22.03.2022

Circular 16/2022 – EU Competition law – Consultation on draft revised Horizontal Guidelines – Please comment until 12 April 2022 COB

Alexis - Alexis.Waravka@IndependentRetailEurope.eu - +32 2 739 60 92

The European Commission organises until 26 April 2022 a consultation on its draft revised Guidelines on horizontal cooperation agreements under EU competition law (the Horizontal Guidelines or HGL). The HGL are very relevant for groups of independent retailers as they regulate a large part of their internal cooperation (e.g. information exchanges, joint commercialisation, joint purchasing, etc.) as well as external cooperation (e.g. retail alliances). In order to prepare a reply to the consultation, we kindly ask members to send us their comments until Tuesday 12 April 2022 COB.

 

Content of the draft revised EU Horizontal Guidelines

Following the various consultations held over the last 2 years (to which Independent Retail Europe responded), the European Commission published on 1 March 2022 its draft revised Horizontal Guidelines (HGL), on which stakeholders are invited to comment until 26 April 2022.

The draft revised HGL propose to amend the existing HGL by adapting them to the digitalisation of the economy and to the need to pursue sustainability goals, in line with the EU green deal, while also taking into account important case law from the last 10 years.

To this end, it contains important changes in the introduction of the HGL, and in the Chapters on joint purchasing, commercialisation, and information exchange. It also contains a new Chapter on sustainability agreements.

  1. Novelties in the introduction of the HGL

The introduction is restructured to provide more details on certain key aspects which are relevant for the entire HGL. It contains new guidance on the following aspects:

a) How to assess the centre of gravity of agreements which encompass different types of cooperation (para 6-8)

In such cases, although all parts of the HGL are relevant, the part of the HGL which covers the centre of gravity of the cooperation prevails to assess whether the agreement creates by object or by effect restrictions of competition. The centre of gravity is assessed using the ‘starting point’ of the cooperation and the degree of integration of the different functions.

For instance an agreement combining specialisation and commercialisation will be considered as a specialisation agreement as the commercialisation cannot happen without the first component.

b) Definition of undertakings (para 11-14) and competitors (para 17)

Reference to recent case law is added as to how to define the notion of undertakings in the case of parent companies and their joint ventures. Moreover, the draft HGL propose additional guidance to assess whether two undertakings are ‘potential competitors’.

  1. Changes in the joint purchasing guidance (Chapter IV)

a) Types of joint purchasing agreements (para 312)

The guidance clarifies that there are several types of joint purchasing agreements: both, pooling of actual purchases and joint negotiation (of the price, of certain elements of the price, or of other terms and conditions – while leaving actual purchases to individual members), are considered as joint purchasing agreements subject to this Chapter.

Retail alliances are explicitly recognised as a form of joint purchasing, with direct reference to the JRC Report on Retail Alliances in the agricultural and food supply chain.

Comment: This seems to be a positive addition to the HGL. Independent Retail Europe asked for an explicit recognition that retail alliances in their diversity are joint purchasing agreements.

b) Benefits of joint purchasing (para 313)

A novelty is the reference that joint purchasing ‘usually’ aims to create buying power versus ‘large’ suppliers.

Another addition is the explanation that joint purchasing can lead to more variety of products (on top of lower prices and better quality already referred in the existing Guidelines). Moreover, joint purchasing can allow undertakings to prevent shortages/disruptions in the production of certain products.

Comment: These novelties seem positive, as there are references to new efficiencies created by joint purchasing.

c) Clarification of the guidance on ‘by object’ restrictions and distinction between buyer cartels and joint purchasing agreements (para 316-319)

Para 316-318 include a (new) clarification on the difference between buyer cartels and joint purchasing, while para 319 provides a non-exhaustive list of factors to assess in practice the distinction.

In summary, buyer cartels are restrictions by object aiming at:

  • coordinating those purchasers’ individual competitive behaviour on the market or influencing the relevant parameters of competition through practices such as, but not limited to, the fixing or coordination of purchase prices or parts thereof (including agreements fixing wages or not to pay a price for a product) or other trading conditions, the allocation of purchase quotas, the sharing of markets and suppliers”; and
  • influencing those purchasers’ individual negotiations with or individual purchases from suppliers, for example through coordination on the purchasers’ price negotiation strategies or exchanges on the status of such negotiations with suppliers.”

