A strange Scottish case on evaluation of tenders -- Boston Sci. Ltd v Common Service Agency [2016] CSOH 132

I find the recent Scottish case Boston Scientific Limited v The Common Service Agency [2016] CSOH 132 most confusing. This is a case of healthcare-related procurement whereby the Scottish NHS' central purchasing body, the Common Service Agency, was tendering framework contracts for the supply of certain types of medical equipment--for simplicity, pacemakers and implantable defibrillators.

The litigation concerned the applicable award criteria and the ensuing evaluation of the tender submitted by Boston Scientific Limited. Even if the case seems to be decided mainly on procedural grounds (the claimant, or pursuer in Scottish terminology, seemed to have been time-barred in raising a challenge against the published award criteria), it raises substantive issues that, in my view, should have been dealt with differently by the Court.

The tender had been advertised and the relevant invitation to tender (ITT) had published the applicable award criteria. For each of the lots in which the framework agreement was to be divided, the tenders would be assessed against a pass/fail criterion of essential features (ie mandatory technical specifications) and then evaluated on a 60:20:20 split of the maximum score of 100. The offered price would carry a 60% weight, whereas two quality criteria would carry 20% each: (a) the inclusion of certain defined desirable features and (b) the longevity of the devices.

The challenge was based on Boston Scientific's submission that the assessment of the longevity of the pacemakers must have been wrong. In a nutshell, Boston Scientific claimed that their position as market leaders and the existence of independent tests that demonstrated that their devices had a very long individual life led them to the conclusion that 'there must have been a failure to compare like with like because if there had not been such a failure [Boston Scientific] would have had the highest longevity scores' [para 15]. In short, the tenderer was not convinced that its competitors could (truthfully) have offered devices with a superior longevity.

Boston Scientific's submission of improper technical evaluation of the tenders is complicated by two additional factors. First, that tenderers had only been asked to declare (or self-certify) the longevity of their devices without providing any supporting evidence. Second, that the criteria applicable to the evaluation of the longevity component were not all that clear.

Self-certification of verifiable technical characteristics?

On the first issue, the complaint considers that the contracting authority was not allowed to include as award criteria elements based on pure self-declaration and which it intended not to verify. Indeed, the case seems peculiar because the Common Service Agency 'had stated clearly prior to the date for submission of tenders that supporting evidence was not sought.  It had protected itself in a different way by making clear that the framework agreement would include a clawback provision if battery life fell short of the figure submitted in a tender' [para 18].

This seems to me to be a peculiar way of conducting business because the longevity of devices that need to be implanted in the human body seems a rather important technical characteristic (as submitted by Boston Scientific, but dismissed by Lord Tyre in his Opinion, despite the relevance of this issue for the purposes of EU consumer law as discussed here), and the abrogation of the power to check compliance with technical specifications in this regard seems odd, regardless of the inclusion of financial penalties in the contract. The Judgment relies on two English precedents that would support the legality of relying on self-certification of compliance with contractual terms. Most importantly, it ignores the EU precedent in EVN and Wienstrom (C-448/01, EU:C:2003:651), to which one of the English cases refers, though. A reference to EVN paras [50]-[51] would have sufficed to quash the award procedure (I am thankful to Karen Wontner and Erik Plas for having raised this point in private correspondence).

First, Lord Tyre relies on Public Interest Lawyers v Legal Services Commission [2012] EWHC 3277 (Admin), Cranston J at para [64] to justify the acceptability of self-certification. However, in my view, this precedent is inapplicable here. First, because it concerned an on-going requirement to be discharged during the execution of the contract (ie an element closer to a contract performance clause than a technical requirement) but, most importantly, because in the previous paragraph of that speech Cranston J stressed that 

... the principle behind its decision was the need to ensure the equal treatment of tenderers through the objective and uniform application of the criteria in their assessment. The principle applies whether or not the public authority is able to verify the criteria. If it is able but omits to do so, that is as much a breach of the duty as if it sets criteria which cannot be verified. That is because the outcome may be an inequality of treatment of tenderers through the equal treatment of unequals, i.e. the equal treatment of those meeting and those failing to meet the tender requirements. After all, it is trite law that equality of treatment means not only treating like cases alike but unlike cases differently [at 63].

And this led Cranston J [at para 65] to insist on the need for robust verification where the contracting authority relies on self-certification by the tenderers. This is important in the context of the Boston Scientific v Common Service Agency dispute because, this case, 'Although it was accepted that in some cases a contracting authority might have a duty to validate information provided by a tenderer, this was not such a case. The defender did not have the means to verify independently the figures for longevity provided by tenderers' [para 18]. The issue here would have been whether this inability of independent verification (a) covered a complete lack of engagement with existing technical information and (b) was not attributable to the contracting authority itself and its decisions on how to organise the procurement procedure. Generally, one would expect that the entity running framework contracts for medical supplies has (or has access to) necessary technical knowledge in any case. Thus, this point of the case remains obscure and, in my opinion, shows excessive deference to the contracting authority.

