Burgers, Booze and Competition Law
Last time I wrote on Uber, I mentioned the recent ECJ ruling on Scottish legislation setting a minimum price per unit of alcohol. These measures are intended to address binge drinking, which, I understand, has become a serious issue in Scotland and the whole of the UK. The Guardian published the other day an interesting – and shocking – piece on the impact of binge drinking on the National Health Service.
What would be the status of similar measures under competition law? I thought of a couple of hypotheticals to show that, under a proper construction of Article 101(1) TFEU, private measures of this kind would not necessarily be problematic.
Example 1: It is well-known that nutrition is key to prevent heart disease, even to those (including Alfonso and myself) who enjoy a burger at Five Guys. Those with a history of cardiovascular problems, or those who want to prevent them might find it a challenge to eat out. Imagine a group of restaurants that create a common label, ‘Heart Healthy’. These restaurants commit to follow the guidelines issued by cardiologists and would give consumers the guarantee that healthy menus would be available.
I do not think anyone would argue that an agreement of this kind is akin to a cartel. It is closer to a standard-setting agreement or a certification mechanism. But please note that the restaurants would go a long way in the coordination of their conduct. They would agree, inter alia, not to add salt to the dishes, not to use red meat (or not to use meat at all) and not to use certain types of oil (or maybe not to use oil at all). Moreover, they might agree to limit the size of their portions (cardiologists insist that portions in some restaurants are too big; Alfonso is not convinced). All of these restraints, examined in isolation, look like those found in cartel agreements, but the rationale of the agreements and the context are clearly different.
Example 2: What if a group of small, independent grocery stores do something similar? A group of retailers could create a common label, ‘Ethical groceries’, and commit, inter alia, to strict animal welfare standards. For instance, they would not sell products like foie gras. They could also commit not to contribute to the negative impact of binge drinking on the community and agree not to sell alcohol in their premises. The label would allow consumers to recognise the stores that share certain views on business practices. A certification scheme of this kind could not be more different from a cartel. Again, I struggle to see how it could be argued that an agreement of this kind is restrictive of competition by object, but I would love to read the views of those who disagree.
Example 3: I have just argued that an agreement between small, independent grocery stores not to sell alcohol and other products is most probably not a restriction of competition by object. What if the same independent stores decide to set minimum prices for alcohol? If an agreement not to sell a product in this particular context is not a cartel, I fail to see why price-fixing should be treated differently. If anything, an agreement not to sell a product at all should be treated more harshly than an agreement setting minimum prices.
The reason why this agreement between small, independent grocery stores should not be treated like a price-fixing cartel seems clear. These small retailers would not have significant market power. And we know very well that a cartel worthy of the name only exists where its participants have market power. If a group of small rivals agrees on price-fixing, they would be harming their profitability. They are in fact foregoing profits to bigger rivals who sell alcohol or who compete on price. If price-fixing in this particular context is not an obvious profit-maximising strategy, the explanation for the arrangement must be a different one (in this context, it would be part of a certification mechanism).
I do not think I am saying anything new. We have acquired sufficient experience to know that not every price-fixing agreement between competitors is restrictive by object. Just take a look at the examples cited by the Commission in its recent Guidance on the question. And I guess the lesson is clear. The nature of the agreement and the economic and legal context of which it is part must always be considered. There are no shortcuts and certainly no exceptions to the principle!
Chillin'Competition 
I like the examples and the justifications given, but I’d like to add a real life example, which has triggered quite different approaches by NCAs.
In several countries national beer producers have agreed (or attempted to agree) to refrain from producing high alcohol content beers as a measure to lower overall alcohol consumption and to improve public health. As far as I’m concerned, this has not been consireded a violation of competition rules in some Nordic countries. In Estonia, the NCA condsidered the even though this agreeement restricts competition, it is exempted under individual exemption criteria. At the same time, in Lithuania the NCA considered that such an agreement restricts competition and cannot be exempted under individual exemption criteria, and furtermore, fined the undertakings concerned.