Para 317 and 318 specify that:

  • in a buyer cartel, purchasers coordinate their individual interaction with the suppliers;
  • if purchasers deal individually with suppliers, they should make their own purchasing decisions independently without removing uncertainty among themselves on their future market behaviour;
  • practices of first fixing a purchase price among participants and then each purchaser subsequently negotiate/purchase individually are likely to be considered as a buyer cartel.
  • a buyer cartel may exist if the purchasers exchange commercially sensitive information about their individual purchasing intentions/negotiations with suppliers (e.g. purchase prices, T&Cs, sources of supply, volumes, quantities, quality, timing, delivery and innovation), outside any genuine joint purchasing arrangement that interacts collectively, on behalf of its members, with suppliers.

Para 319 provides two sets of (non-exhaustive) criteria helping to assess joint purchasing agreements that are not a buyer cartel:

  • The joint purchaser has made clear to suppliers that it jointly negotiates and binds its members on T&Cs of their individual purchases or purchases jointly for them. However, this does not require disclosing the identity of the members, especially when they are SMEs. Indirect knowledge of this condition by suppliers is not sufficient.
  • Parties to the joint purchasing have a written agreement defining the form of their cooperation, its scope and functioning (to allow to check compliance ex-post the actual operation) – although a written agreement cannot shield from competition law in itself.

Comment: These clarifications were required by many national competition authorities. They do not seem to raise particular concerns, as they are meant to reflect existing case law. Do you agree? Are the criteria under para 319 raising any practical difficulty for retail alliances?

d) Updated guidance on ‘by effect’ restrictions

Para 325 updates the generic guidance on restrictions by effect on the basis of the case law, indicating that certain contractual provisions fall outside Article 101(1) TFEU if objectively necessary for the joint purchasing agreement. For instance a prohibition for parties to the joint purchasing agreement from participating in a competing arrangement if this can jeopardise the operations and buying power of the joint purchasing agreement. However, exclusive purchasing obligations (i.e. forcing members to purchase all/most of their requirements through the arrangement) may have anticompetitive effects requiring a specific assessment.

e) Effect of market power

The draft revised HGL introduced two new paragraphs (331 and 332) on the potential impact of the buyers’ market power on upstream suppliers.

There is a risk that joint purchasing may harm competition upstream if the parties have a significant degree of buying power on the purchasing market (para 331). E.g. harming investment incentives or forcing suppliers to reduce the range or quality of products. Specifically:

  • risks are higher for large purchasers (that jointly account for a larger proportion of purchases) in particular when dealing with small suppliers (para 332);
  • harm is less likely to occur if suppliers have a significant degree of countervailing seller power (not necessarily amounting to dominance) on the purchasing market (e.g. when selling must-have products).

Moreover, (new) para 333 clarifies that a boycott of non-sustainable products in a joint purchasing agreement should not be considered as a ‘by object’ restriction. However, the ‘effect’ restrictions must be assessed on the basis of the nature of the products, the market position of the purchasers and of the suppliers (e.g. do they have other customers/can they easily start producing sustainable products).

Furthermore, (new) para 334 reformulates the old para 201 (of existing HGL) about the potential foreclosing effect of joint purchasing for competing purchasers. For instance when there are a limited number of suppliers and barriers to entry on the supply side upstream. Para 334 also clarifies that joint purchasing with the objective of excluding actual or potential competitors on the selling market is a restriction of competition by object.

Para 337 slightly amends existing para 212 HGL highlighting that if parties to a joint purchasing agreement are not active on the selling market, the agreement is less likely to have restrictive effects, by stressing that restrictive effects may still happen upstream if the purchasing power is such that it will harm the competitive process of other players (not party to the agreement).

Comments:

  • As expected (and announced by the Commission and in the JRC report on retail alliances) this part of the draft HGL describes in more detail potential harm by effect upstream.
  • We retain as positive the fact that the guidance mentions that restrictive effects are less likely when suppliers have a significant degree of countervailing power, as well as the reference to must-have products in this context.
  • Do you think that some provisions in para 331/332 need clarification? E.g the notion of small supplier? Do these new para 331/332 risk to create complications for ‘national’ alliances, or for groups of independent retailers when jointly buying from small national suppliers?

f) Collusive outcome – Stopping orders/delisting of products

Para 343 recognises explicitly that threats of stopping orders or delisting (temporarily) are “typically part of a bargaining process and may involve collective actions by purchasers”, while recognising that suppliers may use the same technique in a negotiation.

