of damage – by which stage, the protective scope of the Treaty free movement provisions for the purposes of Francovich liability is to be reasoned instead through the presumption of national procedural autonomy, subject to the principles of equivalence and effectiveness.Footnote ³⁹

Indeed, one might argue that the Court’s tickbox approach to the ‘intention to confer rights’ requirement can mask some complex protective scope issues, or at least push such issues back to a later stage in the Francovich analysis. Consider the essentially conditional and indeed partial nature of the Treaty free movement provisions: the precise scope and extent of the abstract right to free movement must be evaluated in any given case, since it will meet its limit where EU law in fact permits the Member State to regulate the domestic market, notwithstanding any negative consequences that may have for cross-border trade. That evaluation will require the reconciliation of competing public interests, whose outcome will often hinge upon a detailed proportionality assessment – such that national measures may only marginally overstep the Member State’s legitimate power to regulate its domestic market.Footnote ⁴⁰ The primary Treaty provisions are nevertheless unforgiving of even such minor infractions: all national measures infringing directly effective free movement rights should be disapplied in those situations covered by EU law.Footnote ⁴¹ That state of affairs creates the concern that many claimants are delivered a great deal more free movement than they might have been entitled to expect, given the only limited degree to which the Member State overstepped the boundaries of legitimate market regulation.Footnote ⁴²

When it comes to the potential for non-contractual liability in such disputes – that is, of compensating individuals for the losses they have suffered, over and above the mere disapplication of conflicting national legislation – the legal structures of Francovich might appear to be almost as unforgiving as the directly effective free movement provisions themselves. For a start, we would expect for the Court to wave through as obvious that the Treaty free movement provisions are intended to confer rights on individuals. Instead, judicial attention would surely focus on whether the Member State’s breach of EU law should be considered sufficiently culpable, given the circumstances, as to justify imposing liability to make reparation. For those purposes, the substantive question of just how far the Member State might have gone in excessively regulating the marketplace is only one factor relevant to determining the existence of a ‘sufficiently serious breach’; even a marginal infringement of the free movement provisions might suffice to establish Francovich liability if the constellation of other relevant considerations so suggests. A tickbox approach to the ‘intention to confer rights’ therefore offers little recognition for the inherently conditional and partial nature of the Treaty free movement provisions: far from possessing an EU law right that corresponds only to the limit of legitimate national regulation, the claimant is treated as enjoying a simple and apparently boundless entitlement to trade on a cross-border basis, so that establishing liability is essentially a matter of interrogating the nature of the Member State’s own culpability.

That discrepancy between surface expectation and actual entitlement may be acknowledged (if at all) only at the stage of verifying the direct causal link and/or the recoverable extent and heads of loss: for example, were it possible under national causation rules to argue that the claimant’s losses were only partially attributable to the sufficiently serious breach, since the Member State would still have been entitled under EU law to enact similar if marginally less restrictive regulatory standards. At the very least, EU law is thus leaving an important question about its own protective scope to the application of diverse national rules. But at worst, the neglect fully to articulate and delimit the degree of legal protection the claimant was really entitled to expect under the Treaty free movement provisions, might lead such nuances to be lost entirely from consideration, thereby shifting the emphasis under Francovich away from effective judicial protection of the individual and towards effective punishment of the defaulting Member State.

To some degree, it may simply be that such issues have not yet been posed directly to the Court for suitable guidance. But there are plenty of similar problems that also await further clarification. Consider the relationship between first order EU law rights (such as the Treaty free movement provisions) that automatically govern situations falling within their own personal and material scope; and second order EU law rights (the general principles of EU law and the Charter of Fundamental Rights) that apply only within the scope of EU law, i.e. when their potential application has already been directly triggered by some first order provision.Footnote ⁴³ It is certainly possible to bring a claim for reparation based upon the Member State’s breach of the general principles and/or the Charter.Footnote ⁴⁴ But should those first and second order EU norms count as separate ‘intended rights’, each providing the basis for its own Francovich action? Or should they instead be treated as a single albeit composite right for the purposes of seeking reparation? That choice may decisively affect the rest of the Court’s analysis, by either broadening or narrowing the range of factors relevant to identifying, inter alia, a sufficiently serious breach, the direct causal link and the recoverable heads of damage.

By way of illustration, take a dispute inspired by the ruling in Carpenter.Footnote ⁴⁵ The third country national spouse of a British national is deported by the UK authorities (even though the marriage is accepted to be genuine). The British national succeeds in arguing that his situation falls within the scope of EU law: as an occasional provider of services to clients in other Member States, he is entitled to rely on Article 56 TFEU; since deportation of his wife is likely to interfere with his ability smoothly to provide those services, the UK must justify its deportation decision – and for those purposes, the UK must respect the fundamental right to private and family life as protected under the Charter as well as the general principles of EU law.Footnote ⁴⁶ If the claimant were to bring a Francovich action against the UK, should liability be founded on his first order right to provide services under Article 56 TFEU, so that he is arguing for compensation in respect of his lost contracts in other Member States? Or should liability be based upon the secondary right to respect for family life, so that the claimant is seeking compensation for the personal damage caused by his wife’s deportation? Or would the Francovich claim be underpinned by an undifferentiated combination of both rights to free movement and to family life? Obviously, each choice implies a rather different approach to assessing not only the recoverable losses, but also the requirements of breach and causation (which might seem more straightforward to prove in relation to the right to family life than as regards the right to provide services).