Hence, even though the legal and economic context is pretty much the same and there should be uniform application of competition law accross EU, quite opostite interpretaions of the same measure can be possible. In this sense, I really feel for enterprenuers having to align their conduct to competition rules. Even if there are no exceptions possible to the principle that “the nature of the agreement and the economic and legal context of which it is part must always be considered”, there is still little clarity as to what is allowed and what not.
Katri Paas-Mohando
26 January 2016 at 12:59 pm
It’s an interesting issue. I think you are right that the majority of such schemes would not violate the competition rules, but it’s not entirely clear on what basis they would be saved (no restriction at all? a Wouters-type justification? 101/3?). It could, of course, happen that the companies are using the cover of the “ethical groceries” label to put in place a real cartel.
By the way, the EU sponsors a database of voluntary commitments that is full of schemes similar to your examples (http://ec.europa.eu/health/nutrition_physical_activity/platform/platform_db_en.htm). Also, the Horizontal Cooperation Guidelines give a hypothetical of a “government encouraged” standard for recommended maximum fat levels in processed food, saying that it is unlikely to restrict competition because the parties “will be able to continue to compete with regard to a number of other characteristics of the products” (para. 330).
I published a paper on this a while back in case you are interested (http://ebooks.cambridge.org/chapter.jsf?bid=CBO9781107478114&cid=CBO9781107478114A022).
Mislav Mataija
26 January 2016 at 3:13 pm
Not every NCA is automatically convinced that these type of agreements would fall short of the application of article 101 (1) TFEU. In the Netherlands, supermarkets, poultry farmers, and broiler meat processors tried to make arrangements that would improve the welfare of chickens. However, the Autoriteit Consument en Markt found these arrangements to be restrictive of the Dutch Competition Act, because it involved removing chicken meat from the shelves and was therefore viewed as illegal output limitation that was disproportionate in relation to the desired objective.
https://www.acm.nl/en/publications/publication/13761/Industry-wide-arrangements-for-the-so-called-Chicken-of-Tomorrow-restrict-competition/
Bram Nijhof
26 January 2016 at 4:20 pm
Indeed, proportionality is the key concept to address this issue. Taking the Pablo’s example of small grocery stores setting a minimum price for the sale of alcohol, I think such a measure would be found disproportionate in a view to achieve the intended objective: a measure requiring not to sell alcohol in certain hours of the day or not selling super alcoholics would seem a less restrictive alternative to me. Further, being price-fixing a “per object” restriction (which is the same of a hardcore restriction) it would not be exempted under the de-minimis rule and therefore the fact the the companies involved are small grocery stores with no market peer should not be decisive to whether the practice infringes Article 101.
Enzo Marasà
26 January 2016 at 5:36 pm
Thanks to all for the comments!
The examples you have given are very useful to illustrate the key ideas I intended to convey with the post.
The first key idea: it all depends on the context and the nature of the agreement. This is what the Court has consistently held.
I was not making a general statement about agreements of this kind. I was discussing three carefully chosen examples in which the firms lack market power and the scheme is a voluntary one: there is no obligation to join the scheme.
Depending on the nature of the agreement and its context, the conclusion may be very different. I have to thank Katri and Bram for the great examples they have shared.
About Bram’s (there are some details that readers can check): I fully agree with the decision of the Dutch Competition Authority.
Unlike my examples, the Dutch decision concerns an industry-wide arrangement: market power is therefore present. More importantly, there is an element of compulsion. The part of the arrangement that is expressly criticised is the obligation to remove regular chicken from the supermarket shelves.
How are the cases different, then? In the examples I gave above, the participants are willing to forego profits to convey a particular brand image. They know they will lose money to rivals by not selling foie gras, or by serving only healthy meals, as opposed to burgers, fries and steaks. But they assume the consequences of their choice.
In the Dutch case, instead, the farmers devised a mechanism to ensure that they would not lose profits. They agree to enhance animal welfare standards, but only because they are sure that they will not lose sales and profits to regular chicken. They want, in other words, the cake and eat it. A clear restriction of competition.