Such threats/actions do not usually amount to a restriction of competition by object, and their possible restrictive effects are not to be assessed separately, but in light of the overall effect of the joint purchasing agreement.

Comment: This recognition seems positive. Do you agree?

g) Pass-on of lower prices to consumers

Para 347 confirms that companies in a joint purchasing agreement have an incentive to pass on at least part of their reduction in variable costs resulting from the joint purchasing (the higher the profit margin from the variable cost reduction, the higher the commercial incentive to pass it on). However, this paragraph provides two exceptions to this general benefit:

  • when members have a significant power on the selling market (this idea was already in the existing HGL para 219);
  • in cases of pure reduction in fixed costs (e.g. lump-sum payments by suppliers), as this does not provide incentives to expand output.

This paragraph should be read in conjunction with para 335 (under the effect of market power), which confirms the idea that if members of the joint purchasing agreement hold a significant market power on the selling market, they will be less inclined to pass on cost reductions to consumers. The risk is particularly high if the agreement limits or disincentives the ability to independently purchase additional volumes in the purchasing market (through the agreement or outside of it). If the agreement contains an obligation to buy all/most of their requirements through the joint purchasing, a specific assessment is required on the possible restrictive effects.

Comment: What are your views on para 335 and 347? Is the second exception mentioned under para 347 (lump-sum payments) an issue? If so, can you provide a counter-argument?

h) New example of joint purchasing and European retail alliances

Para 350 provides a new example of joint purchasing in the form of a European Retail Alliance (ERA).

The example includes seven retailers from different Member States which jointly negotiate with a large brand manufacturer additional conditions (i.e. an additional rebate in exchange of certain promotional services covering the seven Member States where the retailers are selling) for their future supply agreement. The ERA has a market share of no more than 18% on each relevant purchasing market, and each ERA member has a market share between 15% and 20% on their national retail market. The negotiations include hard bargaining techniques, including a decision to temporarily stop ordering certain products (each member deciding exactly which products it stops ordering).

This example is likely to satisfy the conditions of Article 101(1), even if it only negotiates certain conditions because the members are not active on the same selling markets. The temporary delisting will not harm consumers insofar as consumers can purchase substitutable products from competing retailers and will benefit in the long term with lower prices.

  1. Relevant changes in the Chapter on commercialisation

a) Groups of independent retailers and a common price policy online

The Chapter on commercialisation contains light amendments to the existing Guidelines which can be useful for groups of independent retailers willing to set up a joint platform to sell online with a common price policy.

Building on para 246 of the existing Guidelines, para 380 of the draft revised Guidelines now explicitly mentions that joint distribution can generate significant efficiencies “especially for groups of independent retailers, for instance in the case they take advantage of new distribution platforms in order to compete with global or major operators”.

Moreover, as a support to this explicit recognition, the draft revised HGL amend the previous example of a joint platform with fixed price (see para 400 based on para 254 of existing Guidelines) by explicitly recognising that the price fixing in this example generates efficiencies resulting from the common branding online.

Comments:

  • These are very positive changes introduced in the draft revised HGL, as they can help groups of independent retailers to set up under certain conditions a joint platform with fixed prices, especially in cases where they hold a market share of less than 15% (benefits to consumers are presumed to be passed on). The Directorate General for Competition explained us that this change is made to support our request, without changing the fundamental approach of competition law to price fixing.
  • We propose to slightly amend the example used in para 400 to ensure that it covers cases of click and collect, with consumers choosing the store. Would you agree?
  • Do you have any suggestion as to how to improve further this example (without a radical change of the HGL concerning price fixing, given the horizontal nature of the HGL)?

b) Other changes

The draft revised HGL contain other changes about aspects on which we did not take positions so far:

  • specific rules for commercialisation agreements of agricultural products (para 359);
  • additional guidance on the main risks of output limitation in commercialisation agreements (para 364 and 367);
  • further clarifications on the affected markets and on anticompetitive effects (Chapter 5.2.3);
  • a specific section on bidding consortia (Chapter 5.4), and, in particular, on the assessment of consortia agreements between parties that would be able to take part individually in tenders, and on the analysis to be carried out in such cases.

Comment: As we did not take any position so far on these aspects, please let us know if some of the changes raise any particular issue.

  1. Information exchanges (Chapter 6)

a) General aspects

The introduction of this Chapter proposes in para 407 a definition of information (e.g. raw data, pre-processed data, data that has been manipulated to produce meaningful information, and any other type of information).

