EU antitrust in support of the Green Deal. Why better is not good enough

Abstract

The European Union (EU) Commission proposes to ‘green up’ its enforcement of Article 101(3) TFEU to allow producers to collectively overcome so-called first mover disadvantages that would result from inefficient market regulation. The Commission's reboot focuses on the last three exemption conditions. First, the consumer benefit condition is customized to use collective consumer benefits to determine whether consumers receive a ‘fair share’ of the benefits established under the efficiency condition. Here, the Commission bypasses the Dutch proposition to also take account of non-consumer benefits when investigating whether consumers are compensated for anticompetitive harm. Second, the indispensability condition is tasked to filter out greenwashing. Third, the residual competition condition is trusted to allow private collective action insofar it does not eliminate competition on price and/or innovation. Discussing both EU and Dutch proposals, this article finds that greening up Article 101(3) brings competition policy outside the limiting principles that define objective and effective competition enforcement in terms of voluntary exchange.

I. INTRODUCTION

Climate change confronts mankind with one of its biggest crises ever. To literally stem the tide, the European Union (EU) has established an ambitious ‘Green Deal’ to fulfil the United Nations (UN) Sustainable Development Goals as well as the long-term temperature goal of the Paris Agreement.¹ Green antitrust proponents claim that antitrust must ‘green up’ because mainstream antitrust would hinder producers to invest in green production.² The European Commission partly agrees according to the consulted revision of the Horizontal Guidelines (EU Guidelines).³ The basis of the Commission’s position is that mainstream antitrust is sufficiently green as it is. For it is in protecting competition, which ‘spurs innovation, increases the quality and choice of products, ensures an efficient allocation of resources, [and] reduces the costs of production’, that competition law enforcement contributes to welfare.⁴ But the Commission does concede that mainstream antitrust does not suffice in situations where individual production and consumption decisions have negative effects on third parties, and public policies and regulation do not address, or do not address fully, these market failures.⁵ In these cases, mainstream antitrust impedes producers to collectively overcome so-called first mover disadvantage, a situation where none of them would initiate more sustainable production because consumers would not be willing to pay the higher prices concomitant with greener production processes.⁶ Green antitrust addresses this problem by facilitating private collective action.⁷ The Commission further concedes that, in situations where consumers are unwilling to pay a higher price for more costly, greener products, individual consumer benefits may not suffice to establish ‘fair share’ under the consumer benefit condition.⁸ To effectively green up antitrust, the Commission therefore commits to take account of collective consumer benefits occurring outside the relevant market provided ‘consumers in the relevant market substantially overlap [...] with the beneficiaries outside that market’, and these benefits ‘are significant enough to compensate consumers in the relevant market for the [anticompetitive] harm suffered’.⁹

The Commission hereby rejects the greener antitrust proposition put forward by the Dutch Authority of Consumers & Markets (ACM; Dutch Guidelines).¹⁰ The Dutch Guidelines are ‘greener’ because ACM suggests replacing the consumer welfare standard with a so-called citizen welfare standard that uses non-consumer benefits to determine ‘fair share’ in the case of sustainability agreements that help society realize a binding international or national environmental standard.¹¹ The use of a citizen welfare standard to establish consumer compensation broadens the scope of green antitrust: because non-consumer benefits are potentially larger than collective consumer benefits, they may outweigh anticompetitive harm more easily.

This article discusses the merits of green and greener antitrust, which I define as competition enforcement that prioritizes the promotion of sustainability over the protection of competition. This definition connects the need for private collective action to the economic concept of negative externalities and mainstream antitrust. Negative externalities occur when the production and consumption of a product have negative effects on non-consumers. Not all negative externalities lead to a market failure however. Simple negative externalities, which concern only a few stakeholders, can be corrected through additional cooperative arrangements that do not eliminate competition. Correction of complex negative externalities, on the other hand, requires collective action due to high transaction costs and extensive free riding. In other words, it is only in case of complex negative externalities that the voluntary exchange mechanism of the market fails. At the same time, antitrust protects against another market failure: market power in terms of coercion.¹² This means that mainstream antitrust suffices to potentially clear agreements that correct simple negative externalities. Not eliminating competition, these agreements do not lead to market power in terms of coercion. But mainstream antitrust is not able to clear agreements that correct complex negative externalities. Protecting against market power in terms of coercion, mainstream antitrust stands in the way of private collective action. Green antitrust aims to fill the gap between complex negative externalities and mainstream antitrust by allowing private collective action in those situations where the legislative government has failed to correct, or to correct fully, an externality. References to negative externalities in this article therefore pertain to complex negative externalities. References to market power pertain to market power in terms of coercion.

The central question is thus whether antitrust can be used to correct one market failure, a negative externality, by accommodating the very market failure that competition law is tasked to protect against—market power.

To answer this question, I study private collective action aiming to promote green as a multipurpose cartel under Article 101(3). On the one hand, private collective action aiming to promote green qualifies as a ‘restriction by object’ under Article 101(1) TFEU since it has ‘the object of controlling a product characteristic on which there is competition in the market’.¹³ On the other hand, private collective action improves economic efficiency under the first condition of Article 101(3) when it leads to ‘cleaner production or distribution technologies, less pollution, […] better quality products, etc.’¹⁴ This approach differs from the Commission’s, which limits object restrictions to sustainability agreements that ‘do not genuinely pursue a sustainability objective but cover up price fixing, market or customer allocation, limitations of output or limitations of quality or innovation (italics added)’,¹⁵ and allows common minimum standards to profit from a ‘soft safe harbour’.¹⁶ Instead, I rely on case law that qualifies ‘eliminations of competition that change the structure of the market’ as object restrictions ‘even without limiting output or innovation’.¹⁷ This implies that the category of object restrictions also includes coordinated foreclosure of products that are marketed based on alternative standards like the existing regulatory standard. The result hereof is that it cannot be determined upfront whether private collective action genuinely promotes green, whilst common minimum standards cannot profit from a ‘soft safe harbour’ under Article 101(1). In lieu, the legal framework of Article 101(3) serves to determine whether private collective action should be cleared because it pursues the promotion of green rather than cartel overcharges.