Such questions can arise not just across different (first and second order) provisions of EU law, but even within one and the same substantive EU norm. There has so far been little judicial engagement with the question of whether, specifically for the purposes of Francovich liability, any given EU law right can or should be divided into the pursuit of primary and secondary objectives – such that only the interests legally protected by the primary objectives of the relevant provision are apt to provide the basis for a right to reparation; while those interests legally protected only on a purely secondary basis would fall outside the legitimate protective scope of the EU system of non-contractual state liability. In fact, a first-order-norm/second-order-norm dispute like Carpenter could equally well be framed in such terms: the claimant’s freedom to provide economic services constituted the primary objective of his EU law rights; the protection of his private and family life was merely a secondary objective and as such might be treated as not amenable to vindication through the Member State’s non-contractual liability. But such an analysis is even more pointed in disputes which do not engage two distinct legally binding norms, only the interpretation of a single EU law right.

By way of illustration, take a dispute inspired by the ruling in Ruiz Zambrano.Footnote ⁴⁷ The third country national parent of an infant Belgian citizen is treated as unlawfully resident by the Belgian authorities and thus ordered with deportation as well as prevented from undertaking paid employment. However, Article 20 TFEU requires that the parent should enjoy a right to reside in Belgium, as the primary carer of a minor EU citizen, where the parent’s deportation would in fact endanger the child’s own continuing residence within the EU territory as a whole. Article 20 TFEU also requires that the parent should be issued with a work permit entitling them to engage in gainful employment during their period of lawful residency under EU law. For the purposes of Francovich liability, should we treat the parent’s bundle of derived entitlements to residency and employment as falling entirely within the protective scope of Article 20 TFEU – whether the consequences of that choice are worked out explicitly at the initial stage of the ‘intention to confer rights’ assessment, or only later (for example) when double-checking the heads of damage considered recoverable under national law? Or might the Court in such a dispute instead be persuaded to distinguish between the primary and secondary objectives of Article 20 TFEU, such that the Member State can be held responsible only for damage wrongly inflicted upon the family’s security of residence, but not for any losses incurred through the parent’s unlawful exclusion from the employment market?Footnote ⁴⁸

In short, how the Court chooses to define the ‘intention to confer rights’ can play a crucial role in the subsequent character of the entire Francovich assessment. Yet in practice, the Court rarely offers any detailed guidance. Instead, the ‘intention to confer rights’ is often treated in a tickbox fashion; and often also as a requirement to be fulfilled on an all-or-nothing basis. In consequence, the relationship between the autonomous Francovich criteria risks becoming excessively abstract: there is a right, there is a breach, and provided there is a causal relationship between the latter and the harm sustained, there will be liability. In many disputes, it may well be that the nature of the intended right and its relationship to the issues of breach and causation, as well as all other consequential liability questions, is too obvious to warrant being explored in greater detail. But that risks becoming a bad habit when extended into cases where the nature of the intended right and its influence upon the remainder of the Francovich assessment is more complex or contestable, i.e. where there is indeed a right, and there may well be a breach, but the final prospect of liability needs to emerge through a process of greater definition within and mutual interaction between these otherwise autonomous legal concepts. For those purposes, we are not suggesting that all relevant protective scope issues must be fully dealt with under the initial ‘intention to confer rights’ requirement. It is perfectly possible for such issues to be addressed elsewhere within the subsequent legal structures of the Francovich action. But as things stand, one may worry about the potential for some important protective scope issues, especially those concerning the precise nature and extent of the claimant’s ‘intended right’, to get altogether lost from consideration; or to be addressed only in a relatively superficial and disjointed manner; and in any case, to be mediated through concepts premised upon the (qualified) autonomy of the various domestic legal systems.

Some preliminary observations concerning how EU law deals with a diverse range of ‘rights’

We can begin with some preliminary (doctrinal, not jurisprudential) observations about how EU law handles the very idea of ‘rights’.Footnote ⁴⁹ Defining the capacity directly to enforce EU law – that is, identifying the range of persons who are entitled to rely upon any given EU provision as a direct source of legal rights/obligations before the domestic courts – is primarily the responsibility of EU law itself.Footnote ⁵⁰ The Member State’s procedural autonomy has but a subsidiary role to play: usually in regulating the various executive issues that arise once EU law has done the job of defining its own protective scope,Footnote ⁵¹ though sometimes also in making default choices about the full range of persons entitled to enforce EU rights / obligations before the national courts, where EU law has explicitly or implicitly delegated such authority to the Member State.Footnote ⁵²