Second, Lord Tyre relies on Parker Rhodes Hickmotts Solicitors v Legal Services Commission [2011] EWHC 1323 (Admin), McCombe J at paras [35]-[40], which in turn refers back to Public Interest Lawyers v Legal Services Commission. The difficulty with this second case is that its ratio rests on the construction or interpretation of the tender documentation, rather than an assessment of the requirements of the principle of non-discrimination of tenderers--which was the legal basis for the challenge in Boston Scientific. Importantly, in Parker Rhodes, the relevant part of the Judgment focuses on the fact that the Information for Applicants (IFA) document had not indicated how the contracting authority would proceed to verifying specific aspects of the offers, which the Court considered to cover the possibility of relying on self-certification.

To me, this makes both precedents irrelevant for (if not contradictory to) the assessment of the claims raised by Boston Scientific, which aimed to strike down the procurement process on the basis that the contracting authority had appended a significant weight to a criterion it actually decided not to verify at all. In my view, there are good arguments under the principle of good administration (Art 41 CFR) to demand that contracting authorities only evaluate what they can assess and, even more, that they do not claim not to be in a position to assess technical characteristics of the products they are buying--if nothing else, by reliance on the rules on test reports, certification and other means of proof (now under Art 44 Dir 2014/24/EU). 

Longevity, price, both or none of the above?

Additionally, and focusing on the point of the need to construct or interpret the tender documents as published, the second argument raised by Boston Scientific deserves attention as well because, indeed, the criteria applicable to the evaluation of the longevity component were not all that clear. In that regard, it must be noted that the ITT had established that:

In relation to longevity, the tender receiving the highest total longevity score would receive 20 points.  Each other tender would receive “20‑X points where X = 0.2 x the percentage by which each price in each tender exceeded the lowest price tender achieving the lowest total price score” (Boston Scientific v Common Service Agency, para [5], emphasis added).

This seems odd because the criterion that is aimed at scoring longevity is (or, at least, seems to be) referential to the price of all tenders except that of the tender with (self-certified) longer individual device life. In my view, this is a breach of the general scoring rule included in the ITT, according to which price would carry a weight of 60%. This would not be true except for the tender self-certifying highest longevity, and all other offers' price would be taken into account twice (once for the price component itself, and a second time for the scoring of longevity). This is, simply, technically incorrect and, in my view, should have sufficed to cancel the tender.

However, this does not seem to be the whole story and a mistake must have happened in the preparation of the ITT (there seems to be an obvious explanation if one thinks in terms of copy and paste ...) because, in a debriefing letter, the contracting authority had indicated to Boston Scientific that:

The weighting for Longevity was 20% therefore in each lot the longest longevity submitted received 20 points. The scoring guidance in section 3.3 clearly identifies the points allocated to longevity and how the tender would be scored ... (i.e. if one product had longevity or 100 months (longest) it would score 20 points and if a different product submitted had a longevity of 50 months it would score 10 points)  (Boston Scientific v Common Service Agency, para [10], emphasis added).

Now, this is the natural understanding of a relative scoring for longevity, but it happens not to be the scoring rule disclosed in the ITT, which made reference to relative prices rather than relative longevity. Such a substantial deviation between disclosed scoring rule (even if absurd) and its application seems to run against the basic requirements of the principles of transparency and equal treatment, as recently recast by the Court of Justice of the European Union in TNS Dimarso (for a comment, see here).

In my view, this should also have been taken into account by the Court and, rather than dismissing the challenge, Lord Tyre should have sought to understand better whether the longevity criterion had been assessed as the debriefing letter said, or rather as the ITT established (which could have led to abnormal results ultimately preventing Boston Scientific from making much sense of the scores obtained). Most likely, a divergence between the published scoring rules and the actual evaluation of the tender should have led to a cancellation of the award in any case.

Overall, I think that there are two main problems with the Judgment in Boston Scientific v Common Service Agency, and both of them seem to me to result from a lack of engagement with the case law of the Court of Justice of the European Union by the Scottish court. First, because it is truly abnormal to allow for self-certification of a technical requirement that can be assessed and verified by the contracting authority--at least, by reference to technical documents. Second, because it is also truly remarkable that a contracting authority can evaluate tenders in a way that deviates from the published criteria without the reviewing court picking up on this important aspect (or anomaly) of the process. Ultimately, in my opinion, this is a strange case. But also a very technically deficient Judgment and an incorrect decision.

