The European Commission’s horizontal guidelines are an invaluable tool for practitioners in antitrust compliance work. The team at Unit A1 in DG COMP have done a great job at further developing the guidelines in the new draft that was published for stakeholder comments on 1 March 2022.
Today, the EU & Competition team at Szecskay Attorneys at Law (Hungary) has submitted its consultation response with comments and suggestions on how to further improve the guidance from a perspective of legal certainty and consumer welfare. We have included our main observations in this post.
Need for broader notion of ‘collective benefits’ for ESG agreements
We warmly welcome the Commission’s new chapter on sustainability agreements. As the Commission points out in para. 543, it will enable undertakings to implement strategies and behaviour serving to pursue environmental and social goals such as human rights development.
Our main observation is on the Commission’s position on which consumers or individuals need to benefit in order for restrictions in sustainability agreements to be justified under Article 101(3) TFEU. In this regard, the Commission states in paras. 602-603 that, for a restriction to be justified by collective benefits, there needs to be (a significant) overlap between those individuals who suffer the harm of the restriction and the individuals who benefit. We consider this to be too restrictive a standard.
For example, some environmental benefits may only be realised generations from now. A reduced choice of car engines means short term consumer harm for car buyers today but will benefit the grandchildren and future generations of those same car buyers. There are therefore legitimate collective benefits to be achieved that will not necessarily be enjoyed by any of the consumers who are enduring the sacrifice in the short term.
The same applies to the pursuit of human rights goals outside the EU/EEA. For example, the collective boycott by EU companies of an undertaking established in a third-country that has undisputedly engaged in human rights violations against its workers should be permissible regardless of whether EU citizens obtain any direct benefit from the boycott.
We have provided another, rather topical, example below of why the notion of collective benefits must go beyond the narrow concept of consumer welfare for EU citizens.
Suggested example: Liner shipping companies’ collective boycott of third-country ports
Third-country A has launched a war of aggression against its smaller neighbouring country, Third-country B. Third-country A’s military aggression has been condemned by the European Commission as being against international law and the Council has imposed sanctions against Third-country A in response.
In order to put pressure on the government of Third-country A, the three largest liner shipping companies in the EU agree to stop shipping goods into the EU from all ports in Third-country A.
The collective boycott harms consumers in the EU as consumers can no longer order goods from suppliers in Third-country A. However, the harm is outweighed by the pursuit of peace and human rights for the citizens of Third-country B.
To the extent the Commission does not consider the collective boycott to be legally justified under Article 101(3) TFEU based on a narrow notion of EU consumer welfare, the Commission can, in the alternative, provide a commitment in the Guidelines to the effect that it will, as a matter of enforcement policy, not intervene against such an agreement.
Buyer cartels – Is there an economist in the room?
We respectfully consider that the Commission is deeming buyer coordination to be anti-competitive and illegal beyond what is justified under the consumer welfare standard. As we raised back in 2019 in an article titled “Buyer Beware! – are purchase price fixing cases exempt from the consumer welfare standard?”, the distinction between a (legal) buying alliance and an (illegal) buyer cartel is, at least from the perspective of consumer welfare, one of form rather than substance.
When competitors collude, the effects trickle down through the supply chain ultimately affecting consumers. This is why fixing of sales prices at manufacturer level causes harm to consumers even when consumers do not buy directly from the manufacturer but from a retailer further down in the supply chain. Moreover, sales price collusion is harmful regardless of whether the price fixing is done secretly in a smoked-filled room or openly in the public domain. This is the whole premise for subjecting bidding consortia between competitors to the same standard as opaque bid rigging.
However, if indeed inflated prices due to secret or open sales collusion trickle down and hurt consumers, then – all other things being equal – deflated prices due to secret or open buyer collusion must also presumptively trickle down and benefit consumers. The likelihood of price changes – whether positive or negative – passing through various levels of the supply chain depends on a number of factors – addressed and explained in the Commission’s 2016 Study on the passing on of overcharges. But importantly – all other things being equal – an undercharge as a result of secret purchase price fixing ought to have the same likelihood of benefitting consumers as a secret overcharge has of harming them.