The Chairman of the Dutch authority raises the fundamental question: is choice enhanced or restricted by the certification mechanism? In the Dutch case, the agreement restricts choice (the agreement would lead to the removal of regular chicken). In my examples, these mechanisms increase choice: a group of restaurants would become more visible and give choice to consumers who want ‘heart healthy’ meals, a group of stores give consumers the guarantee that certain ethical standards are respected. But consumers would still be able to choose from a wealth of alternative restaurants and supermarket chains.
The second key idea: ‘price-fixing’ is an empty label in itself. It does not mean anything as such. In this sense, Enzo’s analysis is incorrect. It appears to contradict the case law. It is always necessary to consider the nature of the agreement and the economic and legal context of which it is part. Experience shows that a price-fixing agreement between competitors is not always by object. The most obvious examples are probably Gottrup-Klim and Tournier (I am sure Mislav would also add Cipolla to the mix!).
It is also useful to think of the various examples provided by the Commission in the Guidance on by object restrictions. It is clear from the document that price-fixing in the context of a joint venture is not a by object infringement of Article 101(1) TFEU. If the examples I gave in my post are not joint ventures, they come very close.
Pablo Ibanez Colomo
26 January 2016 at 10:45 pm
Thanks Pablo for your useful and stimulating comments – I always learn a lot from your and Alfonso’s blog. I appreciate, and agree with, your argument that, theoretically speaking, price-fixing cannot be automatically regarded as a restriction by object since the Court’s case-law requires to carefully consider the economic, factual and legal context as well as the nature of the agreement in the specific case at hand as a pre-condition to get to such a conclusion (and this principle also applies to price-fixing).
However, you will appreciate that the Court’s case-law (including the judgements mentioned in your comments, which in fact do not specifically refer to price-fixing agreements) does not contradict the statement that voluntary (i.e. non State-mandated) price-fixing has – in practice – consistently be deemed a restriction by object under Article 101. Rather, the Court and the Commission’s guidelines have affirmed that, in principle, even price-fixing could fall out of the scope of Article 101 where it is strictly indispensable to attain a legitimate objective that could not otherwise be attained as effectively through a less restrictive measure/agreement (like in certain joint ventures). Yet, I do not think your example fits to such a case (I do not see how an “ethical” price-fixing agreement between grocery stores can be construed as a joint venture, but maybe you have better ideas in mind).
According to this reasoning, I would insist that “proportionality”, rather than market power, is the key concept to assess compatibility of the price-fixing agreement referred to in your example with Article 101 (rectius: with the applicable national mirror provision): because price-fixing is not (in my view – but that can be disputed) the least restrictive means to achieve the goal of discouraging the consummation of alcohol (e.g. not selling alcohol in certain hours of the day or not selling super-alcholics may preserve a greater degree of competition between the grocery stores compared to setting a price floor; and it seems even more effective towards financially-capable drinkers), than such an agreement is disproportionate and would therefore fall within the scope of Article 101 (rectius: of the applicable national mirror provision) as a “per object” restriction.
This being said, our positions may riconciliate in the event that the price-fixing agreement were concluded between distant grocery stores, i.e. where they were so far from one another that could not even be regarded as competitors (since they would belong to unrelated geographical markets). I agree that in such a case the agreement would fall outside Article 101 (or the national mirror provision) because it clearly could not have as object or effect the distortion or restriction of competition. On the contrary, if the grocery stores were close enough, the principle of effective and consistent application of competition law would probably require the application of the “per object” notion to the price-fixing agreement at hand, otherwise (as suggested by Mislav) it would be too easy to use the cover of the “ethical groceries” label to put in place a real cartel.This does not mean that such an agreement would automatically be deemed prohibited and void. It only means that the grocery stores will bear the burden to prove that the conditions for an individual exemption under Article 101(3) (rectius: the applicable national mirror provision) are satisfied – which indeed is unlikely where hardcore retractions are at stake
Needless to say, this is just sparring through academic hypothesis, as we competition-geeks like to do (sometimes I have the impression that competition law experts would be able to state everything and its contrary without contradicting themselves, though I should not say that really). Please let me know your thoughts.
