Under EU competition law, collective boycotts can be most readily assessed under the framework of anti-competitive agreements pursuant to Article 101 (1) of the Treaty on the Functioning of the European Union (hereinafter referred to as “TFEU”). Such agreements can take many forms: If the collective advertisement boycott was organized e.g. by a trade association of advertisers via their decision-making body or between individual advertisers via multi-party agreements, meetings, e-mails etc., it could qualify as such an agreement.

The question may arise, however, whether a horizontal agreement between advertisers concerning their advertisement spending is even capable of having a restrictive effect on competition and thus violating Article 101 of the TFEU. We are used to think of cartels as being organized on the supply side (by the sellers) and taking the forms of fixing prices, restricting output or market sharing. In contrast, an advertisement boycott is organized on the demand side, by buyers of advertising spaces through the withholding of their spending from a targeted platform.

While agreements between buyers appear less frequently in case law and may be treated more leniently by authorities, it is important to note that EU competition law aims to prohibit anticompetitive behaviour in all its forms. If buyers on a certain market engage in a coordinated conduct that has the object or effect of restricting competition, the conduct may be regarded as an antitrust violation. Similarly, it is of no import whether the anti-competitive effects of an agreement arise on the market where the participants are active or on an adjacent market. A boycott agreed on by a group of advertisers could be found anti-competitive even if it did not restrict competition between the advertisers themselves but had restrictive effects elsewhere, e.g. on the market where their vendors operate – in our case, on the market of advertising spaces on online platforms.

Finally, the goal to be achieved by the boycott is not a decisive factor in assessing its legality under competition law. While authorities may take the participants’ stated goals into account, good intentions do not automatically relieve them of responsibility. In previous cases, it has been held that agreements between competitors which, by their express terms, merely aimed at the prevention of the sale of faulty, used or dumping products could still be anti-competitive in view of their effects. Accordingly, it cannot be excluded that an advertising boycott that is aimed at enforcing certain brand safety standards could unlawfully restrict competition.

Having established that our hypothetical advertiser boycott could qualify as a horizontal agreement, the next question is whether it is capable of preventing, restricting, or distorting competition in the EU.

The assessment of whether a given agreement violates the prohibition of Article 101 of the TFEU consists of three main steps. The first is to determine whether the agreement between undertakings has an anti-competitive object. If not, the question becomes whether it has nonetheless actual or potential restrictive effects on competition. Finally, if the agreement is found to restrict competition either by its object or by its effect, it must be determined whether there are any economic advantages produced by the agreement that would offset the harm suffered by competition.

A. Is there a by-object restriction?

Article 101 of the TFEU lists certain types of agreements that have an inherently anti-competitive objective (fixing prices or any other trading condition, limiting production or investment, dividing markets or sources of supply, applying dissimilar conditions to equivalent transactions, etc.). Theoretically, a collective advertising boycott to enforce certain brand safety standards may be regarded as falling under these categories. It could be viewed, for example, as:

• an agreement among advertisers to fix certain conditions (namely on brand safety) under which advertisers will buy advertising spaces from certain online platforms,
• an agreement to limit the production of advertisements or the investment into the purchase of advertising spaces, particularly if the exclusion of the targeted platform led to an overall decrease in social media advertising, 
• an agreement to apply dissimilar conditions to equivalent transactions, particularly if the organizers of the boycott discriminated against the targeted platform by not enforcing the same brand safety standards against other similar sites.

Indeed, case law specifically recognizes buyer cartels as agreements that are anti-competitive by object. These are defined as agreements between two or more purchasers which, without engaging in joint negotiations vis-à-vis the supplier, align the purchasers’ individual behaviour on the purchasing market, for example through coordination of the purchasers’ negotiation strategies. Buyer cartels unlawfully restrict competition as purchasers are not supposed to first fix the conditions of purchase between themselves before each purchaser individually negotiates its purchases from suppliers.

One could argue that an advertising boycott whereby each participating advertiser demands compliance with a brand safety standard that they previously agreed upon among themselves as a condition for buying advertising space from an online platform is just this kind of buyer cartel and therefore qualifies as a restriction of competition by object.

This is the point, however, where the intentions of the participants may play a role. According to case law, although an agreement between purchasers not to buy from particular suppliers may amount to a restriction of competition by object, this is not generally the case. If the collective boycott of a supplier is based on certain characteristics of its product or its manufacturing processes (e.g. unsustainable technology, slave labour), such an agreement – called a vertical buyer boycott – does not in itself have the object of restricting competition. Accordingly, it does not qualify as a buyer cartel.

If the advertisers withheld advertisement spending from an online platform by reference to brand safety standards according to a behind-the-scenes agreement among themselves, the issue would thus boil down to whether the invoked brand safety standards were to be regarded as a negotiable commercial term of the advertising contracts or rather a characteristic of the platform where the advertisers were considering buying advertisement spaces.

If the brand safety standards were regarded as commercial terms that an advertiser and a platform may want to agree upon when concluding and advertisement deal, the advertisers’ behind-the-scenes agreement to fix these terms for all future purchases would likely qualify as a buyer cartel and as such a by-object restriction of competition. If, by contrast, the level of brand safety were to be regarded as an inherent characteristic of the platform offering advertisement spaces, it would be more likely that a behind-the-scenes agreement of the advertisers not to place advertisements on the platform in question until the issue was fixed would qualify as a vertical buyer boycott whose object in itself was not restrictive.