In the example below, we have sought to illustrate why the distinction between a buying alliance and a “buyer cartel” is one of form rather than substance.
Suggested example: Buying Alliance vs. Buyer Cartel
Scenario 1 – Legal buying alliance
Companies A and B are direct competitors and together hold a (sales) market share of 10%. They agree to setup a buying alliance for the procurement of Product X which is a component for their respective and competing downstream products. The buying alliance comprises joint price negotiations with suppliers of Product X but the ultimate purchase contracts with suppliers will remain separate. Their upstream / purchase market share for Product X is also 10% and they both currently pay an average of EUR 120 per unit for Product X.
The parties formalise the buying alliance through an agreement and notify the suppliers of Product X that the parties will be negotiating jointly rather than individually. Before negotiations with the suppliers for the forthcoming year begin, Company A and B agree that they will not accept a purchase price for Product X of more than EUR 100 per unit which they convey to suppliers in a joint meeting. The suppliers agree to selling Product X at EUR 100 to Companies A and B and contracts are subsequently concluded individually.
Scenario 2 – Illegal buyer cartel
Companies C and D are competitors on the same market as Companies A and B in Scenario 1 above. They hold the same upstream and downstream market shares as Companies A and B and also pay an average of EUR 120 per unit for Product X. Companies C and D likewise agree that they will not accept a purchase price for Product X of more then EUR 100 per unit. However, unlike Companies A and B, they do not formalise the buying alliance in an agreement, they do not inform the suppliers of Product X of their cooperation, and they negotiate with suppliers of Product X individually. Following individual negotiations, suppliers of Product X agree to sell Product X to Companies C and D at a price of EUR 100 per unit.
Based on the Commission’s decisional practice and draft Guidelines, Companies A and B have concluded a legal buying alliance whereas Companies C and D have engaged in an illegal buyer cartel. However, the difference in the legal treatment of Scenarios 1 and 2 is not justified by any notion of economics or consumer welfare. All differences are a matter of form, labelling, and transparency. Most importantly, the cost savings achieved by Companies C and D are just as likely to be passed on to consumers as with Companies A and B.
For these reasons, we do not consider there to be any justification based on consumer welfare or any other legitimate competition policy for subjecting so-called buyer cartels to a different legal standard than other forms of buyer coordination, e.g. formalised buying alliances. In practical terms, if a secret fixing of purchase prices between competitors would be legal if it were formalised through a buying alliance agreement and announced to the public, then the secret fixing of purchase prices should not be illegal under Article 101 TFEU either.
Uncertainty is further exacerbated by the Commission’s statement in footnote 180 on page 70 of the draft Guidelines where the Commission states that “However, secrecy is not a requirement for finding a buyer cartel.” This means that there is absolutely no objective, distinguishing feature between a legal buying alliance and an illegal buyer cartel.
The exception: Non-poaching clauses and other labour market restrictions
Our remarks above on buyer cartels do not mean that we endorse lack of enforcement against non-poaching agreements and other agreements between undertakings that cause harm to workers. However, intervention against ‘employer cartels’ as a specific subsegment of ‘buyer coordination’ has a very specific justification and basis in the EU Treaties. Indeed, workers’ rights are protected at Treaty level in several places including Article 3(3) of the Treaty of the European Union, Article 152 TFEU, and Article 28 of the Charter. The special protection of workers’ rights in the context of EU competition policy is also reflected in the Commission’s forthcoming guidance on the right to collective bargaining for solo self-employed.
Antitrust intervention against ‘employer cartels’ should thus be considered the exception where we accept that higher production costs – and thus higher prices for goods – are outweighed by the need to protect workers’ rights. It should not serve as justification for targeting buyer coordination in general where there is no harm to consumer welfare.
