Some additional thoughts on the interaction between procurement remedies and the principle of State liability—re Fosen-Linjen (E-16/16) and Nuclear Decommissioning Authority ([2017] UKSC 34)


After I published some comments on the EFTA Court’s Judgment in Fosen-Linjen AS v AtB AS (E-16/16, see here) some three weeks ago, I have had some interesting exchanges and discussions with some academic colleagues (Dr K-M Halonen, Dr R Vornicu, Dr P Bogdanowicz, Prof R Caranta, Dr A Georgopoulos, Dr Herrera Anchustegui and Aris Christidis) and with policy-makers and practitioners (which mostly wish to remain anonymous). I am grateful to all of them for forcing me to think harder about some of the issues that derive from the Fosen-Linjen case and, in particular, for their repeated invitations to consider it by comparison to the Judgment of the UK Supreme Court in Nuclear Decommissioning Authority v EnergySolutions EU Ltd (now ATK Energy Ltd) [2017] UKSC 34 (the ‘NDA’ judgment; for my views on an interim decision at the start of the litigation, see here).

Indeed, comparing those cases is interesting, for the Fosen-Linjen and NDA judgments offer diametrically opposed views of the interaction between the use of damages as a procurement remedy and the principle of State liability for breach of EU law, in particular concerning the threshold for liability under the so-called second Francovich condition—ie whether liability arises from a ‘sufficiently serious breach’ of EU public procurement law, or from any (unqualified) infringement of the rules.

In this post, (1) I compare the approach to the procurement remedies-State liability interaction in both judgments, to then offer some brief reflections on (2) the implications of minimum harmonization of this subject-matter through the Remedies Directive (ie, Dir 89/665/EEC, as amended by Dir 2007/66/EC; see its consolidated version), (3) the possibility to reform the Remedies Directive so as to achieve maximum harmonization, and (4) the potential implications of a damages-based procurement enforcement strategy in the context of the emergence of EU tort law. This post is meant, more than anything, as an invitation for further discussion.

(1) Opposing approaches to the procurement remedies-State liability interaction

One of the contended issues in academic, and now also judicial, debate around public procurement remedies is the relationship between, on the one hand, the liability in damages derived from the Remedies Directive (art 2(1)(c), requiring a power for review bodies or courts to ‘award damages to persons harmed by an infringement’ of relevant EU public procurement rules) and, on the other, the liability derived from the general principle of State liability for breaches of EU law (following Francovich and Others, C‑6/90 and C‑9/90, EU:C:1991:428, and Brasserie du Pêcheur and Factortame, C‑46/93 and C‑48/93, EU:C:1996:79).

This is an issue that the Court of Justice of the European Union (ECJ) explicitly addressed in Combinatie Spijker Infrabouw-De Jonge Konstruktie and Others, C-568/08, EU:C:2010:751 ('Spijker'), when it stated that Art 2(1)(c) of the Remedies Directive

gives concrete expression to the principle of State liability for loss and damage caused to individuals as a result of breaches of EU law for which the State can be held responsible …

… as regards state liability for damage caused to individuals by infringements of EU law for which the state may be held responsible, the individuals harmed have a right to redress where the rule of EU law which has been infringed is intended to confer rights on them, the breach of that rule is sufficiently serious, and there is a direct causal link between the breach and the loss or damage sustained by the individuals. In the absence of any provision of EU law in that area, it is for the internal legal order of each member state, once those conditions have been complied with, to determine the criteria on the basis of which the damage arising from an infringement of EU law on the award of public contracts must be determined and estimated, provided the principles of equivalence and effectiveness are complied with (Spijker, paras 87 and 92, emphases added).

However, maybe surprisingly, Spijker is not (yet) universally seen as having settled the issue of the interaction between the actions for damages under the Remedies Directive and the Francovich doctrine.

As mentioned above, the main point of contention rests on what could be seen as a lex specialis understanding of the interaction between the two regulatory frameworks (which could formally match a literal reading of para 87 of Spijker, but is more difficult to square with its para 92)—ie a view that the general condition for there to be a ‘sufficiently serious breach’ of EU law under Francovich is relaxed by the Remedies Directive by solely mentioning the need for an (unqualified) infringement as sufficient ground for a damages claim. This is specifically a point where the UK Supreme Court and the EFTA Court have taken opposing views in their recent judgments.

The UK Supreme Court's approach

Indeed, in its NDA Judgment (as per Lord Mance, with Lord Neuberger, Lady Hale, Lord Sumption and Lord Carnwath agreeing), the UK Supreme Court followed what I think is the correct reading of Spijker and established that

para 87 [of Spijker] proceeds by making clear that the liability of an awarding authority is to be assessed by reference to the Francovich conditions. Subject to these conditions being met, paras 88 to 90 go on to make clear that the criteria for damages are to be determined and estimated by national law, with the further caveat that the general principles of equivalence and effectiveness must also be met (para 91). Finally, para 92 summarises what has gone before, repeating the need to satisfy the Francovich conditions (NDA, per Lord Mance, at [23]).