ECJ confirms that procurement rules do not apply to licences or authorisations (“concessions”) for betting and gambling services (C-225/15)

In its Judgment of 8 September 2016 in Politanò, C-225/15, EU:C:2016:645, the European Court of Justice (ECJ) followed the Opinion of Advocate General Wahl (see here) and confirmed that the 2004 procurement rules were not applicable to a public contest for the award of concessions (ie licences or authorisations) for the provision of betting and gambling services to the public.

The ECJ did not address AG Wahl’s obiter comments concerning the theoretical applicability of the 2014 Concessions Directive to an equivalent case but, in my view, the stress put by the ECJ in the analysis of the essential elements of remuneration and risk transfer in the definition of a (services) concession indicates that the ECJ would have likely ruled against that applicability.

In the Politanò Judgment, the ECJ addressed the question whether Directive 2004/18 was applicable to a call for tenders for the grant of “concessions” in the field of betting and gambling by focusing on the remuneration element that is necessary for the existence of a public contract.

After distinguishing public service contracts from services concessions by reference to the different modalities of remuneration they imply and the different risk structure that underlies those modalities of remuneration (paras 30 to 31), the ECJ focused on the plain and simple fact that ‘in the case in the main proceedings, the service provider receives no remuneration from the contracting authority and bears the entire risk associated with the activity of collecting and transmitting bets’ (para 32, emphasis added).

This led the ECJ to conclude that such “concessions” could not ‘be classified as a public contract for services within the meaning of … Directive 2004/18’ (para 33), which leaves them outside of its scope of application (para 34). The ECJ does not make an equivalent explicit conclusion concerning the classification of those “concessions” for the provision of betting and gambling services as services concessions because those were explicitly excluded from the scope of application of Directive 2004/18 in any case (para 29).

In my view, however, that conclusion would be unavoidable in an equivalent case that took place after the entry into force of Directive 2014/23 because its Art 5(1)(b) defines a services concession in the following terms:

a contract for pecuniary interest concluded in writing by means of which one or more contracting authorities or contracting entities entrust the provision and the management of services other than the execution of works ... to one or more economic operators, the consideration of which consists either solely in the right to exploit the services that are the subject of the contract or in that right together with payment (emphasis added).

Importantly, Art 5(1) in fine of Directive 2014/23 also requires that

The award of a works or services concession shall involve the transfer to the concessionaire of an operating risk in exploiting those works or services encompassing demand or supply risk or both ... (emphasis added).

Taking all of this into account, and on the basis of the same factual finding that in the case of licences or authorisations for the provision of betting and gambling services (typically) ‘the service provider receives no remuneration from the contracting authority and bears the entire risk associated with the activity of collecting and transmitting bets’ and, possibly more importantly, the fact that 'the "service" under analysis [is] not provided on behalf of the contracting authority' (see Opinion of AG Wahl, para 51), the only possible conclusion is that Directive 2014/23 is equally inapplicable to a call for tenders for the grant of “concessions” in the field of betting and gambling.

This simple and unsurprising conclusion is slightly more interesting when taken together with the also recent Judgment in Promoimpresa (see here) because, together, they provide some additional clarity on the limits of application (or rather, outright inapplicability) of public procurement rules to “concessions” in name that are actually regulatory systems of licences or authorisations to carry out specific economic activities, whether they involve the use of public assets (generally, parts of the public domain) or not.

ECJ deviates from AG Sharpston’s Opinion and accepts use of “sure-refund” good conduct guarantees in public procurement litigation (C-439/14 and C-488/14)

In its Judgment of 15 September 2016 in Star Storage, joined cases C-439/14 and C-488/14, EU:C:2016:688, the European Court of Justice (ECJ) has deviated from the Opinion of AG Sharpston (see here) and ruled that, read in the light of Article 47 of the Charter of Fundamental Rights of the European Union, the Remedies Directives must be interpreted as not precluding national legislation that makes the admissibility of any action against an act of the contracting authority subject to the obligation for the applicant to constitute a good conduct guarantee if that guarantee must be refunded to the applicant whatever the outcome of the action.

I find this Judgment seriously troubling for two reasons: (a) it goes against basic intuitions of the effect of financial requirements on access to justice and (b) I do not grasp the purpose of “sure-refund” good conduct guarantees, which seem to be useless procedural hurdles. I develop these points below.

The main legal issue in Star Storage

The legal issue raised by the joined cases decided in Star Storage has been a rather moving target because the underlying Romanian rules have been altered in the period between the referral of the question to the ECJ and its Judgment. The initial question concerned the compatibility with EU law of requirements to furnish a good conduct guarantee in order to challenge procurement decisions under the risk that the guarantee would be executed in case of negative results for the litigant.