On ‘commonality of costs’ as theory of harm in supermarket retail buying alliances
The underlying premise behind the ‘commonality of costs’ theory is that competitors will be able to guess each other’s sales prices if they know enough of each other’s variable production costs, thus resulting in coordinated effects. This reasoning, however, can only apply to the joint purchase of intermediary products and cannot be extended to supermarket buying alliances’ purchase of goods for resale. This is due to several factors:
Firstly, when two supermarkets jointly procure chocolate bars, the insight and information exchange brought about by this cooperation does not make it easier for the supermarkets to estimate each other’s sales prices for apples. This is because the chocolate bars are purchased and resold 1:1 and do not constitute a component for the production of apples. Similarly, the insight into purchase prices for milk does not increase transparency as to the sales price for shampoo, and so on. Therefore, in the Commission’s Example 4 under para. 352 of the draft Guidelines, we do not think it is appropriate for the Commission to factor into its assessment how many consumer goods the supermarkets in question procure jointly.
Secondly, from a practical perspective when it comes to supermarkets, it is questionable how much a buying alliance can really contribute to the transparency of sales prices. The sales prices are already fully transparent in the supermarket stores as well as possibly on their websites. Competitors can easily do legal ‘mystery shopping’ and it therefore seems a bit academic to discuss how insight on input prices can put supermarkets in a position to guess each others’ sales prices.
For these reasons, we recommend that the Commission exclude the ‘commonality of costs’ theory of harm from the analysis of buying alliances in the specific context of supermarket retail.
Joint commercialisation – Too strict approach to non-exclusive supply of competitor?
We request the Commission to clarify whether it considers the supply of a competitor on a non-exclusive basis as being a restriction of competition ‘by object’. If an undertaking supplies a competitor, it inevitably involves some form of joint commercialisation as the competitor is now selling its own as well as a competing product. In this regard, the Commission states the following under para. 366 in the Guidelines:
“That assessment does not change if the agreement is non-exclusive (that is to say, where the parties are free to sell individually outside the agreement), as long as it can be concluded that the agreement will lead to a coordination of the prices charged by the parties to all or part of their customers.”
With this in mind, we ask the Commission to consider the following example.
Suggested example: Fashion brand retailing of competing products
Fashion companies A and B produce competing lines of clothing. Both companies are vertically integrated with own retail stores in which they sell their own clothing lines. Fashion company A’s retail stores do not stock any third-party brands whereas Fashion company B’s stores stock a variety of brands alongside its own.
Fashion company A wishes to sell its clothing line in Fashion company B’s stores on a non-exclusive basis which Fashion Company B agrees to.
The parties are engaging in joint commercialisation of their competing products as Fashion company B is retailing Fashion company A’s products alongside its own. With para. 366 of the draft Guidelines in mind, does the Commission consider this non-exclusive supply agreement to constitute a restriction of competition ‘by object’?
Bidding consortia – Is there are requirement to expand capacity?
In para. 391 of the draft Guidelines, the Commission states the following:
“A joint bidding consortium agreement – irrespective of its legal qualification – does not restrict competition if it allows the undertakings involved to participate in projects that they would not be able to undertake individually.”(underlining added)
We ask the Commission to provide more guidance on what it understands by inability to undertake a project individually. For example, assume for the sake of argument that a company does not have the capacity in terms of machines and workers to perform a particular project. The company could, however, obtain the machines and workers in question if the company’s owner provided a capital injection that doubled the equity of the company or if the company obtained a loan that was guaranteed by the parent company.
How much of a capacity increase and, to this end, how much exploration of financing are undertakings required to carry out in order to comfortably assume that they cannot undertake the project individually, thus allowing them to engage in a bidding consortium?
Moreover, when undertakings assess whether they have capacity for a specific project, are they allowed do disregard capacity that they have reserved for re-occurring customer contracts or projects that are anticipated but not yet concluded?