My very best,
Enzo
Enzo Marasà
27 January 2016 at 4:41 pm
Thanks so much for sharing your thoughts, Enzo! This is all really interesting!
Gottrup-Klim and Tournier are indeed about price-fixing between competitors. Yet, the Court held that they were not as such restrictive of competition under Article 101(1) TFEU,
Gottrup-Klim was about a joint purchasing co-operative between small firms (in that sense, similar to my example). When firms buy in common, they fix jointly the purchasing price. This is recognised by the Commission in the Guidance on by object restrictions. The Commission clarifies that price-fixing in this context is not restrictive by object (one of the examples it gives).
Tournier was about the licensing practices of copyright collecting societies. Under an arrangement of this type, authors fix in common the price and the conditions under which their work is made available to customers. This is the example I use with my students to show that not every price-fixing agreement between competitors is restrictive by object. In BMI v CBS, the US Supreme Court acknowledged that these arrangements amount to price-fixing, but have absolutely nothing to do with price-fixing in the context of a cartel.
About the rest of your points: you appear to suggest that an agreement not to sell alcohol between, say, 8pm and 6am is less restrictive than an agreement to sell alcohol around the clock but at a minimum price. This is not a legal question, but, as I mention in my post, I really struggle to see how an agreement not to sell a product at all is less restrictive than an agreement to sell it at a minimum price.
On distant grocery stores: when the participating grocery stores lack market power, it does not matter whether they are close or distant. In my example, consumers who want alcohol would be able to buy it from large supermarket chains, night shops and specialised stores. The independent stores of my example would only represent a very small fraction of the market. Does it really matter, then, that they are distant or close? When consumers have a wealth of alternative outlets, does it really matter that a very small fraction of them agrees on price? In the absence of market power, these independent stores would only be (knowingly) harming their sales and profits. This practice is not a cartel and cannot be treated like a cartel. Even from an enforcement perspective, an authority should not be devoting its resources to investigating it.
Thanks again!
Pablo Ibanez Colomo
28 January 2016 at 9:18 am
On one thing I totally agree with you: a competition authority should not be devoting its resourcing investigating it! 🙂
Thanks much again for your thoughts!
Enzo Marasà
28 January 2016 at 12:37 pm
Not to get mixed up in a conversation that already covers most of the relevant ground, but I would like to point out that joint purchasing is something completely different from price fixing between sellers, and has never been held to be anticompetitive by object. So I’m not sure what the relevance of Gottrup-Klim is here.
Apart from that, isn’t this conversation “simply” about the law on ancillary restraints? My copy of Whish cites all sorts of caselaw about what he calls commercial ancillarity, where “the Court shave held that a particular agreement was necessary to enable the parties to achieve a legitimate purpose.” (p. 136)
Martin Holterman
28 January 2016 at 12:46 pm
I second your thoughts, Martin. Indeed, for a restrictive provision to be deemed ancillary to an agreement a proportionality/indispensability test is required.
Enzo Marasà
28 January 2016 at 2:17 pm
I am really glad that this discussion has sparked so much interest
Very briefly, Martin: I am sure you agree with me that a buyers’ cartel is as bad and as harmful as a sellers’ cartel. And the two are restrictive of competition by object without any doubt. The former is an agreement to set the purchasing price; the latter is an agreement to set the selling price.
The bottom line, again, is that the labels you use (‘price fixing between sellers’; ‘joint purchasing’) are meaningless and misleading in the absence of an analysis of the economic and legal context. An agreement between purchasers may be pro-competitive, as in Gottrup-Klim, or may be a hardcore cartel, as explained by the Commission in the Guidelines on horizontal co-operation agreements. Same with ‘price fixing between sellers’: it can be a plain vanilla cartel or one of the pro-competitive arrangements described by the Commission in its Guidance on by object restrictions.
Ancillary restraints: the doctrine tells you that restraints that are ancillary to a pro-competitive transaction are not restrictive of competition at all. They fall outside the scope of Article 101(1) TFEU altogether.
This is not really the discussion we are having here. The question is whether an agreement can be said to restrict of competition by object simply because it has an element of price-fixing.