In any event, the capacity directly to enforce EU law can cover a wide range of possibilities: it goes without saying that not every right to invoke EU provisions before the domestic courts will be of the same kind or quality. Consider the entirely familiar distinction between (on the one hand) substantive rights/obligations, which envisage a particular level of protection within any given dispute or relationship; and (on the other hand) purely procedural rights/obligations, which foresee only a degree of participation in some decision-making process but without predetermining its final outcome. Or consider the equally important distinction between (on the one hand) subjective individual rights, which are personal to their beneficiary, the breach of which should give rise to an equally personal remedy such as compensation; and (on the other hand) mere rights to standing, usually in vindication of some collective interest, where infringement should not lead the legal system to enrich the claimant personally. In particular, while many public law obligations can be assumed to create subjective rights for the benefit of particular individuals, the potential enforcement of which should thus be reserved to and for the direct benefit of those right-holders, other public law obligations are evidently intended to benefit society as a whole rather than any specific person/s, and imply the need for broad standing rules, as well as appropriate though not necessarily personal remedies, if they are to be effectively enforced: that will often appear to be the case, for example, with general environmental legislation. Indeed, the same may be true also for private law. Most such obligations will create personal rights for an identifiable group of individuals, the latter being responsible for the vindication of their own legal entitlements. But certain private law obligations may also require broader potential enforcement mechanisms if they are to be properly effective: for example, where individual right holders (such as consumers or employees) are in practice ill-equipped to defend their own interests, so that representative organisations should also be recognised as competent to initiate some form of collective legal action.Footnote ⁵³

Even playing just with those two variables, EU law can thus create a wide spectrum of possible scenarios, when it comes to the decentralised enforcement of Treaty norms before the national courts. At one extreme, the claimant might rely upon a subjective individual right to some particular substantive benefit or level of protection. At another extreme, the claimant might be entitled to no more than a mere right to standing before the domestic courts to promote the enforcement of certain general interest procedural obligations incumbent upon the competent public authorities.

However, whatever the abstract theory, it can undoubtedly prove more difficult in practice to express the potentially varied nature of one’s capacity to enforce EU law in terms which also capture the appropriate corresponding character of the claimant’s entitlement to judicial protection (in general) and when it comes to determining the availability of reparation pursuant to the Francovich case law (in particular).Footnote ⁵⁴ When will the claimant’s capacity directly to enforce any given provision of EU law also qualify as an ‘intention to confer rights’ specifically for the purposes of establishing the non-contractual liability of the Member State? And conversely, when should we assume that the lack of an appropriate ‘intention to confer rights’ for the purposes of the Francovich action should also disqualify the claimant from being recognised to enjoy any other capacity directly to enforce the relevant provisions of EU law before the domestic courts?

To help illustrate those issues, consider the ruling in Muñoz.Footnote ⁵⁵ The Court held that one private party must be able to bring a civil action to enforce certain EU agricultural regulations against another private undertaking (notwithstanding that, under domestic law, enforcement powers were reserved exclusively to a designated public authority). The claimant in Muñoz clearly fell within the protective scope of EU law: it was recognised as enjoying capacity directly to enforce the relevant regulations. But the Court did not clarify any further the precise nature of the claimant’s right under the Treaties. In particular, it remains unclear whether the claimant in Muñoz had an EU law right not only to seek an injunction to prevent its commercial rival from breaching the applicable EU quality control standards; but also to seek compensation in respect of any losses sustained as a result of the impact of that unlawful activity upon its own market position (by analogy with the ruling in Courage v Crehan).Footnote ⁵⁶ In other words, did its direct commercial concern confer a right of standing upon the claimant in Muñoz, but only so as to protect the public interest in a system of fair competition? Or was the claimant also endowed with the capacity to enforce its own subjective right to fair competition against the predations of its market rivals?Footnote ⁵⁷ Conversely, consider the ruling Peter Paul.Footnote ⁵⁸ As we know, the Court held that the various EU directives on deposit guarantee schemes and banking supervision are not intended to confer rights on depositors and cannot provide the basis for a Francovich claim against the Member State based on alleged shortcomings in discharging its regulatory duties; nor was there any obligation upon the Member State otherwise to establish a system for compensating depositors in respect of their losses, over and above the minimum requirements laid down by the relevant EU legislation. But even if the claimants in Peter Paul were not entitled to Francovich damages in respect of their non-existent subjective rights, would they at least have enjoyed standing (say) to seek judicial review before the national courts, in respect of a possible failure by the domestic regulatory authorities adequately to enforce the applicable EU directives?Footnote ⁵⁹

The task of translating one’s capacity directly to enforce EU law into an appropriate form of judicial protection is not limited to the distinction between subjective personal rights and mere rights of standing. Even assuming that EU law has indeed envisaged the creation of an individualised right, issues can also arise when it comes to the distinction between substantive and procedural levels of protection. After all, the treatment of substantive provisions within a system of non-contractual state liability obviously calls for a very different assessment than that appropriate to purely process-based provisions. For example, breach of a right merely to participate in any given decision-making process will present specific challenges when it comes to establishing a causal link between the Member State’s infringement and the claimant’s damage (especially where the national authorities argue that the outcome would have been the same in any case). Similarly, such a breach raises problems in deciding the extent and heads of recoverable loss (including whether they should be limited to the nominal value of the process rights itself, unrelated to any actual loss resulting from the contested final decision).