The forfeiture of the guarantee was later declared unconstitutional by the Romanian Constitutional Court and, as a result, the only question left for the ECJ to consider revolved around the compatibility of such good conduct guarantees in the scenario where they would be refunded to the challenger of the procurement decision, whatever the outcome of the review process.

So, in short, the ECJ had to consider whether the Remedies Directives and Art 47 CFR excluded the possibility to require the provision of a “sure-refund” good conduct guarantee in order to challenge public procurement decisions under Romanian law (C-439/14, para 38).

After rehearsing its standard case law concerning the Remedies Directives’ objective of ensuring the effectiveness of the substantive EU public procurement rules (paras 41-44) and stressing that the 2007 review of those rules aimed at ensuring ‘full respect for the right to an effective remedy and to a fair hearing, in accordance with the first and second paragraphs of Article 47 of the Charter’ (para 45), the ECJ focuses on the specific assessment of the “sure-refund” good conduct guarantee and follows the analytical framework proposed by AG Sharpston in her Opinion, which started from the position that

the good conduct guarantee … constitutes, as a pre-condition for getting any challenge examined, a limitation on the right to an effective remedy before a tribunal within the meaning of Article 47 of the Charter which, in accordance with Article 52(1) of the Charter can therefore be justified only if it is provided for by law, if it respects the essence of that right and, subject to the principle of proportionality, if it is necessary and genuinely meets objectives of general interest recognised by the EU or the need to protect the rights and freedoms of others (see judgment of 4 May 2016, Pillbox 38, C-477/14, EU:C:2016:324, paragraph 160) (C-439/14, para 49, emphasis added).

In assessing this test, the ECJ considers that the first requirement of explicit legal basis is met (para 50). This does not seem controversial. However, and this is where the ECJ starts to deviate from the analysis of AG Sharpston, the Court also considers that ‘the fact that the good conduct guarantee may reach the substantial amount of EUR 25 000 or EUR 100 000 cannot lead to the conclusion that the obligation to give such a guarantee undermines the fundamental content of the right to an effective remedy since, in any event, that guarantee, cannot be kept by the contracting authority, whatever the outcome of the action’ (C-439/14, para 50, emphasis added).

The ECJ further considers the measure adequate because the aim of the good conduct guarantee is justified by the legislative aim of avoiding the abuse of the remedies system so as to ensure the administrability of the procurement process (paras 52 and 53), and that ‘A financial condition such as the good conduct guarantee … is a measure liable to discourage frivolous challenges and ensure that all individuals have their actions dealt with as rapidly as possible, in the interest of the proper administration of justice, in accordance with Article 47, first and second paragraphs, of the Charter’ and that this is so even if ‘the obligation to provide a good conduct guarantee is a less dissuasive measure in its current version than in its initial version, since it can no longer be automatically and unconditionally kept by the contracting authority in the case that the appeal is rejected or withdrawn, [because] that obligation is still able to achieve the objective of combating frivolous actions pursued by the Romanian legislation’ (C-439/14, paras 54 and 56, emphasis added).

It finally considers the measure proportionate, mainly because ‘The good conduct guarantee of 1% of the value of the public contract, limited in accordance with the type of contract remains modest (see judgment of 6 October 2015, Orizzonte Salute, C-61/14, EU:C:2015:655, paragraph 58), in particular for tenderers which must normally demonstrate a certain financial capacity. That guarantee may, next, and in any event, be constituted in the form of a bank guarantee. Finally, it has to be constituted only for the period between the filing of the application and final judgment’ (C-439/14, para 61, emphasis added).

Issues around access to (administrative) justice

The first aspect in which I find the Star Storage Judgment criticisable concerns the analysis of proportionality. I think that the imposition of financial requirements and costs in order to challenge procurement decisions—including the payment of (non-negligible) courts fees—should be considered more clearly contrary to Art 47 CFR and the Remedies Directives. These rules require the recognition of standing to challenge procurement decisions ‘at least to any person having or having had an interest in obtaining a particular contract and who has been or risks being harmed by an alleged infringement’, and not only to those that can foot the bill of a (bank-issued) financial guarantee or absorb the opportunity cost of having a significant amount of money idle for the duration of the review procedures. Moreover, the assessment of proportionality in the terms carried out by the ECJ in the Star Storage Judgment can be particularly burdensome for SMEs, for whom the effects of a financial requirement proportional to the value of a contract they were not awarded can be clearly disproportionate, or at least imply an excessive risk.

Also, an assessment of proportionality should include a consideration of whether less restrictive measures are available. In the specific setting of aiming to discourage frivolous litigation, it would seem that the creation of a system of court-administered fines would be superior and reduce the ex ante restriction of access to review procedures. Moreover, that system could include provisions allowing the review body or court to ask for financial guarantees as interim measures only where they are necessary to ensure the possibility of the fine having to be paid at the end of the procedure (although I am not sure that this mechanism to avoid bankruptcy proofness is necessary).

