Recent Case Law on EU Institutional Procurement under the Financial Regulation (II): Abnormally Low Tenders

Before the summer recess, the General Court adopted two interesting decisions on public procurement carried by the EU Institutions. One concerns the debarment of tenderers that have been found to breach EU procurement rules and negatively affect the financial interests of the Union (T-151/16). The other concerns the obligation to state reasons in the context of allegations that a tender is abnormally low (T-392/15). The first case was discussed in a previous post, while this blog now discusses the second case.

In its Judgment of 4 July 2017, European Dynamics Luxembourg and Others v Agence, T-392/15, EU:T:2017:462, the GC assessed once more the limits of the obligation incumbent upon contracting authorities to state reasons in the context of an assessment of an apparently abnormally low tender. The case is decided under the rules of EU Institutional Procurement (ie the Financial Regulation and Rules of Application), but its basic principles seem to me to be also of relevance for procurement covered by the 2014 Public Procurement Package and, in particular, Article 69 of Directive 2014/24/EU.

The distinctive peculiarity of the case is that the challenge concerns the retendering of lots of a previous procedure that had been partially cancelled. As a result of the cancellation of the original procedure post-evaluation and the disclosure of information in the debriefing linked to that tender, participants in the retendering had the advantage of availability of substantial pricing information concerning their competitors (which is certainly one more reason to take confidentiality of information in these processes very carefully, in particular where disclosure of information allows for a 'reverse engineering' of the prices offered by other tenderers--see the discussion in A Sanchez-Graells, 'Transparency in Procurement by the EU Institutions' (August 16, 2017). As a result of having that information, one of the tenderers challenged the award decision in the retendering on the basis that some of the values of the preferred tenders were 'excessively low' and that the contracting authority, having access to that information, was under a duty to provide explicit reasons why it did not consider the tenders received in the second run abnormally low (see paras 68-69) .

In order to decide on the dispute, the GC first recasts the existing provisions and case law on the duty to provide reasons as part of the right to good administration under Article 41 of the Charter of Fundamental Rights of the EU (paras 72-80) and stresses that 'the obligation to state reasons for an act depends on the factual and legal context in which it was adopted' which in the specific requires that 'account ... be taken of the ... regulatory framework applicable in the present case governing abnormally low tenders' (para 81). The GC then discusses such regulatory framework (paras 82-90), stressing that previous case law 'has held that the contracting authority’s obligation to check the seriousness of a tender arises where there are doubts beforehand as to its reliability, bearing in mind that the main purpose of that [investigation] is to enable a tenderer not to be excluded from the procedure without having had an opportunity to explain the terms of its tender which appears abnormally low. Thus, it is only where such doubts exist that the evaluation committee is required to request relevant information on the composition of the tender, before, if necessary, rejecting it' (para 85, references omitted). This creates a two-stage approach to the analysis, where first the authority needs to assess if there is an appearance or suspicion of abnormally low values and,only in that case, engage in the inter partes detailed investigation that will trigger the need for additional justification of its final position on the abnormality or not of the tender. In the analysis of the GC, thus, whether there is a duty to investigate in detail and the extent to which reasons need to be given depend on whether 'there is evidence which arouses a suspicion that a tender may be abnormally low' (para 89).

Elaborating on this, the GC establishes that 'the contracting authority need, in the first stage, only carry out a prima facie assessment of the abnormally low character of a tender, that its duty to state reasons is limited in scope. To require the contracting authority to set out in detail why a tender does not appear to be abnormally low does not take into account the distinction between the two stages of the examination' (para 92). Thus, in even clearer terms, 'where a contracting authority accepts a tender, it is not required to state explicitly in response to any request for a statement of reasons ... [why] the tender it accepted does not appear to it to be abnormally low. If that tender is accepted by the contracting authority, it follows implicitly, although not necessarily, that the contracting authority considers that there was no evidence that that tender was abnormally low. However, such reasons must be brought to the attention of an unsuccessful tenderer which has expressly requested them' (para 93).

In my view, this test is helpful, as it sets a clear balance of duties between the contracting authority -- a duty to assess whether there is evidence to support a suspicion of abnormality, but no duty to justify why it does not consider that this is the case in each and every single instance -- and the tenderers -- which can express their concerns about the appearance of abnormality of competing tenders and demand that the contracting authority clarifies the reasons for its disagreement, where prompted to do so. In my view, this is a useful and practical approach generally applicable to procurement, both under the rules of EU Institutional procurement and that covered by the 2014 Public Procurement Package.