Indeed, nor need the challenges end there. Just as we have seen that it is possible to distinguish between the primary and secondary objectives of a given EU measure for the purposes of ascertaining the material scope of the claimant’s protected interests under EU law, so too it may be possible to distinguish between the primary and secondary beneficiaries of a given EU measure when it comes to defining the latter’s intended personal protective scope. Consider situations where the protective scope of EU law is not defined directly by the European Court of Justice, but rather by devolving responsibility onto the Member States through the principle of national procedural autonomy (subject to the requirements of equivalence and effectiveness). That was the case, for example, in Verholen: the claimant fell outside the personal scope of Directive 79/7 on equal treatment between men and women in the field of social security;Footnote ⁶⁰ nevertheless, as the husband of a woman who had suffered unlawful discrimination, Mr Verholen claimed that he actually suffered its adverse consequences in practice.Footnote ⁶¹ The Court held that, in principle, national law should determine an individual’s standing and legal interest in bringing proceedings; but here, a man who actually bore the adverse effects of sex discrimination against his wife should be entitled to bring legal proceedings before the national courts. That begs the question: what sort of judicial protection should such a claimant, who enjoys only a derived capacity to enforce EU provisions, be entitled to expect under the Treaties? In particular, would a claimant like Verholen be entitled to rely on their status as a secondary beneficiary of Directive 79/7, in order to seek reparations from the Member State under Francovich? In any event, it seems unlikely that the Court would sanction the availability of Francovich protection in situations where the Member State has recognised an individual’s capacity directly to enforce EU measures entirely as a matter of national discretion and free from any compunction under EU law itself.Footnote ⁶² After all, it seems difficult to argue that there is an ‘intention to confer rights’, when EU law itself appears so agnostic or even nonchalant about its own protective scope.

In other words, although they are legally distinct tasks, there is nevertheless a close inter-relationship between defining the capacity directly to enforce EU law (on the one hand) and identifying an ‘intention to confer rights’ for the purposes of the Francovich action (on the other hand). Deciding whether there is an ‘intention to confer rights’ for the purposes of Francovich will often involve, and indeed hinge upon, verifying the existence and nature of the claimant’s prior capacity directly to enforce EU law in the first place. As such, we should not be too surprised if uncertainty or confusion when it comes to EU law’s general approach to ‘rights’ also extends into the process of determining the scope of the Member State’s non-contractual liability under Francovich.

And vice versa. Indeed, based upon our analysis so far, one might predict that the Court’s tendency towards a tickbox assessment of the ‘intention to confer rights’ requirement, risks precisely to overlook any explicit discussion of just how far the claimant’s particular capacity to enforce EU law should translate into an entitlement also to claim reparation through the medium of the Francovich action. After all, if the standard formulation of the ‘intention to confer rights’ requirement which emerged from Brasserie and Dillenkofer neglects to focus attention on basic protective scope issues about the personal and material scope of the claimant’s EU law rights, one might expect the same to be true also when it comes to verifying the underlying nature of the claimant’s EU law rights and, in particular, their aptness for protection through compensation as a matter of EU public law. If that risk indeed materialises, the consequences would be similar to before: questions about the true nature of the claimant’s capacity to enforce EU law for the purposes of awarding compensatory damages would be pushed back (if at all) to subsequent stages in the Francovich analysis and thus likely be handled through a very different legal framework; in any event, the whole emphasis of the Francovich system would shift away from protecting the rights of deserving claimants, to focus more upon penalising the Member State for its wrongdoing.

Exploring the case law concerning when mere standing to enforce EU law might also provide the basis for a Francovich claim

How far may such theoretical problems have extended into judicial practice? Indeed as predicted, various rulings suggest that the Court does not always confidently separate out a mere right to invoke EU law before the national courts, from the creation of subjective personal rights apt for protection specifically through the Francovich case law. Moreover, it is noteworthy that these rulings typically involve a claim to enforce certain procedural obligations, rather than to benefit from a particular substantive level of protection under EU law. For example, Diahatsu Deutschland concerned a trade union action claiming standing to seek the imposition of penalties for the non-publication by a company of its accounts, as required by an unimplemented EU directive.Footnote ⁶³ Although there could be no (horizontal) direct effect of the relevant EU rules as against the company itself, the Court nevertheless referred explicitly to the possibility of the trade union seeking Francovich damages against the Member State for its failure to implement the directive correctly – without offering any clarification of whether the Court considered that the trade union’s mere right of standing to enforce a third party’s obligation was nevertheless sufficient to bring the claimant within the protective scope of Francovich liability (let alone how the national court should seek to value that right in compensatory terms).