More importantly, the UK Supreme Court considered that

… there is … very clear authority of the Court of Justice confirming that the liability of a contracting authority under the Remedies Directive for the breach of the [public procurement rules] is assimilated to that of the state or of a public body for which the state is responsible. It is in particular only required to exist where the minimum Francovich conditions are met, although it is open to States in their domestic law to introduce wider liability free of those conditions (NDA, per Lord Mance, at [25], emphasis added).

Therefore, the UK Supreme Court takes the clear view that the existence of grounds for an EU damages action based on the Remedies Directive requires the existence of a ‘sufficiently serious breach’ of EU public procurement law. At the same time, it takes no issue with the possibility for more generous domestic grounds for actions for damages (although it eventually decided that this was not the case in relation to the Public Contract Regulations 2006; see NDA, per Lord Mance at [37], with which I also agree).

The EFTA Court's approach

Conversely, in its Fosen-Linjen Judgment, and despite the fact that similar arguments on the interpretation of Spijker were made before it (in particular by the Norwegian Government), the EFTA Court considered that

Article 2(1)(c) of the Remedies Directive … precludes national legislation which makes the right to damages for an infringement of public procurement law by a contracting authority conditional on that infringement being culpable … The same must apply where there exists a general exclusion or a limitation of the remedy of damages to only specific cases. This would be the case, for example, if only breaches of a certain gravity would be considered sufficient to trigger the contracting authority’s liability, whereas minor breaches would allow the contracting authority to incur no liability

A requirement that only a breach of a certain gravity may give rise to damages could also run contrary to the objective of creating equal conditions for the remedies available in the context of public procurement. Depending on the circumstances, a breach of the same provision of EEA public procurement could lead to liability in one EEA State while not giving rise to damages in another EEA State. In such circumstances, economic operators would encounter substantial difficulties in assessing the potential liability of contracting authorities in different EEA States’ (Fosen-Linjen, paras 77 and 78, emphases added).

This led it to reach the view that

A simple breach of public procurement law is in itself sufficient to trigger the liability of the contracting authority to compensate the person harmed for the damage incurred, pursuant to Article 2(1)(c) of the Remedies Directive, provided that the other conditions for the award of damages are met including, in particular, the existence of a causal link (Fosen-Linjen, para 82, emphasis added).

I already discussed (here) the reasons why I think the EFTA Court’s Judgment does not accord with the ECJ’s case law (notably in Spijker) and why I hope the ECJ will explicitly correct this situation. In the remainder of this post, I briefly discuss the themes of minimum and maximum harmonisation of procurement remedies that emerge from a comparison of the approaches adopted by the UK Supreme Court and by the EFTA Court.

(2) Minimum harmonization through the Remedies Directive

The UK Supreme Court’s approach is implicitly based on a conceptualisation of the Remedies Directive as a minimum harmonization instrument, which sets the basic elements of the (effective and equivalent) remedies that Member States must regulate for, in accordance with the peculiarities of their own domestic systems. I think that this characterisation of the Remedies Directive is uncontroversial (see eg the recent report by the European Commission on its implementation at Member State level, at 4). Following the logic of minimum harmonization, the UK Supreme Court clearly has no problem with the existence of two potential tiers of remedies: a lower or more basic EU tier (subject eg to a requirement of ‘sufficiently serious breach’), and a higher or more protective domestic tier (subject eg to ‘any infringement’), which may or may not exist depending on the policy orientation of each EU/EEA State.

This approach has both the advantage of being in accordance with the current state of the law as interpreted by the ECJ (as above), and of not imposing—as a matter of legal compliance, rather than policy preference—an absolute harmonisation of public procurement remedies (at least as the threshold of liability for damages is concerned).

However, this approach is not without some practical difficulties, as there is a thick mist of uncertainty concerning what is a sufficiently serious breach of procurement rules (but also of what rules in the EU directives are ‘intended to confer rights’ on the tenderers—ie the first Francovich condition, which has been so far largely untested), and the existing ECJ case law on the interpretation of substantive EU procurement rules would require significant reconceptualisation in order to provide clarity in this respect. The existence of the preliminary reference mechanism of Art 267 TFEU can alleviate this legal uncertainty (in the long term, and maybe starting soon with the pending decision in Rudigier, C-518/17), but not without creating a significant risk of collapse of the ECJ (or, at least, an even more significant growth in procurement-related preliminary references). From that perspective, the possibility to engage in maximum harmonization (as rather implicitly advocated by the EFTA Court) deserves some consideration.

(3) Maximum harmonization through a revised remedies directive?

In my view wrongly, the EFTA Court holds the implicit normative position that the Remedies Directive is an instrument of maximum harmonisation when it emphasises its ‘objective of creating equal conditions for the remedies available in the context of public procurement’ (see Fosen-Linjen, para 78 above, emphasis added). The EFTA Court derives this objective in an earlier passage, where it stresses that a 'fundamental objective of the Remedies Directive is to create the framework conditions under which tenderers can seek remedies in the context of public procurement procedures, in a way that is as uniform as possible for all undertakings active on the internal market. Thereby, as is also apparent from the third and fourth recitals to the Remedies Directive, equal conditions shall be secured (sic)' (Fosen-Linjen, para 66, emphasis added).