Recent case law on EU Institutional Procurement under the Financial Regulation (I): Self-Cleaning

Before the summer recess, the General Court adopted two interesting decisions on public procurement carried by the EU Institutions. One concerns the debarment of tenderers that have been found to breach EU procurement rules and negatively affect the financial interests of the Union (T-151/16). The other concerns the obligation to state reasons in the context of allegations that a tender is abnormally low (T-392/15). This blog discusses the first case, while a subsequent post comments on the second.

Judgment of 27 June 2017, NC v Commission, T-151/16, EU:T:2017:437, is concerned with the registration in the Early Warning and Detection System database (ie the registry of tenderers and contractors debarred from EU Institutional procurement, currently relabelled as Early Detection and Exclusion System, EDES) of tenderers that have been found  to have committed serious breaches of contractual obligations--in this case, as established by OLAF, the simulation of procurement procedures for the acquisition of equipment ultimately funded by the EU. The case is affected by the additional difficulty that the rules controlling EU Institutional procurement (ie the Financial Regulation and its Rules of Application) were modified in the period between the irregularities were committed (2008 and 2009) and the time of the imposition of the sanction of debarment by the Commission (which crossed over between 2015 and 2016). This triggered two legal complications in terms of retroactivity of most favourable/lenient substantive rules: first, the effect that needed to be given to a reduction in the maximum period of debarment from 5 to 3 years; second, the possibility to neutralise a ground for exclusion on the basis that the affect undertaking had taken sufficient remedial measures demonstrating its reliability (ie had self-cleaned). On top of that, there were procedural complications due to the revised procedures leading to registration in EDES, which currently require a panel opinion that was not part of the pre-2016 procedure for the registration in the Early Warning and Detection System database.

On the procedural point, which the GC examines first, the dispute hinges on the fact that the debarment decision was adopted on 28 January 2016 (which would have required an involvement of the EDES panel, active from 1 January 2016; see para 32), but the Commission considered the administrative procedure 'completed' on 17 December 2015 (thus subjecting it to the 'no-panel' procedure in force until 31 December 2015; see para 34). This ground is ultimately dismissed by the GC on the basis that there is no reason to establish the retroactive application of the procedural rules to investigations started before 1 January 2016, which would 'imply recommencing the preliminary procedure completed properly before that date, in particular having regard to compliance with the adversarial principle' (para 43).

This decision goes against the general principle that new procedural rules that do not contain specific transitional provisions accompanying the fixing of their general application date also apply to on-going/pending procedures (see para 36). The decision is based on an exception to such created in the Judgment of 8 November 2007, Andreasen v Commission, F-40/05, EU:F:2007:189, whereby that rule can be excluded to avoid 'the retroactive annulment of procedures or procedural steps which complied with the rule in force when they were completed' (para 38; see also para 43 of T-151/16).

What I find interesting, though, is that the GC considers that such assessment is not altered '[e]ven if the introduction of that panel was intended to strengthen the rights of the defence of parties contracting with the Union who may be subject to a penalty under the Financial Regulation' (ibid). In my view, this is a very ad hoc finding, which the GC reaches only because it considers the pre-2016 rules already sufficiently protective of individual rights of the affected undertaking, and to have been adequately followed in the specific instance. Had this not been the case (eg, had the previous procedure been seen to fall short of complying with the adversarial principle), the decision by the GC may well have been the opposite. Thus, on this point, the decision of the GC seems difficult to extrapolate to other contexts and the exception that seems to derive from Andreasen and now NC needs to be taken with a pinch of salt.

On the substantive points, first concerning the retroactivity of a more lenient rule allowing for self-cleaning, the GC takes the view that the possibility to self-clean and thus exclude debarment makes the new rules clearly more favourable (para 57). On that basis, the GC takes issue with the fact that the Commission took into account remedial measures for the purpose of setting the duration of the exclusion below the maximum exclusion period (initially at 2 years, later reduced to 18 months) but did not assess it with a view to completely exclude the debarment on the basis of satisfactory self-cleaning. As the GC put it: 'Although the contested decision shows that the remedial measures taken by the applicant were taken into account to determine the duration of the exclusion imposed, no reason is given in that decision as to why those measures were insufficient to satisfy the conditions' for an operator that has taken certain remedial measures demonstrating its reliability not to be excluded from the contracts and grants of the Union (para 58). Second, and along the same lines, on the assessment of the implications of a reduction the maximum debarment period from 5 to 3 years, the GC considers that the new spread of debarment times should have been explicitly taken into account by the Commission (paras 59-60). This eventually leads to an annulment of the debarment decision (para 63).