Certain EU measures have generated entire bodies of case law of their own grappling precisely with these issues. Consider those rulings dealing with the Member State’s procedural obligation to notify to the Commission, then hold in abeyance, draft technical regulations, pursuant to what is now Directive 98/34.Footnote ⁶⁴ Notwithstanding the usual rule that directives cannot of themselves be enforced against individuals,Footnote ⁶⁵ the Court established in CIA Security and Unilever Italia that the duty to notify can indeed be enforced in litigation between two private parties,Footnote ⁶⁶ through disapplication of the relevant national rules, on the rather curious grounds that Directive 98/34 was intended to create neither rights nor obligations for individuals.Footnote ⁶⁷ Yet the Court in Sapod Audic then explicitly suggested that an individual who suffers loss as a result of the disapplication of national technical regulations pursuant to Directive 98/34 may yet seek compensation from the Member State pursuant to the Francovich case law.Footnote ⁶⁸

On the one hand, Sapod Audic could be interpreted as the Court’s attempt to resolve the inherent self-contradiction which would otherwise underlie the rulings in CIA Security and Unilever Italia, whereby a purely procedural directive that apparently creates neither rights nor obligations nevertheless appears uniquely capable of reshaping the legal relationships between individuals, but without offering any legal redress to those who undeniably suffer loss as a consequence of the Member State’s default. On the other hand, Sapod Audic sits uneasily with the ‘intention to confer rights’ requirement imposed by the Francovich action itself: why should a measure which is not designed to create any rights in favour of individuals be capable of furnishing the platform for a compensatory action against the defaulting Member State?

Moreover, the Court in Sapod Audic gave no hint as to its understanding of the material interests which might be protected by Directive 98/34 for the purposes of a Francovich action. Surely there was no subjective individual right to any identifiable substantive benefit or level of protection; yet nor was there any subjective individual right to any identifiable procedural benefit (such as a right to be heard). The only possible ‘right’ enjoyed by individuals pursuant to Directive 98/34 would be the mere right to invoke the Member State’s breach of its notification/standstill duties in proceedings before the national courts. But what financial value might we place upon such a right – especially given that the simple availability of litigation in itself fulfils the claimant’s expectations under EU law? Such logical and practical dilemmas were presented to the Court for resolution in Berlington Hungary.Footnote ⁶⁹ Referring directly to Unilever Italia and without any mention of Sapod Audic, the Court held that, since Directive 98/34 creates neither rights nor obligations for individuals, it is also and in consequence to be treated as incapable of providing the foundation for a Francovich action under EU law. A more rigorous application of the ‘intention to confer rights’ prevailed, albeit at the cost of judicial protection for those undoubtedly harmed by the Member State’s breach of EU law.

A particularly striking illustration is provided by the case law delivered pursuant to Directive 85/337 on the system of environmental impact assessments.Footnote ⁷⁰ The Court had regularly indicated that the range of persons entitled to bring legal action, for the purposes of enforcing Directive 85/337 directly before the national courts, should be conceived in broad terms: rulings such as Kraaijeveld referred to a right of access to the domestic courts simply in favour of ‘those concerned’.Footnote ⁷¹ That broad judicial approach has since been articulated and refined through more detailed provisions adopted by the EU legislature in order to specify the conditions under which members of the public and non-governmental organisations should be able to bring judicial review proceedings.Footnote ⁷² Either way, how are we to translate such a generous conception of environmental locus standi into an appropriate standard of judicial protection and, in particular, a suitable remedy for the claimant?

Signs that the Court might prove unsure about how to define the capacity to enforce EU environmental obligations, in terms of individual rights and their judicial protection, first arose in a series of enforcement proceedings brought by the Commission against Germany: the Court noted that various EU directives designed to increase levels of environmental protection actually created rights for individuals, though of an undefined nature; such rights should be capable of being enforced before the national courts, though it was left unclear by which precise means.Footnote ⁷³ The first real opportunity to clarify further the intended nature of such rights, and their corresponding entitlement to judicial protection, arose in the dispute in Wells.Footnote ⁷⁴ Here, the Court expressly envisaged that an individual seeking to enforce Directive 85/337 against a public authority which had granted permission to another private entity for the carrying out of certain mining works, without conducting the necessary environmental impact assessment, should also be able to seek damages against that public authority for having breached its EU law obligations. The ruling stopped short of suggesting explicitly that Directive 85/337 was intended to confer rights on individual members of the public for the purposes of the Francovich case law. Instead, the Court appeared to assume that the Member State should provide a mechanism for making good any harm caused to the claimant by its own failure to carry out an environmental impact assessment, in accordance with the principles of equivalence and effectiveness, and that that mechanism should include the possibility of obtaining financial compensation as a matter of national law.

Even to that extent, the ruling in Wells could already be seen as problematic. After all, one might fairly have assumed that, when it comes to carrying out environmental impact assessments in accordance with Directive 85/337, individuals merely have an EU right to standing in the public interest, not a subjective personal right that should be valued and compensated in monetary terms.Footnote ⁷⁵

Nevertheless, the Court went further in its subsequent decision in Leth.Footnote ⁷⁶ It turns out that the reference in Wells to the possibility of the defaulting Member State incurring liability under national law, was not intended to exclude the possibility also of the disgruntled citizen seeking Francovich reparation directly under EU law itself. For those purposes, the Court observed that the objective of Directive 85/337 is to protect the environment and quality of life; exposure to noise resulting from a qualifying project can have significant effects on the quality of life and potentially also the health of individuals. The Court concluded from those considerations that Directive 85/337 is intended to confer an individual right to have the environmental effects of qualifying projects assessed by the competent authorities and also to protect against pecuniary damage which is the direct economic consequence of such effects. For those purposes, the ruling also engaged in greater detail with the precise scope of the material interests considered to fall within the protective scope of Directive 85/337 and thus of the Francovich action: qualifying pecuniary damage would cover (for example) a decrease in property value; but not (say) certain competitive disadvantages.