It would also be possible to create a domestic discretionary exclusion ground for spurious litigants on the basis that this conduct makes ‘the economic operator … guilty of grave professional misconduct, which renders its integrity questionable’ [Art 57(4)(c) Dir 2014/24]. In my view, the existence of these potential alternatives should have been taken into account and this could have led to a finding that the upfront requirement of good conduct guarantees is in itself disproportionate.

However, this is not the weakest point of the Judgment.

Uselessness of “sure-refund” good conduct guarantees

Rather, in my view, the weakest and most criticisable aspect of the Star Storage Judgment is that it fails to recognise the futility of “sure-refund” good conduct guarantees. As simply and clearly put by AG Sharpston in her Opinion,

such a procedural requirement does not protect contracting authorities adequately from frivolous challenges … the contracting authority has to return the good conduct guarantee to the applicant within five days following the date on which the decision ... or the judgment has become final, even where the applicant manifestly abused his right to access review procedures. The costs which the … regime involves may therefore not be such as to discourage an economic operator from lodging a challenge that pursues an objective other than those for which the review procedures are established — for example, harming a competitor. They may nevertheless prove an obstacle to an economic operator with an arguable claim but limited means (Opinion of AG Sharpston, para 56, emphasis added).

It is surprising that the ECJ diverges from this assessment. In general terms, the ECJ implicitly dismisses it by stressing that ‘although the obligation to provide a good conduct guarantee is a less dissuasive measure in its current version than in its initial version, since it can no longer be automatically and unconditionally kept by the contracting authority in the case that the appeal is rejected or withdrawn, that obligation is still able to achieve the objective of combating frivolous actions pursued by the Romanian legislation’ (C-439/14, para 56). The rationale for this assessment is developed in the following terms:

Mobilising a sum of that amount by bank transfer, like the requirement to take the steps necessary to constitute a bank guarantee and pay the fees relating to it are such as to encourage applicants to carefully consider bringing an action. Furthermore, in so far as it undermines the applicant’s resources or, at least, its ability to obtain credit until that guarantee is refunded, the good conduct guarantee is of such a nature that it encourages applicants to act prudently in the proceedings they bring, consistent with the requirement … that the review procedures … are conducted as rapidly as possible. … it is conceivable that such a financial condition will encourage potential litigants to seriously evaluate their interest in bringing legal proceedings and their chance of winning and thereby dissuade them from bringing claims which are manifestly unfounded or which only seek to delay the award of a contract (C-439/14, para 59).

In my view, the ECJ fails to address AG Sharpston’s concerns. There is indication of the cost of such a financial guarantee, but it is certainly easy to foresee that (especially for large contracts), it may well be a minor amount for resourceful litigants willing to incur that cost in order to obtain some competitive advantage.

On the contrary, it is surprising that the ECJ does not use the exact same reasons detailed in para 59 of the Star Storage Judgment to acknowledge the barrier that the guarantee represents, which it ‘saves’ by indicating that the absolute value of the guarantees remains ‘modest’ (para 61). In my view, it is not possible to have it both ways. Either the requirement is modest and, therefore, unable to provide sufficient deterrence for resourceful litigants, or it is a serious barrier to the exercise of legal actions (whether legitimate, which should overcome the barrier, or illegitimate, which should not) and, consequently, it cannot overcome an analysis of strict proportionality in the terms discussed above.

Overall, once more, I find the judgement of the ECJ lacking commercial and financial realism and I start to wonder whether we will see a reversal of this trend any time soon—which seems unlikely, particularly when the ECJ deviates from the well thought-through proposals of some of its Advocates General, such as Sharpston’s in the Star Storage case.

What does Brexit mean for public procurement? Short remarks on Arrowsmith's White Paper

Prof Arrowsmith has published a White paper on the implications of Brexit for the law on public and utilities procurement, where she briefly considers the alternative models for the future regulation of public procurement in the UK after an exit from the EU. A fuller academic version of the paper is bound to appear in a Brexit special issue of the Public Procurement Law Review. In her White Paper, Arrowsmith provides the skeletal implications that different UK-EU relationships would have in terms of public procurement regulation, most of which point towards a clear need for (broad) continuity of the existing EU-based model.

Along the same lines already drawn by previous commentators (see here), her White Paper stresses the limited scope (and incentive) for a change of regulatory model if the UK is to have full access to the EU single market (under either the 'Norwegian'/EEA model, or the slightly more flexible 'Swiss' approach). She indicates that there is a (theoretical) possibility for the UK to reach a bespoke agreement with the EU that softens the requirements under the current EU Directives, but she also stresses that the stronger indication is that the EU would rather push for a consolidation of existing rules, both because that is the obvious 'off-the-shelf' solution, and due to the different negotiation dynamics between previously unrelated parties (the EU and Switzerland, for these purposes) and between the UK and the EU, which render arguments based on the need to 'learn' about the EU procedures or to reform internal rules to ensure approximation and consistency moot.

