On April 27, 2020[1] the Council of State upheld two judgments issued by the Regional Administrative Court of Lazio (“TAR Lazio”) in 2016,[2] which had annulled an ICA decision fining Chef Express S.p.A. (“Chef Express”) and My Chef Ristorazione Commerciale S.p.A. (“My Chef”, and together with Chef Express, the “Companies”) for alleged bid rigging in the market for food catering services in Italian motorway restaurants (the “Decision”).[3]

In particular, the Council of State agreed with the TAR Lazio that the ICA had not adequately proved a collusive scheme.


The Decision

In the Decision, the ICA stated that the Companies had entered into a secret anticompetitive horizontal agreement in breach of Article 101 TFEU, aimed at coordinating their conduct in 16 tenders for the sub-licensing of the provision of food catering services on the Italian motorway network managed by Autostrade per l’Italia S.p.A. (“ASPI”). As a consequence, the ICA fined Chef Express and My Chef €8.4 million and €4.9 million, respectively.

The opening of the ICA investigation followed a complaint submitted by Roland Berger Strategy Consultants S.r.l. (“Roland Berger”), the advisor appointed by ASPI to organize the tender procedures, which reported alleged “unusual and symmetrical bidding behavior by two participants” in the procedures launched in 2013 for the provision of catering services in 16 out of 43 motorway restaurants.

In the ICA’s view, the Companies implemented symmetrical bidding behavior aimed at rigging the competitive dynamics of the 16 tenders. As a result, the tenders were awarded equally to the Companies. According to the ICA, the tender awarding could not be the natural result of rational business choices independently made by each of the Companies, but was due to an anticompetitive scheme, which allowed the Companies to get the sub-licensing contracts on more advantageous conditions.

As allegedly proved by numerous documents (including tables containing simulations of the bidding scenarios, found at the Companies’ premises), the Companies planned to implement a complex bidding mechanism to support each other’s offers, taking into account the mathematical formula used by Roland Berger to assess the quality of bids. In particular, where a party offered economic terms much better than the starting economic conditions provided for by tender rules, the difference between the scores attributed to other financial bids were less marked; therefore, in these cases, competition among bidders was based almost entirely on the technical aspects of the offers. Accordingly, the ICA found that where one of the Companies was interested in getting a particular lot, it submitted a strong offer from a technical viewpoint, coupled with a weaker financial offer; by contrast, the other Company submitted for the same lot an offer with weaker technical aspects and a stronger financial component.

The judgments issued by the TAR Lazio

The Companies challenged the Decision before the TAR Lazio. They argued among other things that the ICA’s findings were based on an incomplete and inadequate investigation, as well as conflicting documents, which were misinterpreted by the ICA to support its finding that the Companies’ behavior was “intrinsically anomalous”. According to the Companies, the ICA breached their rights of defense and to a fair trial.

In April 2016, the TAR Lazio quashed the Decision, on the ground that the ICA had failed to provide sufficient evidence of the alleged collusive scheme between My Chef and Chef Express.

In particular, the TAR Lazio held that: (i) the Decision did not adequately explain the legal reasoning leading the ICA to conclude that the Companies’ bidding behavior could not be explained by alternative (and lawful) justifications; (ii) the ICA did not base its findings on objective elements, but merely relied on average data provided by Roland Berger; (iii) the Decision did not take into account the evidence provided by the Companies during the investigation to demonstrate that their bidding behavior was economically rational; (iv) finally, the ICA did not prove specific contacts between the Companies that could have allowed them to devise the alleged anticompetitive scheme.

The Council of State’s ruling

Following the ICA’s appeal against the first instance judgments, the Council of State fully confirmed the conclusion reached by the TAR Lazio.

At the outset, the Council of State restated the well-established principle that mere similarity between the bidding behavior of independent operators (particularly in the context of oligopolistic markets, which have fewer players and may naturally be more inclined to align their conduct and strategies) cannot, in itself, prove the existence of an anticompetitive agreement or practice, unless there is no alternative rational explanation for such parallelism. Accordingly, the ICA is required to prove that the parallel conduct may not be the result of plausible and independent business choices. However, where there is evidence of contacts or exchanges of information between the market players, the burden of proof is reversed, and it is up to the investigated companies to show that their conduct is not the result of anticompetitive coordination.

In light of these principles, the Council of State held that the Decision was not based on sufficient evidence.

First, the Council of State criticized the fact that the ICA merely relied on average data provided by Roland Berger, instead of analyzing the offers submitted by the Companies in all of the 48 tenders in which they participated.

Second, the Council of State noted that the ICA completely ignored the evidence submitted by the Companies during the proceedings. In particular, the Decision did not make any reference to the economic reports presented by the Companies to demonstrate that their offers could not result in the rigging of the tenders at issue. According to the Council of State, this conclusion was supported by the statements made by Roland Berger during the investigation, according to which, even absent the alleged “supporting” offers submitted by each of the Companies, the outcome of the tenders would have remained the same, i.e. they would have been awarded to the same players.

Third, the Council of State held that the documentary evidence gathered by the ICA to support its findings, which mainly consisted of papers found at the Companies’ premises, was not convincing, particularly because it was unclear whether these analyses of the possible bidding scenarios had been drafted before or after the awarding of the tenders (when the offers submitted by the other bidders were publicly known).

Finally, the ICA did not adequately demonstrated the absence of alternative explanations to the alleged coordinated outcomes, as it limited itself to referring to “the total symmetry between the bidding behavior of the Companies, […] as well as the total symmetry of the outcomes of such tenders”. According to the Council of State, the Decision lacked robust and convincing reasoning regarding the existence of the alleged collusive scheme.

[1]              Council of State, Judgments Nos. 2673 and 2674 of April 27, 2020.

[2]              TAR Lazio, Judgments Nos. 3982 and 3983 of April 1, 2016.

[3]              ICA, Decision of April 22, 2015, No. 25435, Case I775 – Procedure di affidamento dei servizi di ristoro su rete autostradale ASPI.