There are good reasons why one might feel uncomfortable with those findings in Leth. After all, not only has the Court gone beyond a right of standing to bring judicial review proceedings in the public interest, so as to extract from the regulatory regime established under Directive 85/337 certain subjective individual rights. In addition, those individual rights are defined not merely in procedural terms, concerning the need to conduct and the possibility to participate in an environmental impact assessment; but also in substantive terms, conferring protection against the direct economic consequences of the environmental effects of a qualifying project. Furthermore, when it comes to articulating the full protective scope of Directive 85/337, the Court’s conception of the material scope of the individual interests safeguarded under EU law is essentially monetarised: the Directive protects against (certain) pecuniary damage to property values, rather than the non-material interests of the individual, let alone the general interests of the natural environment.

But however expansive and controversial the Court’s findings in Leth may appear when it comes to drawing the borderline between subjective personal rights and mere rights to standing, or that between substantive rights and purely procedural rights, we should recall that the Court’s full reasoning through of such protective scope issues, for the purposes of establishing Francovich liability on the part of the Member State, was not confined to the ‘intention to confer rights’ requirement. Rather, it was at the subsequent stage of verifying the existence of a direct causal link that the Court reined in the full potential impact of its previous findings. The essentially procedural nature of the Member State’s obligations under Directive 85/337, which did not in any way predetermine the substantive decision on whether the proposed project should be authorised to proceed, implied that failure to conduct an assessment could not in principle by itself be held responsible for any decrease in the value of the claimant’s property, even if it arose directly from the relevant environmental effects. In other words, the problem in Leth was less about the Court treating the initial ‘intention to confer rights’ criterion in an entirely superficial manner, then having to address the essential protective scope issues only at subsequent stages in its Francovich analysis. Rather, the judgment illustrates how the Court might adopt a very expansive approach to the protective scope of EU law for the purposes of the ‘intention to confer rights’; but then employ a range of alternative legal tools – principally, the denial of a direct causal link between the Member State’s breach of its essentially procedural obligations and the claimant’s pecuniary losses in relation to their admittedly personal and substantive rights – in order finally to settle the protective scope of the Francovich action. As we have seen, that expansive approach to ‘rights’ under EU law is contestable. As we also know, the choice to tackle protective scope issues through legal instruments predicated upon the presumption of national autonomy is equally open to debate.

In short, EU law has yet to develop a coherent framework for addressing the interaction between individual rights and collective interests. Given that the Court can encounter difficulties in seeking to define the appropriate capacity directly to enforce any given provision of EU law, it is perhaps unsurprising that a degree of uncertainty then also seeps into the ‘intention to confer rights’ criterion under the Francovich action, i.e. when it comes to the distinct if closely-related task of defining the appropriate protective scope of EU law specifically for the purposes of establishing non-contractual state liability. But an already difficult challenge is hardly helped by the fact that nothing in the ‘intention to confer rights’ assessment as currently formulated and undertaken by the Court poses such questions with any degree of clarity. That said, the case law does at least suggest that the Court is increasingly willing to provide more explicit guidance to the national courts when faced with EU provisions designed to create individual rights of a purely procedural (rather than more substantive) nature, for example, for the purposes of identifying the direct causal link.Footnote ⁷⁷

The public-private distinction under the EU system of judicial protection

Judicial protection under EU law is characterised, as much as within the national legal systems, by a division between public law and private law situations. In public law situations, the availability of certain remedies will indeed be as of right and follow directly from the mere breach of EU law: that is the case, for example, with the right to recovery of charges unlawfully levied by the public authorities.Footnote ⁷⁹ But the availability of other public law remedies will be conditional upon meeting certain criteria over and above proof of an infringement per se. That is certainly the case with non-contractual liability to make reparation for wrongdoing under the Francovich action: the Francovich criteria are specifically designed to limit the potential for Member State liability to arise in respect of the performance of public functions in the general interest.Footnote ⁸⁰ That is precisely why the Court requires not merely a breach, but a sufficiently serious breach, of the Treaties before a Member State can be required to make reparation for its wrongdoing. The relevant criteria to be taken into consideration for those purposes are focused around the degree of discretion conferred upon the Member State under EU law, but also taking into account factors such as the reasonableness of the Member State’s interpretation, whether it has acted in good faith and whether its approach was encouraged or condoned by the EU institutions themselves. Such criteria all reinforce the character of Francovich as a public law remedy which seeks to strike a fair balance between compensating the individual victim who has suffered loss at the hands of the Member State (on the one hand) and safeguarding the public interest in efficient and effective regulation and administration (on the other hand).Footnote ⁸¹