She then also assesses two scenarios that, in my view, are only interesting from a theoretical perspective. First, the strict application of the WTO GPA, which leads her to suggest that this option would allow the UK to develop a more flexible procurement regime but, realistically, only in the long run because access to the WTO GPA (of which the UK is not a Member in its own right, but only as part of the EU) would be significantly facilitated by keeping the existing EU-based regulations in place. She also mentions that this option would require the UK to agree on coverage with the EU and all other GPA members, and that 'there ... seems to be no reason why the other GPA parties would want to reopen the existing detailed coverage arrangements with the UK, or vice versa'. She is right to stress that business as usual would be the best way of ensuring quick accession to the GPA by the UK, but the question then arises of what is the advantage of such loose relationship with the EU compared to full access to the single market?--and the simple answer is that, in procurement terms, there is none in the short run and that any long run advantage seriously depends on the way the GPA itself evolves, which is not something we can include in our analysis with any meaningful level of predictability.

Second, and maybe in the only controversial or provocative point of her White Paper, Arrowsmith entertains the idea "that Brexit [c]ould see the UK throw off the shackles of EU procurement law, leaving it free to design its own system", in what she labels "the freedom option", which would derive in case the UK was not able to commit to any trade agreements covering public procurement. Arrowsmith rightly considers this situation unlikely and, as far as I can assess her qualitative comments on its implications, probably undesirable. I think that Arrowsmith's assessment of this situation is however partial because it fails to stress the losses in terms of trade that would derive from such "freedom solution", that I would rather label "the self-destructive, isolationist option".

Beyond the possibility of creating a superior public procurement system, which is by no means guaranteed (as Arrowsmith stresses herself), this solution of "absolute regulatory freedom" would not be viable unless the UK had no intention of keeping a meaningful level of international trade in public procurement markets. This scenario would come together with the possibility for any third country to discriminate against UK producers and exporters in their own procurement, as well as reduced incentives for international suppliers to participate in tenders where their ability to enforce individual rights was more reduced than in other jurisdictions. In general, it does not seem far fetched to consider that there would be less overall competition for contracts tendered by the UK government and, in the end, this would harm the UK taxpayer via higher prices and/or reduced quality of supplies, services and works needed to run public services.

I guess that my broader point is that, in this area of economic regulation, as in any other, arguments based on the possibility to develop a (theoretically) better system from a legal / technical perspective need to be considered together with their economic implications. And, from this perspective, any option that implied limited access to the EU single market and, even more, to the international markets, would impose a very heavy burden on UK's public expenditure. That is why it is important not to isolate technical legal analysis from its broader context in this important debate.

Moreover, and this is not a perspective generally included in Brexit assessments, multilateral investment banks also have a stake in the domestic regulation of public procurement. In case the UK wanted to have any chance of securing international funds for large infrastructure projects (which it may well want to preserve, in order to retain some possibilities of, for instance, EBRD investment in the country), it would still need to have a domestic regulation that complied with standards very close to those of the current EU-based regulatory mechanisms. Otherwise, it could not be out of the question that internationally-funded projects would need to be tendered under special rules in the future, thus not leaving the "freedom option" completely unconstrained.

Similarly, these issues of reduced international competition (with its negative economic effects) and difficulties in continuing to attract procurement-related international investment would arise in case Arrowsmith's proposal for a transition period between Brexit (ie, 2 years after the trigger of Art 50 TEU, which now seems likely to happen in early 2017) and the moment trade agreements were reached, in which she considers that a "sensible and likely interim solution would be to retain the award procedures of the regulations in place, but without provision for enforcement by non-domestic suppliers, pending eventual confirmation, modification/replacement, or total repeal of the regulations, depending on the outcome of trade negotiations and other decisions on how procurement will be regulated after Brexit". In my view, this is a bad idea and the UK would be better off by completely keeping the status quo ante Brexit (including remedies for international tenderers and investors) if it wants to preserve its (diplomatic) options of a swift conclusion of procurement-related trade agreements, as well as preventing disruption in investment and infrastructure projects.

Once again, from a broader perspective and like in most other areas of Brexit-related renegotiation, strategies that not only do not consolidate or grandfather rights, but also seek to (temporarily) restrict rights and guarantees, seem not to be conducive to productive future relationships and there is no reason to believe that such moves would not severely damage the UK's chances of reaching satisfactory agreements for the future. Thus, in my view, Arrowsmith's proposal for a transition period of reduced enforcement rights for non-UK bidders should not be followed.