Pablo Ibanez Colomo
28 January 2016 at 4:16 pm
A buyers’ cartel is much, much less bad than a sellers’ cartel. Buyers’ cartels reduce prices, meaning that they are presumptively pro-competitive unless some anti-competitive effect can be shown. Cf. chapter 5 of the Commission guidelines on Horizontal Cooperation Agreements.
As for your final par., I suppose that’s the simple way to summarise the law: Price fixing agreements are anti-competitive agreements by object unless they are ancillary restraints. (Though they may still be justified under art.101(3), of course.)
Martin Holterman
29 January 2016 at 1:28 pm
Hi Martin,
You have to be much more precise in the use of vocabulary. A cartel is not just any agreement between competitors. In EU competition law a cartel is understood to be an agreement between competitors with market power that lacks any redeeming virtues. The sole purpose of a cartel is to restrict competition.
Cartel conduct may take many forms. One thing is clear, however. Cartel conduct is restrictive of competition by object, typically subject to very heavy fines and very unlikely to fulfill the conditions of Article 101(3) TFEU.
It does not matter whether the cartel is concluded between buyers or sellers, or between firms taking part in a tender. The legal status of the practice would be the same. For an example of a cartel between purchasers take a look at Spanish Raw Tobacco, which Alfonso discussed when he came to LSE.
The Commission does not say in the Guidelines on horizontal co-operation agreements that buyers’ cartels are presumptively pro-competitive unless some anti-competitive effect can be shown. That would be an abomination. What is more, the Commission even explicitly clarifies that a joint purchasing agreement that is a cartel in disguise is restrictive of competition by object.
Your second paragraph contains a very categorical statement, which my blog post seeks to dispute 😉
Have a great weekend!
Pablo Ibanez Colomo
29 January 2016 at 6:19 pm
Pablo, I suppose we’ll have to agree to disagree about many of these things. Let me just ask one final question: When has it ever been held that cartel members, individually or collectively, have to have market power in order to fall foul of art. 101?
The whole point of an object cartel is that it doesn’t actually need to impact competitive conditions in order to be illegal. It is not necessary to show an effect on competition. (See, most recently, Maxima Latvija.) If just 3 out of 100 competitors form a cartel, it could still be illegal, if the conduct in question is anticompetitive by object.
Martin Holterman
30 January 2016 at 8:46 pm
Hi Martin!
I did not suggest, let alone write, that ‘cartel members, individually or collectively, have to have market power in order to fall foul of art. 101’
I am afraid you are mixing up two issues, one economic (when do cartels form?) and one legal (when are cartels prohibited?).
A cartel is only a profitable strategy, and therefore a rational one for firms to adopt, when its members are able to raise (decrease) the market price. A group of retailers with a joint market share of 3% would not only not be able to raise prices if they form a cartel. They would be harming themselves, not society. You may want to take a look at Hovenkamp’s Antitrust Enterprise, where this point is very clearly explained. Luc Peeperkorn has recently published a great piece on restrictions by object where he emphasises this issue, often forgotten in EU competition law.
The fact that a cartel is only a profitable strategy when firms are able to raise the market price does not mean that it is necessary to show market power to establish an infringement. Cartel conduct, when detected, will be prohibited by object.
But please note: it is not because a cartel is prohibited by object under Article 101 TFEU that cartel conduct is a profitable strategy when firms lack market power. The law does not determine the conditions under which a cartel is a profitable strategy. That is an economic matter.
The bottom line, and the key message I intended to convey with my post: if a group of retailers with a joint market share of 3% conclude an agreement that has a price-fixing element, this agreement is most probably not a cartel. When firms lack market power and agree on prices, the arrangement is, in all likelihood, not a cartel, but an agreement that promotes competition.
And do not forget how lucid and carefully drafted Cartes Bancaires (and, most recently, Maxima Latvija) are. The Court does not refer to price-fixing, but to ‘horizontal price-fixing by cartels’. As my post emphasised, the nuance is of major importance.
Thanks so much for your active involvement in the blog!
Pablo Ibanez Colomo
1 February 2016 at 10:34 am