When R&D agreement becomes a Concentration under the EUMR
We encourage the Commission to clarify how R&D agreements relate to the notion of a concentration under the EUMR, cf. para. 24 of the Commission’s jurisdictional notice. An express clarification is important due to the Commission’s new referral policy under Article 22 of the EUMR concerning ‘killer acquisitions’ in pharma. The Commission considers that it can take jurisdiction even if neither Commission nor Member State thresholds are satisfied.
When planning and executing an R&D project in the context of pharmaceuticals, the assessment under Article 101 TFEU depends on whether any exclusive licensing of the developed API is concluded at the same time as the R&D agreement or at a later stage.
For example, in M.7872 – NOVARTIS / GSK, the Commission found that acquisition of exclusive license for a single therapeutic use of a compound may constitute a ‘concentration’. This requires, however, that the R&D efforts for the compound are at an advanced stage – at least in Phase II / III clinical trials and so relatively close to being put to market (paras 9-11 of the decision).
Accordingly, if the exclusive license is granted before Phase II/III clinical trials, the licensing is non-concentrative and is to be assessed under Article 101 TFEU as part of the R&D agreement itself. If the parties, however, agree exclusive licensing after Phase II/III clinical trials, that license agreement would (likely) constitute a concentration distinct from the R&D agreement and thus with no (or only limited) assessment under Article 101 TFEU required for the license agreement itself.
Counterfactual analysis when analysing cooperation agreements under Article 101(1) & (3)
In recent years, there has been much debate in the EU competition law community on to what extent considering the ‘legal and economic context’ of an agreement for the purpose of finding a restriction ‘by object’ comprises counterfactual analysis. We urge the Commission to make its position clear in this regard for the purpose of analysing horizontal cooperation agreements.
At the very least, we invite the Commission to confirm that the balancing exercise under Article 101(3) TFEU must indeed be based on a comparison between, on one hand, the factual scenario with the restriction in place, and, on the other hand, the counterfactual scenario without the restriction in place. The assertion that counterfactual analysis is required is reflected indirectly in several places in the draft Guidelines, for example in the following:
“An R&D agreement may bring together different research capabilities and combine complementary skills and assets that may result in improved or new products and technologies being developed and marketed than would otherwise be the case.” (Para. 42, underlining added)
“Mobile infrastructure sharing may also allow the emergence of competition that would not otherwise exist” (Para. 298, underlining added)
“A joint bidding consortium agreement – irrespective of its legal qualification – does not restrict competition if it allows the undertakings involved to participate in projects that they would not be able to undertake individually.” (Para. 391, underlining added)
“However, to the extent that the parties would not have been in a position to enter the market for providing laundry services to institutional customers, either individually or in cooperation with a fewer number of parties than the four currently taking part in the agreement…” (Para. 398, underlining added)
The Commission’s statements reflect an acknowledgement of the need for counterfactual analysis. For clarity, however, it would be helpful if the Commission were to make clear the need for counterfactual analysis under Chapter 1 in the draft Guidelines presenting the general methodology for its analysis.
Balancing exercise under Article 101(3) TFEU in ‘by object’ restrictions
As the Commission states in para. 19 of the Guidelines, Article 101(3) TFEU requires the “balancing of restrictive and pro-competitive effects”.
When the Commission finds an agreement or concerted practice to be a restriction ‘by object’, the Commission typically does not – and not need – identify any anti-competitive effects. This leaves the undertakings in a position of having to demonstrate that the agreement’s pro-competitive effects outweigh the agreement’s anti-competitive object.
However, you can only measure ‘effects’ against ‘effects’. You cannot balance ‘effects’ against ‘object’. Therefore, we urge the Commission to expressly state that if a) it is undisputed that an agreement has pro-competitive effects and b) neither the Commission nor the undertakings can identify negative effects, then the undertakings have met the burden of proof under Article 101(3) TFEU regardless of the agreement containing a restriction ‘by object’.