By contrast, the standard of judicial protection in respect of private law rights and obligations is not governed by the same considerations. In many cases, breach of a binding private law obligation will justify the provision of an effective remedy per se and without any further (fault-based) limitation on the defendant’s liability to make reparation.Footnote ⁸² For example, the availability of private law damages for infringements of EU competition law, in accordance with the Courage v Crehan jurisprudence, is based upon the principle that the defendant’s mere infringement of the Treaties provides sufficient culpability to justify its liability to make reparation in respect of losses that can be directly attributed to its unlawful behaviour.Footnote ⁸³ Of course, that is not to deny that certain public interest considerations also play a role in shaping the nature and limits of private law remedies: the state still has legitimate concerns about striking a fair balance between disputing parties or different social groups, or containing the risks associated with certain behaviours and activities, which can justify limiting or even excluding the possibility of certain private law liabilities.Footnote ⁸⁴ But those concerns are qualitatively different from those at play in public law situations, which are governed by fundamental constitutional considerations about relations between citizen and state in a democratic political system governed by the rule of law.Footnote ⁸⁵ Thus, when the Member State is acting in a purely private law capacity, the standard of judicial protection for individuals required under EU law should be governed by exactly the same framework as that applicable to any other private law actor. For example, under Directive 2006/54, although the Member States exercise a degree of discretion in prescribing an effective remedy for victims of sex discrimination in the sphere of employment, they cannot limit the availability of an effective remedy by introducing additional (fault-based) requirements going beyond the wrongdoing already outlawed by EU law – and all that is true regardless of whether the relevant employer is a public or private entity.Footnote ⁸⁶

In practice, of course, such public-private distinctions are not nearly so easy to draw as they are in theory. After all, EU law itself generally lacks any clear and systematic distinction – indeed, any clear and systematic criteria for distinguishing – between what should count as a public law or a private law situation for the purposes of the Treaties.Footnote ⁸⁷ Consider, for example, the rule embodied in Article 45 TFEU concerning non-discrimination against migrant workers. That rule might be breached by purely private conduct (as in the case of an individual employer whose terms and conditions discriminate against migrant employees);Footnote ⁸⁸ or in what could be seen as a private law context by a public authority (if the latter discriminates in its own terms and conditions of employment);Footnote ⁸⁹ or instead by what would undoubtedly be classified as the administrative conduct of a public authority (for example, in setting the criteria for access to employment training programmes);Footnote ⁹⁰ or by the regional/national assembly in exercising its legislative prerogatives (to set out the statutory framework of applicable employment law);Footnote ⁹¹ or indeed by means of improper judicial conduct (based on the Köbler case) which consists in refusing to vindicate the migrant worker’s EU law rights.Footnote ⁹² One and the same provision of EU law might thus be capable of manifesting itself in a variety of public and private law contexts. Moreover, the Court has long held (in rulings such as Bozzetti) that Member States are competent to designate the national courts having jurisdiction over any given category of EU action.Footnote ⁹³ The Member State may therefore classify a particular dispute either as one of private law to be adjudicated through its civil courts, or as one of public law to be pursued through its administrative courts.Footnote ⁹⁴ To the extent that EU law thus relies upon national law to categorise the public or private nature of certain legal actions, differing domestic approaches towards the liability of public authorities might end up applying to one and the same ‘right’ under the Treaties.

Without diminishing the importance or difficulty of such challenges, we can nevertheless argue that they should not cloud our judgment at least when it comes to the Francovich right to reparation: as an autonomous EU law action, designed specifically to incorporate safeguards for the Member State’s performance of its general interest functions, the Court should have little choice but to lay down a coherent and uniform approach to identifying those public law powers which are deserving of specially crafted protection from the imposition of excessive non-contractual liability – thereby separating them from the Member State’s ordinary private law powers, which should justly be governed by different principles of judicial protection.Footnote ⁹⁵

Exploring the case law concerning Francovich as a public (not private) law remedy

What does the available case law tell us about how well the Court has discharged that responsibility? The striking point to note is that nothing within the existing legal structures of the Francovich action expressly requires the Court to ensure that the protective scope of the system of Member State liability is limited to situations involving the exercise of public law powers under the Treaties.

In particular, we know that there is a distinction between (on the one hand) the EU law rights relied upon by the claimant to establish their entitlement to protection under Francovich and (on the other hand) the EU law obligations whose breach is the subject of scrutiny in order to determine the culpability of the Member State. But the ‘intention to confer rights’ criterion emerges as an inappropriate forum for checking that the claimant’s action is properly grounded in public law. In many cases, the claimant will be invoking an intention by EU law to confer what are undoubtedly private law rights – for example, relating to employment or consumer relationships (as was true in disputes such as Francovich and Dillenkofer respectively).Footnote ⁹⁶ Evidently, it is not the public or private law classification of the claimant’s intended rights which is to be considered decisive; this is not the relevant stage at which to delimit the protective scope of Francovich conceived as an essentially public law remedy.