Since brand safety standards only make sense in the context of advertising and are in fact demanded by advertisers when buying advertising spaces, it is more likely that they should be regarded as commercial terms of advertising contracts rather than an inherent characteristic of the platform where the advertisement is placed. Consequently, a behind-the-scenes agreement of advertisers not to place advertisements on an online site until a set of brand safety standards is fulfilled by the platform will likely qualify as a restriction of competition by object.

Agreements between buyers concerning the terms of their purchases may, however, be exempt from the strict treatment of buyer cartels if they qualify as joint purchasing agreements. In this case, negotiations are conducted collectively on behalf of all the buyers involved, who become bound by the agreed terms and conditions for all their purchases. In contrast to behind-the-scenes buyer cartels, joint purchasing agreements are not seen as anti-competitive by object as they aim to create a degree of buying power vis-à-vis dominant suppliers which individual members of the joint purchasing arrangement might not attain if they acted independently, thereby enhancing competition.

As the negotiations with a supplier on behalf of a joint purchasing agreement might involve suspending purchases as a negotiation tactic to achieve more favourable terms, temporary buyer boycotts implemented in the context of a joint purchasing agreement are not considered unlawful restrictions of the competition by object. If the hypothetical advertisers engaged in joint negotiations with the online platform with the aim of making sure that their brand safety standards were complied with, their conduct would not be considered anti-competitive by object even if it involved withholding advertisement from the site until these requirements were met.

The question whether the hypothetical advertiser boycott qualifies as a buyer cartel, a vertical buyer boycott or a joint purchasing agreement is of paramount importance for the outcome of the analysis: while buyer cartels qualify as restrictions by object, vertical buyer boycotts and joint purchasing agreements do not – these can at most prove anti-competitive in their effects.

B. Is there a by-effect restriction?

If the advertisers’ boycott was implemented in the context of a joint purchasing agreement (or at least not as a buyer cartel), the next step would be to assess whether the agreement had any negative effect on competition. The assessment would need to cover the possible restrictive effects both upstream (in our case on the market of social media advertising), and downstream (on the market of the products advertised) and look at whether the coordinated conduct of the advertisers led to distortions on the market (increased prices or the reduction of output, quality or innovation) that would not have occurred without such coordination.

One factor would be the market power of the advertisers participating in the boycott. If they represented a significant share of overall advertising spending on social media, there would be a greater risk that the arrangement may lead to market distortions, which may ultimately also cause harm to consumers. First of all, there would be the risk of reduced competition among the advertisers themselves due to the information exchange between participants of the joint purchasing agreement. Second, case law recognizes that the exercise of joint buying power by purchasers may harm suppliers’ investment incentives and force them to reduce the range or quality of services offered, ultimately leading to sub-optimal supply.

In our case, such market distortions could manifest in the decrease of advertising offered by the boycotted site as well as a slowdown of the roll-out of more innovative ways of placing advertisements on the platform. This would reduce consumer reach by advertisers and would make it more difficult for consumers to identify available products or services. The boycott could also artificially shift advertising spending to other platforms and other media, leading to price increase and loss of efficiency. Overall, if the market share of the participating advertisers were significant, it would seem likely that the boycott would have at least some negative effects on competition.

EU case law unfortunately provides little guidance to advertisers on the boundaries of collective boycotts. Some decisions on the national level, however, are worthy of attention as they deal with buyer boycotts targeted at suppliers.

The Irish Competition Authority, for example, initiated competition proceedings against the Irish Dental Association following allegations of a collective boycott against a private dental insurance company. The boycott was orchestrated through informal and formal communication with members of the trade association advising them not to complete a certain claim form provided by the insurance company unless it agreed to change the form. The case resulted in a settlement between the competition authority and the trade association. In another matter, the Italian Competition Authority found that the collective boycott organized by the Italian National Association of Labor Consultants against a software supplier, in which the Association urged its members not to do business with the supplier, constituted a violation of the Italian Competition Act.

In a case involving advertising boycott, the Danish Competition Authority found that the Danish Association of Peugeot Dealers engaged in an unlawful boycott against bilbasen.dk, a Danish platform for advertisement and sale of cars. According to the decision, members of the association were instructed not to advertise on bilbasen.dk in order to promote its competitor. The Authority found that such a conduct was to be regarded as a by-object restriction of competition.

In case the advertising boycott qualified as a restriction by object or by effect, the final step would be to examine whether the agreement had an economic benefit (e.g. improving the production or distribution of goods or promoting technical or economic progress) that consumers can profit from and that outweighs the restrictive effects of the agreement.

In our hypothetical scenario, it is doubtful whether such efficiency gains would prevail. While brand safety standards may be important for advertisers, they do not seem to contribute to improved economic production or technical innovation unless it could be shown that they enhance consumer reach on online platforms. Short of such proof, it is unlikely that any alleged positive effect of brand safety standards would outweigh the negative effects that an advertising boycott enforcing such standards can have on competition.

Here again, the motives of the advertisers may prove relevant. Even if brand safety standards were to be regarded as being able to contribute to the improved production or distribution of goods, any concerted action undertaken for the enforcement of such standards would need to be proportionate to the purported objective. If it turned out, for example, that the boycotted online platform had already accepted reasonable brand safety standards but the advertisers still continued the boycott citing further demands, authorities could conclude that the boycott was not indispensable to attain the justified objective and therefore did not warrant an exemption.