In my view, this strict approach adopted by the GC on the basis of the guarantees enshrined in Article 49 of the Charter of Fundamental Rights of the EU and interpretive case law (paras 53-55) comes to strengthen the procedural guarantees involved in the adoption of debarment decisions. Extrapolating this to procedures not covered by the rules on EU Institutional procurement, but rather by the 2014 Public Procurement Package and its transposition at domestic level by the Member States, it seems clearer than ever to me that there is a need for the revision of the remedies directive in order to ensure the effectiveness of the same level of protection--as discussed, over a year ago, in A Sanchez-Graells, '"If It Ain't Broke, Don't Fix It"? EU Requirements of Administrative Oversight and Judicial Protection for Public Contracts' (August 11, 2016), to be published in S Torricelli & F Folliot Lalliot (eds), Administrative oversight and judicial protection for public contracts (forthc). Available at SSRN: https://ssrn.com/abstract=2821828.

The curious case of the open envelope inside the envelope - a propos GC's Gfi PSF v Commission Judgment (T-200/16)

The General Court of the Court of Justice of the European Union (GC) has issued Judgment in case Gfi PSF v Commission, T-200/16, EU:T:2017:294 (available in French only). This is a curious case about the physical formalities imposed in a procurement procedure carried out by the European Commission (Publications Office), which required a set of envelopes to be enclosed in multiple layers so as to avoid their tampering prior to the official opening of bids. It is also interesting because it raises some issues around the difficulties in the fact finding of processes dominated by formal documentary evidence.

I am also afraid that the factual circumstances of the case are probably rather common in practice (they remind me of the submission of a tender for a multi-million concession contract that had to be submitted in suitcases locked with padlocks in Mexico some 10 years ago), and I guess that the case also reflects some of the advantages that could be gained by a proper migration to e-procurement (or at least to electronic submission of tenders).

In the case at hand, tenderers had to prepare their tenders in two separate envelopes. An envelope (a) containing their technical offer and an envelope (b) containing their financial offer. Both envelopes then had to be enclosed in a third envelope (c) marked as "Tender - not to be opened by the internal mail service" and placed in a fourth envelop (d), which had to be sent by registered mail or courier service, or be submitted in the offices of the European Commission as indicated in the letter of invitation to tender.

Gfi PSF prepared its tender in accordance with these instructions and sent it to the European Commission via UPS. However, t is worth noting that, inside envelope (c), Gfi PSF did not only include envelopes (a) and (b), but also several binders including additional information. The tender was submitted in time and there is an electronic receipt issued by UPS with a signature from a Commission official. However, an acknowledgement of receipt was also prepared by the mail service of the European Commission indicating that the offer had been received, but not in good state, and also including the following:

two headings, relating respectively to the "first container" and the "second container". In the section on the first container, the pre-printed indications "open" and "damaged" have both been checked. In the section on the second container, the same information was also checked. This last heading also contains the words "did not include the words "Do not open by the mail office"" and "there were no double envelopes", which were not checked (T-200/16, para 7, own translation from French).

After proceeding to the formal opening of the tenders, Gfi PSF's was rejected on the basis that the tender was already open when the contracting authority received it, which is a cause for rejection under Art 111(4)(b) of the Financial Regulation. After Gfi PSF challenged this decision and asked for additional details, the European Commission wrote a letter indicating that

even if the electronic receipt [issued by UPS] did not contain any remarks as to the status of the consignment containing the applicant's tender, this was because of the technical constraints of the terminal used by the courier acting for UPS. The [Commission's] note of receipt acknowledging the damage of the consignment was signed jointly by the same courier and a representative of the [Commission]. Copies of this note and a photograph of the said item were annexed to the" Commission's letter (T-200/16, para 13, own translation from French).

After inspecting the envelopes still in the custody of the Commission, counsel for Gfi PSF challenged the time, the content, and the probationary value of this acknowledgement of receipt issued by the Commission. It also sought to prove that the document had been issued unilaterally by the Commission because it was not clear that UPS had signed the note (which is however later proven wrong by the GC, see paras 53-54), and submitted that in any case the state of the offer resulted from improper handling by the Commission's mail services, rather than as a result of defective compliance with the physical requirements for the submission of the offer by Gfi PSF at the time of its expedition.