CJEU ignores commercial reality and sets unjustified contractual boilerplate requirements for contractual modifications (C-549/14)

In its Judgment of 7 September 2016 in Finn Frogne, C-549/14, EU:C:2016:634, the European Court of Justice (ECJ) issued guidance on the requirements (and constraints) derived from the principle of equal treatment in situations where the difficulties in the performance of a contract are such that the contracting authority decides to settle its early termination in a way that implies a material amendment to the initial contract. This case is relevant in the early stages of the new rules on contract modification and termination in Articles 72 and 73 of Directive 2014/24. However, the compatibility between the Finn Frogne Judgment and these new rules raises several questions.

In Finn Frogne, and according to the rather limited facts given in the Judgment, the dispute concerned the contract for the supply of a global communications system common to all emergency response services and for the maintenance of that system for several years, which was awarded after a competitive dialogue. The execution of the contract was subsequently delayed due to difficulties for which neither the contracting authority nor the supplier accepted responsibility (in the terms of the ECJ, both parties disagreed "as to which party was responsible for making it impossible to perform the contract as stipulated", para 10), which eventually led them to enter into a settlement involving the reduction of the contract and each party waiving all other rights arising from the original contract (para 11).

The main point of contention was that the settlement not only included the supply of equipment initially covered by the original contract (a radio communications system), but also the sale of two central server farms which the contractor had itself acquired with a view to leasing them to the contracting authority in performance of the original contract (paras 11 and 19). The settlement was the object of a voluntary ex ante transparency notice and subsequently challenged by a third party.

The legal issue in front of the ECJ was "in essence, whether Article 2 of Directive 2004/18 must be interpreted as meaning that, following the award of a public contract, a material amendment cannot be made to it without a new tendering procedure being initiated, even in the case where the amendment is, objectively, a type of settlement agreement, with both parties agreeing to mutual waivers, designed to bring an end to a dispute with an uncertain outcome, which arose from the difficulties encountered in the performance of that contract" (para 27). Or, in simple terms, whether settling the disputes that had made the commercial relationship between the supplier and the contracting authority non-viable in a way that implied a substantive amendment of the initial contract breached the principle of equal treatment and the obligation of transparency.

In Finn Frogne, the ECJ first took the opportunity to clarify its case law in pressetext (C‑454/06, EU:C:2008:351) and in Wall (C‑91/08, EU:C:2010:182) in the sense of emphasising that a material reduction of the scope of a public contract is equally caught by the restrictions on contract modification as a material extension of the scope of that contract. The reasons for this are as follows:

an amendment of the elements of a contract consisting in a reduction in the scope of that contract’s subject matter may result in it being brought within reach of a greater number of economic operators. Provided that the original scope of the contract meant that only certain undertakings were capable of presenting an application or submitting a tender, any reduction in the scope of that contract may result in that contract being of interest also to smaller economic operators. Moreover, since the minimum levels of ability required for a specific contract must ... be related and proportionate to the subject matter of the contract, a reduction in that contract’s scope is capable of resulting in a proportional reduction of the level of the abilities required of the candidates or tenderers (C-549/14, para 29).

This makes logical sense and is generally linked with the discussion of the division of contracts into lots and how to manage volume-related restrictions of competition for public contracts. However, in the context of a contractual settlement aimed at terminating the commercial relationship between the original (larger) provider and the contracting authority, this would lead to the conclusion that, in a case of breakdown of the commercial relationship implicit in all public contracts, "the principle of equal treatment and the obligation of transparency imply that a contracting authority cannot consider entering into a settlement to resolve the difficulties arising from the performance of a public contract without this automatically giving rise to the obligation to organise a new tendering procedure relating to the terms of that settlement" (para 24), which the referring court considered problematic.

Indeed, in my opinion, taking this position would create situations where the contracting authority is simply in a catch 22 by having to either remain committed to a non-functioning contractual relationship that is not allowing it to perform its public functions to which the contract is instrumental, or having to spend significant funds in the creation of an alternative commercial relationship that may not be the best solution for its needs--particularly if there are economies to be had from preserving part of the original contract or the preparatory actions which the parties had already undertaken in view of its performance.

Regardless of this clear practical difficulty, the ECJ considered that

neither (i) the fact that a material amendment of the terms of a contract results not from the deliberate intention of the contracting authority and the successful tenderer to renegotiate the terms of that contract, but from their intention to reach a settlement in order to resolve objective difficulties encountered in the performance of the contract nor (ii) the objectively unpredictable nature of the performance of certain aspects of the contract can provide justification for the decision to carry out that amendment without respecting the principle of equal treatment from which all operators potentially interested in a public contract must benefit (C-549/14, para 29).