Rather, it is the public or private law character of the Member State’s obligations that needs to be checked and verified. But at no subsequent stage in the existing legal structures of the Francovich action is the Court explicitly prompted or required to make such an assessment. One might assume the most natural point would come when evaluating the existence of a ‘sufficiently serious breach’, i.e. by asking whether that breach relates specifically to the Member State’s public law obligations under the Treaties. But if the Court is indeed undertaking that enquiry, whether here or elsewhere in its overall Francovich analysis, it must be doing so only implicitly. And to be fair, by and large, not unsuccessfully. After all, the vast majority of Francovich cases handled by the Court do indeed involve a breach of Member State obligations which are recognisably of a public law nature: non- or faulty transposition of directives into national law;Footnote ⁹⁷ legislative or administrative action which infringes a binding Treaty obligation;Footnote ⁹⁸ and shortcomings in judicial protection either as regards the applicable remedies and procedural rules or when it comes to the proper handling of final disputes (regardless of whether the specific disputes themselves relate to public or private law).Footnote ⁹⁹

But perhaps it is inevitable that, in a system where a crucial element in defining the proper protective scope of the Member State’s non-contractual liability relies upon an implicit rather than express process of assessment and verification, the Court will sometimes seem to sanction the public law remedy of Francovich reparations in a situation where both the claimant’s intended rights and the Member State’s relevant obligations appear to be of a purely private law character. Consider the ruling in Fuβ,Footnote ¹⁰⁰ which concerned an action brought by an employee against his public employer for having been required to work excessive hours in contravention of the Working Time Directive.Footnote ¹⁰¹ When asked whether the claimant was entitled to reparation in respect of the damage he had suffered as a result of an infringement of his basic employment rights, the Court simply observed that the Working Time Directive does not contain any specific provisions on sanctions or remedies, then proceeded to direct the national court to consider the claimant’s action for reparation in accordance with the Francovich case law – observing that the latter’s system of Member State liability ‘holds good for any case in which a Member State breaches EU law, whichever public authority is responsible for the breach’.Footnote ¹⁰²

Fuβ is a remarkable judgment.Footnote ¹⁰³ It is perfectly true that the Working Time Directive does not contain any explicit provisions on sanctions or remedies for its infringement. The same is true for the great majority of EU measures (including many that create private rights in fields such as employment or consumer law). The real question is what should fill that silence: a system of judicial protection suited to the exercise of public law powers in the general interest; or one which is more appropriate to the ordinary private law relationships regulated under EU law? The latter provides the most natural and compelling answer: whether employed in the public or private sector, Mr Fuβ was intended to benefit from certain employment rights whose infringement should generate the right under EU law to an effective remedy per se; or at least, if the availability of a remedy was to be made conditional upon some requirement of fault, that requirement should apply regardless of the public or private character of the employer.Footnote ¹⁰⁴ Instead, the Court decided that the Francovich action furnished the appropriate vehicle for analysing Fuβ’s claim to effective judicial protection – meaning that the claimant was not entitled to a remedy in respect of the infringement of his employment rights per se, or at least on the same basis as would any purely private sector counterpart, but was instead expected to demonstrate that his public employer had committed a ‘sufficiently serious breach’ of its obligations under EU law.Footnote ¹⁰⁵

To that extent, the ruling in Fuβ illustrates exactly the sort of confusion we identified and warned against before: without a clear and explicit understanding that the protective scope of the Francovich action is limited to the public law defaults of national authorities, a standard of judicial protection specifically designed to limit the provision of reparation to individuals so as to protect the general interest in effective and efficient public administration, is extended into an essentially private law relationship where no such countervailing considerations apply. Moreover, the Court’s approach poses the clear risk of introducing an unjustifiable discrepancy between the judicial protection of employees depending on (the irrelevant factor of) the identify of their employer: whereas public employees are now expected to fulfil the Francovich criteria in order to secure reparation for breach of their basic employment rights, we can assume that private employees in an entirely comparable position will continue to benefit from the general right to an effective remedy under EU law. The outcome is Fuβ is perhaps all the more frustrating because there was indeed an opportunity for the Court consciously to consider these issues, when it explicitly discussed the potential scope of the system of Francovich liability and its aptness to apply in this particular dispute. But instead of identifying and respecting the division between situations in which a national authority acts in a public or instead a purely private capacity for the diverse purposes of EU law, the Court simply cited as authority for its reasoning the Brasserie principle that Member States are to be held ultimately responsible under Francovich for the unlawful acts of all their public bodies – which is very true, but entirely beside the point, since such responsibility should only become engaged through Francovich where the dispute relates to the exercise of public law powers.Footnote ¹⁰⁶

In short, the question of the public law character of the dispute, which should be crucial to defining the protective scope of the Francovich action, is not really appropriate for discussion in the context of the initial ‘intention to confer rights’ criterion and it is not explicitly being asked under the subsequent ‘sufficiently serious breach’ requirement. Once the analysis has passed those points and an autonomous right to reparation under Francovich is already on the cusp of being fully vested – or just as likely, dismissed for want of sufficient Member State culpability – it is difficult to see how the situation can be salvaged through either the direct causal link test or the remaining consequential analyses that complete a Francovich analysis. As a result, judgments like Fuβ illustrate the very real risk that Francovich liability will extend beyond the public law situations it was specifically designed to cater for, into purely private law disputes which should be governed by a very different framework of judicial protection. As a result, the failure of EU law to find an appropriate home for this basic protective scope question within the existing legal structures of the Francovich action might well offer Member States undue protection from liability and thus undermine the effective protection of individual rights.