The Commission opposed this interpretation of the documents (in particular the lack of signature by a UPS representative) and the physical evidence (ie damaged envelopes), and provided additional evidence downloaded from UPS' tracking webpage, where the following remarks appeared from entries logged in the weeks following the delivery of the package to the Commission:

"[t] he goods are lacking. UPS will notify the additional details to the consignor / goods entrusted to the consignee "; and ... " [t]he damage to the contents of the packages has been reported [;] We will notify the consignor / We are investigating the claim for damages" (T-200/16, para 18, own translation from French).

Overall, then, the dispute concerns the factual circumstances of the delivery, the documentation of its receipt, as well as the ensuing investigation of the Commission's claim that the package was delivered by UPS. It is hard to imagine the physical state of the envelopes (it would have helped to have the pictures attached to the GC's Judgment), but it is certainly plausible that the envelopes where stacked in such a way that opening envelop (d) also ripped envelop (c) (particularly if they were of similar sizes). Be it as it may, the reasoning of the GC is interesting beyond the specific issues leading to the discrepancy in the state of the offer at the time of submission and of the formal opening of the tenders, which exonerates the Commission from any responsibility.

In the GC's view

... the acknowledgement of receipt is of significant probative value, since its content is attested by the signatures, on the one hand, of a person subject to the contracting authority and, on the other hand, by a third party not directly involved in the procurement tender, but rather acting indirectly on behalf of the tenderer whose tender is considered irregular by the contracting authority. However, the acknowledgement of receipt contains indications that both the first and second containers were both open and damaged, and that the second container displayed the words "Not to be open by the mail office" and contained a double envelope (T-200/16, paras 57-58, own translation from French).

This would create a difficulty in establishing the moment in which envelop (c) had been opened despite including the prescribed label against it. However, given the very peculiar circumstances of the case, where envelop (c) contained not only envelops (a) and (b)--which may not have been compromised--but also the binders including additional information, the GC found a way out by adopting a functional approach to the rules in Art 111(4)(b) Financial Regulation from the perspective of the integrity of the process. In that regard, it stressed that the submission instructions and Art 111(4)(b) aimed to ensure the confidentiality of all tenders until they are simultaneously open. From that perspective, the factual circumstances of the case led to the assessment that

On the one hand, the applicant does not claim that the binders were themselves placed in a closed envelope, the binders being visible in the photograph to which it refers, annexed to the letter of [the Commission sent during the debriefing and complaints procedure]. Furthermore, the fact that the binders were placed by the applicant in an envelope on which it indicated that it had affixed the words 'invitation to tender - not to be opened by the courier' [ie envelop (c)], which is established by the acknowledgment of receipt, shows that in the applicants' own view, the binders contained documents constituting its tender. Consequently, and due to the fact that it must be held that the outer envelope and the intermediate envelope of the item containing the applicant's tender had been presented open on the premises of the [European Commission], it must be found that certain data forming part of the applicant's offer were directly accessible. Consequently, it appears that the applicant's tender was submitted in such a way that its confidentiality, as required by Article 111 (1) of the Financial Regulation, was not guaranteed, as it was "already open" within the meaning of paragraph 4 (b) of that Article. Consequently, this offer was regularly rejected (T-200/16, paras 65-66, own translation from French).

This is important because the GC has no interest in (and probably no possibility to) establishing the way in which envelope (c) came to be open despite it being labelled as not to be open by the mail service. This fits with the burden of proof derived from a claim for damages based on Art 340 TFEU, which was the relevant underlying legal basis for this case. However, this leaves important questions unanswered, such as what would have happened if the binders were sealed in envelops, so that the opening of envelope (c) would not have made any of the contents of the offer directly accessible. In that case, the rejection of the offer on the basis that it had already been opened would be very problematic and would probably have required further investigation of how this came to be.

On balance, it seems that the GC places the burden of ensuring that the offers remain confidential on the tenderers, at least implicitly, by supporting a broad approach to the rejection of offers which confidentiality may have been compromised. As a matter of general trend, this seems preferable to the opposite. However, this also shows the unavoidable limitations of paper-based procurement procedures. Had the Commission been running an e-procurement (or at least electronic submission) process, this situation could have been easily avoided. It seems that, once more, the adaptation of procurement (and administrative activity, more generally) to new technologies cannot come quickly enough.