I think this is a clear judicial excess and I do not think the Remedies Directive can be considered an instrument of maximum harmonization (ie a tool that sets a ceiling, or even a common core of protections that must be uniformly provided in all EEA States) in the way the EFTA Court does. In my view, this is particularly clear from recital (6) of the Remedies Directive, according to which: ‘it is necessary to ensure that adequate procedures exist in all the Member States to permit the setting aside of decisions taken unlawfully and compensation of persons harmed by an infringement’ (emphasis added; note that adequate procedures are not necessarily homogeneous or identical procedures)--which the EFTA Court includes in its Judgment (para 3), but then largely ignores.

However, the EFTA Court does have a point when it stresses that the divergence of rules on (damages) remedies can distort the procurement field and, in particular, discourage cross-border participation—which could be alleviated by a reform of the Remedies Directive to create such maximum harmonization. Such revision and an explicit view on the elements of a uniform system of maximum harmonisation could bring a much needed clarification of the function and position of different types of remedies under its architecture—notably, it would clarify whether damages are a perfect substitute for other remedies (as the EFTA Court seems to believe) or an ancillary remedy [as I posit, maybe not in the clearest terms, in A Sanchez-Graells, '"If It Ain't Broke, Don't Fix It"? EU Requirements of Administrative Oversight and Judicial Protection for Public Contracts' in S Torricelli & F Folliot Lalliot (eds), Contrôles et contentieux des contrats publics (Bruylant, 2018)]. Maximum harmonisation could also provide an opportunity to consider the creation of safe harbours (at least of damages liability) for purely procedural errors, or in the context of certain general guidelines.

Nonetheless, despite potential advantages derived from a revision of the system to consider maximum harmonization, given the vast differences in the rules on damages claims across EU jurisdictions, it would be certainly difficult, if not outright impossible, to reach an agreement on the adequate level of protection and the relevant procedural mechanisms [for comparative discussion, see for example, the contributions to S Treumer & F Lichère (eds), Enforcement of the EU Public Procurement Rules (DJØF, 2011), and to D Fairgrieve & F Lichère (eds), Public Procurement Law. Damages as an Effective Remedy (Hart, 2011); see also H Schebesta, Damages in EU Public Procurement Law (Springer, 2016)].

Given these practical difficulties, I would not think the European Commission would be willing to engage in the exercise of designing such maximum harmonization, even if it decided to revise the Remedies Directive in the future (which, unfortunately, seems very unlikely at least for now). What then should not be acceptable is for such maximum harmonisation to be achieved or imposed through an excessively broad interpretation of the Remedies Directive as, in my view, the EFTA Court's Fosen-Linjen judgment does.

(4) Damages-based enforcement of procurement rules & EU tort law

As a last thought, I think it is worth stressing that, in addition to the practical difficulties derived from the current minimum harmonization of procurement remedies, and the not smaller difficulties in attempting a maximum harmonization, there are also structural tensions in the use of damages actions for the enforcement of EU public procurement rules. As recent research has clearly shown (see P Giliker (ed), Research Handbook on EU Tort Law (Elgar, 2017)), the use of damages actions (either based on Francovich liability, or sector-specific rules) for the enforcement of substantive EU law creates distortions in the domestic legal systems of the Member States. From that perspective, both the minimum and maximum harmonization approaches are problematic.

From the minimum harmonization perspective, because the existence of two tiers of protection can also result in two tiers of regulation and/or case law concerning the interpretation and application of the rules, which is bound to create legal uncertainty (eg if issues around the effectiveness of the remedy in the EU-tier create pressures on the interpretation of the domestic-tier remedies as a result of reverse pressures resulting from the principle of equivalence—ie the domestic remedy can hardly be both broader in scope and less effective in its consequences).

From the maximum harmonization perspective, because the creation of a one-size-fits-all remedy (such as that derived from the lower threshold for damages liability in the EFTA Court’s Judgment) can have rather drastic impacts for some Member States (in particular, those without a ‘higher-tier’ domestic protection), not only in the area of procurement law, but also in other areas of (economic) law which regulation and case law can be distorted as a result of the EU rules.

Thus, it seems adequate (and it may not be too late…) to reconsider a drastic change in the enforcement strategy to reduce the current over-reliance on tenderer-led administrative and/or judicial reviews, and start to move away from damages-fueled private enforcement of EU public procurement law and towards a more robust architecture of public enforcement with a restriction of damages compensation solely in exceptional cases—certainly where that compensation goes beyond direct participation costs.

Discussing the possibilities of doing so and the challenges it would imply far exceeds the possibilities of this post, but given that reaching a ‘happy median’ in the regulation of (private) damages actions in the context of procurement remedies in the EU would not be a minor feat, it may be time to (re)open that discussion.