Para 411 also clarifies the case of information exchanges resulting from legal obligations, stating that Article 101(1) fully applies to these types of exchanges. In such a case, companies obliged to share information must restrict the information to the strict minimum and may have to implement precautionary measures. An example is provided.

b) The case of algorithms

Para 418 (part of the guidance on collusive outcomes in information exchanges) addresses the case of algorithms. ‘Collusion by code’ refers to the deliberate application by competitors of common behavioural coordination algorithms and is therefore a restriction of competition by object.

It should be differentiated from ‘algorithmic collusion’, where the use of algorithms by competitors may increase the risk of a collusive outcome on the market (due to increased market transparency, which allows to detect in real time price deviations and act accordingly to ‘punish’ effectively companies initiating the deviation). To be possible, algorithmic collusion needs a specific design of the algorithms, as well as specific market conditions (high frequency of interactions, limited buyer power and homogeneous products).

Para 432 also states that unilateral disclosure of information that may give rise to breaches of Article 101(1) also concerns “input in a shared algorithmic tool” (therefore indirectly).

Finally, para 436 also provides an example of indirect information exchange through a shared optimisation algorithm (that would take business decisions based on commercially sensitive data feeds or the implementation in automated tools of aligned/coordinated mechanisms of optimisation) between competitors. In such a setting, while the use of publicly available data to feed the software is legal, the aggregation of sensitive data into a pricing tool provided by a single IT company to which various competitors have access could amount to a collusion. This last example may be particularly relevant in a group of independent retailers.

Comment: Do you think that the provisions on ‘algorithmic collusion’ are clear enough? Or are they so general that they may catch very generic cases of algorithms use in (online) retail? Shall we request to add further criteria? (If so, which ones?) Any other comment concerning the provisions on algorithms?

c) Example of commercially sensitive information

Para 424 provides a list of information which are considered to be particularly commercially sensitive, and the exchange of which is a by object restriction:

  • price and pricing intention;
  • current and future production capacities;
  • intended commercial strategy;
  • arrangements to current and future demand;
  • future sales;
  • current state and business strategy;
  • future product characteristics which are relevant for consumers;
  • positions on the market and strategies at auctions for financial products.

Comment: This seems to be a positive clarification.

d) Age of the data

Para 431 provides additional element on how to appreciate if data is old enough not to be considered as sensitive, stressing the need for a case-by-case assessment depending on the specific market. It provides an example where information up to one-year-old will not be considered as ‘historic’ but sensitive (i.e. in sectors relying on consumer preferences to optimise brands’ strategic decisions).

e) Measures to limit/control how data is used and accessed

Para 440 to 442 provide guidance on measures to control the access to the data.

Para 440 in particular provides guidance on:

  • the setting up of ‘clean teams’ (i.e. a restricted group of individuals not involved in the day-to-day commercial operations and bound by strict confidentiality protocols with regard to the commercially sensitive information);
  • data pools, where participants can access their own information and the final aggregated information of other participants. In such a case, measures shall be taken to ensure that a participant is unable to obtain commercially sensitive information from other participants, while giving access only to the information necessary for the implementation of the legitimate purpose of the pool. Management of the data pool can be given to a third party with strict confidentiality rules (this paragraph is very relevant for groups of independent retailers).

Lastly, para 441 provides that where the information exchange is strategic for competition and covers a significant part of the relevant market but does not pose a risk of collusive outcome, such information exchange may be permissible only if the data is accessible in a non-discriminatory manner to all undertakings active on the relevant market (as otherwise, there is a risk of foreclosure).

Comment: Independent Retail Europe was asking for more detailed guidance on data sharing, as it may facilitate such information exchange inside a group of independent retailers. The inclusion of new detailed guidance is therefore positive. Do you have comments or request of clarifications about how data exchange can be made using clean teams and data pools in the context of a group of independent retailers?