Consequently, it stuck to its previous line of case law in Succhi di Frutta (C‑496/99 P, EU:C:2004:236) whereby any material modification of a public contract requires a new tender (para 38), but placed significant emphasis on the fact that

Although the principle of equal treatment and the obligation of transparency must be guaranteed even in regard to specific public contracts, this does not mean that the particular aspects of those contracts cannot be taken into account. That legal imperative and that practical necessity are reconciled, first, through strict compliance with the conditions of a contract as they were laid down in the contract documents up to the end of the implementation phase of that contract, but also, second, through the possibility of making express provision, in those documents, for the option for the contracting authority to adjust certain conditions, even material ones, of that contract after it has been awarded. By expressly providing for that option and setting the rules for the application thereof in those documents, the contracting authority ensures that all economic operators interested in participating in the procurement procedure are aware of that possibility from the outset and are therefore on an equal footing when formulating their respective tenders (C-549/14, para 37, emphasis added).

Ultimately, the ECJ ruled that

Article 2 of Directive 2004/18 must be interpreted as meaning that, following the award of a public contract, a material amendment cannot be made to that contract without a new tendering procedure being initiated even in the case where that amendment is, objectively, a type of settlement agreement, with both parties agreeing to mutual waivers, designed to bring an end to a dispute the outcome of which is uncertain, which arose from the difficulties encountered in the performance of that contract. The position would be different only if the contract documents provided for the possibility of adjusting certain conditions, even material ones, after the contract had been awarded and fixed the detailed rules for the application of that possibility (C-549/14, para 40, emphasis added).

In my view, the Finn Frogne Judgment must be criticised, at least for two reasons.

First, because it is very difficult to coordinate with the functional approach of Art 72 (and to some extent, 73) of Directive 2014/24 and gives excessive deference to the creation of contractual modification mechanisms. Strictly on the coordination aspect, it is worth stressing that Art 72 seems to be concerned with extensions of the contractual object, but not with its reduction (Art 72(4)(c)), and with qualitative or technical changes that would have allowed other tenderers to participate (Art 72(4)(a)). In the Finn Frogne case, there would have seemed to be more reason to challenge the content of the settlement on the basis that it changed one of those conditions (sale rather than lease of the central server farms) rather than on the change of overall value of the contract. 

Moreover, it is worth stressing that Art 72 also provides significant leeway for the modification of contracts up to 50% of their value (per modification, without a maximum cap) where a diligent contracting authority could not have foreseen the circumstances leading to the need for the contractual modification. Implicitly, the ECJ seems to indicate that every diligent contracting authority needs to foresee the possibility of the commercial relationship breaking down (which may be fair enough), but it also goes on to require a full contractual regulation of how such termination of the contractual relationship needs to unfold.

In that regard, it must be stressed that the requirements for the inclusion of "general" contractual review clauses foreseen in Art 72(1)(a) demands them to be "clear, precise and unequivocal", which may or not be coincidental with the ECJ's requirement for the contractual arrangements to fix "the detailed rules for the application [of] the possibility of adjusting certain conditions"--and which may not be (feasibly) applicable to "termination through settlement" clauses, whereby the parties must necessarily engage in negotiations.

In my view, the ECJ has fallen in the same problematic assumption of the possibility to design "perfect contracts" explicitly and exhaustively regulating all consequences of their (un)foreseeable non-viability or imperfection that also affects the provision in Art 72(1)(a) of Dir 2014/24 [for criticism, see A Sanchez-Graells, Public procurement and the EU competition rules, 2nd edn (Oxford, Hart, 2015) 428], but with the aggravating factor of not acknowledging that they may also be totally ineffective in scenarios where the commercial relationship is broken and, consequently, the parties need to settle, mediate, arbitrate or litigate those consequences regardless of the prior inclusion of such contractual clause.

The second reason why the Finn Frogne Judgment needs to be criticised is because it does not make any effort to attempt to distinguish between settlement conditions that remain strictly within the scope of the original contract and, consequently, only entail its partial enforcement (in its own terms) from settlements which include substantive changes in either their scope or the conditions for (partial) performance. While the first imply a consolidation of the effects already (de facto) created by the original contract, the latter seem to indicate the appearance of different needs of the contracting authority and/or different ways of satisfying them by the supplier. And, in my opinion, while the latter may justify the imposition of strict restrictions and (depending on the circumstances and the proportionality of the requirement) a new tender, the former do not seem to warrant such an approach.

These are issues that will necessarily arise again in litigation concerning the termination of contracts under the combined effect of Arts 72 and 73 of Directive 2014/24 and I would hope that the ECJ will adopt a more analytically rigorous approach when that happens because following the path started in Finn Frogne does not make commercial sense.