  1. Sustainability agreements (Chapter 9)

The draft revised HGL contains a new Chapter on agreements between competitors which pursue a sustainability goal (the definition of sustainability used in para 543 reflects the EU’s vision for sustainability), and how such agreements will be assessed. The guidance recognises that competition law enforcement contributes to sustainable development (para 544). However, it recognises that market failures not addressed by legislation can be mitigated through sustainability agreements between competitors (para 546).

a) Sustainability agreements not raising competition concerns (para 551-554)

The following agreements do not raise competition concerns:

  • agreements that do not affect parameters of competition (e.g. proven quantity, quality, choice, innovation);
  • agreements that do not concern economic activity of competitors but their internal corporate conduct (e.g. agree measures to eliminate single use plastics in their buildings, etc.);
  • agreements creating databases of information on sustainability-friendly suppliers or distributors that do not require the parties to purchase/sell to/from those suppliers/distributors;
  • agreements to organise awareness campaigns for consumers (not amounting to joint advertising of a product).

b) General principle – classification of the type of agreement

Para 556 provides that when a sustainability agreement takes the form of a specific type of cooperation covered by a specific section of the HGL (e.g. joint purchasing, commercialisation, etc.), it should be assessed using the guidance for this type of agreements. Sustainability will be taken into account to determine any by object of by effect restriction (para 560). Moreover, when such agreements pursue specific by object restrictions, parties to the agreements will need to demonstrate that the goal is truly to achieve sustainability, and not to hide an anti-competitive goal (para 560).

c) Sustainability standardisation agreements (Chapter 9.3.2)

In such agreements, competitors agree to adopt and comply with certain sustainability standards (see para 561 to 567). While they may bring benefits to consumers, if the agreement covers matters related to how to translate increased costs from the adoption of the standard into increased sales prices, such agreements restrict competition by object (para 571).

For the assessment of restrictive effects on competition, a ‘soft safe harbour’ is created (para 572) if the following conditions are met:

  • first: the procedure to develop the standard is transparent and all interested competitors can participate;
  • second: the standard should not impose on undertakings that do not wish to participate in the standard an obligation (directly or indirectly) to comply with the standard;
  • third: participating undertakings should remain free to adopt for themselves a higher standard;
  • fourth: parties to the standard should not exchange commercially sensitive information not necessary for the development, the adoption or the modification of the standard;
  • fifth: effective and non-discriminatory access to the outcome of the standardisation procedure should be ensured, including effective and non-discriminatory access to the requirements and the conditions for obtaining the agreed label or for its adoption at a later stage;
  • sixth: the standard should not lead to a significant increase in price or reduction of choice;
  • seventh: there should be a monitoring system to ensure compliance.

If some of these conditions are not met, an assessment of the agreement’s effect on competition is needed. If a restrictive effect is found, an exemption can still be found on the basis of the four criteria mentioned in Article 101(3). For instance efficiencies, indispensability, pass-on to consumers, no restriction of competition.

Concerning the indispensability criteria: the guidance clarifies that:

  • where EU or national law requires to comply with concrete sustainability goals, agreements to achieve them cannot be considered indispensable (para 583);
  • this criterion is met if the parties can show that consumers cannot objectively balance the future benefits (e.g. improved quality/innovation) in terms of sustainability against the immediate harm they suffer (e.g. increased prices) (para 586);
  • the obligations in the agreements should not go beyond what is required to achieve the aim of the agreement (para 587).

Concerning the ‘pass-on to consumers’ (para 588-609), the guidance distinguishes three types of pass-on:

  • the individual use value benefits (resulting from the use of the product). For instance improved product quality or variety or price decrease (para 590 to 593);
  • the individual non-use value benefits (indirect benefits resulting from the consumers’ appreciation of the impact of their sustainable consumption). For instance using a certain type of product because less harmful to the environment, or willingness to pay more for a product harvested sustainably (para 594 to 600);
  • collective benefits (e.g. when necessary to internalise negative externalities, irrespective of the individual benefits for consumers). For instance agreements to phase out pollution when consumers are not willing to pay a higher price (para 601 to 608). Para 606 explains which conditions must be met to account for collective benefits (e.g. describe clearly the benefits with evidence that they already occurred or are likely to occur, define the beneficiaries, consumers must overlap with the beneficiaries, demonstrate what part of the collective benefits outside the relevant market accrue to the consumers of the product on the relevant market). In this context, evidence based on public authorities’ reports or prepared by recognised academic organisations are particularly valuable.

Concerning the ‘restriction of competition’ criteria: some degree of residual competition must remain (para 610), through competition on at least one important parameter of competition such as price, variety or quality (para 611).

The draft revised HGL provides five examples of such sustainability agreements (some meeting the conditions of Article 101(3), some not) in para 617 to para 621.

Comment: Do you have specific comments or suggestions concerning sustainability agreements?

 

Reply to the consultation – Feedback welcome until 12 April 2022 COB

In order to prepare a constructive reply to the consultation, we kindly ask members to send us their feedback on the draft revised HGL, and in particular on the questions made throughout this Circular until 12 April 2022 COB.