The central question thus becomes whether the legal framework of Article 101(3) can be customized as claimed. The first condition states that anticompetitive agreements must improve productive efficiency and/or promote economic progress (efficiency condition). The second condition requires that consumers must receive a ‘fair share’ of said benefits (consumer benefit condition). The third condition specifies that the anticompetitive agreement at issue must be necessary to attain said benefits (indispensability condition). The fourth condition demands that anticompetitive agreements do not eliminate competition in respect of a substantial part of the products in question (residual competition condition). The interpretation of the efficiency condition is not up for debate since a correction of external costs improves economic efficiency. Instead, the focus is on the interpretation of the other three exemption conditions and how they interact with the first.

Based on the above, the article is structured as follows. Section II investigates whether the Commission rightfully rejects the Dutch proposition to use a citizen welfare standard to determine ‘fair share’ under the consumer benefit condition. To do so, this section examines the main Dutch arguments. First, that the presence of binding international or national environmental public objectives justifies said use of a citizen welfare standard in the case of so-called environmental damage correction agreements. Second, that case law, in particular MasterCard, allows competition agencies to take account of non-consumer benefits when determining ‘fair share’.¹⁸ Third, that the ‘polluter pays’ principle laid down in Article 191(2) TFEU mandates competition agencies to take account of non-consumer benefits when determining ‘fair share’. Section III analyses whether the EU Guidelines fare better than the Dutch Guidelines in complementing the Green Deal. This section investigates four main assertions. First, that the consumer benefit condition allows ‘fair share’ to be based on collective consumer benefits on the condition that these benefits objectively accrue to consumers in the relevant market. Second, that the indispensability condition can legitimize private collective action by way of filtering out ‘greenwashing' (where promoting green covers for pursuing cartel overcharges). Third, that the residual competition condition allows private collective action provided that the agreement at issue does not eliminate competition on price and/or innovation. Fourth, that CECED constitutes the prototype of a private collective action that genuinely promotes sustainability.¹⁹ Section IV presents an overall conclusion.

II. GREENER ANTITRUST ACCORDING TO ACM

The Dutch Guidelines clarify how ACM will enforce the Dutch counterpart of Article 101 TFEU for agreements that aim to improve sustainability. These guidelines contain two parts that both define ‘consumer’ in the mainstream way: they who buy or use a product or service.²⁰ A general part that identifies sustainability agreements in broad terms, thus covering all agreements ‘aimed at the [prevention or mitigation] of the negative impact of economic activities on people (including their working conditions), animals, the environment, or nature’.²¹ In this part, the consumer welfare standard will govern the enforcement of the consumer benefit condition. Here, ‘fair share’ requires that potential sustainability benefits must compensate consumers fully for the anticompetitive harm they suffer from the agreement.²² The other part narrows sustainability agreements down to agreements that reduce environmental damage resulting, among other things, ‘from the emission of harmful air pollutants and greenhouse gases, and from the waste of raw materials’.²³ Reducing negative externalities in terms of environmental costs, these ‘environmental damage correction agreements’ fulfil the efficiency condition because they lead to a more efficient use of natural resources.²⁴ This time, ACM employs a citizen welfare standard under the consumer benefit condition that allows it to base ‘fair share’ on non-consumer benefits when consumer benefits do not fully compensate anticompetitive harm.

The use of a citizen welfare standard is justified in two separate tracks. The Dutch Guidelines purport that in case an agreement ‘helps, in an efficient manner, to comply with an international or national standard targeting the reduction of environmental costs’,²⁵ particularly by reducing the risks of global warming,²⁶ ‘it is justified to take the sustainability gain for society as a whole into account’ when determining ‘fair share’ under the consumer benefit condition.²⁷ To make this work, ACM introduces a ‘public objective’ defence under Article 101(3) for clearing anticompetitive sustainability agreements that efficiently correct environmental costs.²⁸ The landmark Urgenda case,²⁹ in which the Supreme Court of the Netherlands (SCOTN) confirmed that the international goal to reduce greenhouse gas emissions by 25% of 1990 levels in 2020 binds the Dutch State, would corroborate the introduction of said defence.³⁰ In a separate legal memo, ACM submits two additional arguments to motivate the use of a citizen welfare standard.³¹ Case law, in particular MasterCard,³² would allow competition agencies to determine ‘fair share’ based on non-consumer benefits.³³ The ‘polluter pays’ principle laid down in Article 191(2) TFEU would mandate competition agencies to take account of non-consumer benefits.³⁴ But are these claims correct?

The citizen welfare standard and the public objective defence

The risk of using a citizen welfare standard under the consumer benefit condition is that it politicizes competition law. ACM acknowledges this point: ‘Normally [we are] very reluctant to allow [anticompetitive] agreements […] that lead to redistribution between users and non-users’ because ‘it is up to the democratically elected legislature to determine who contributes to what extent to the achievement of public interest goals’.³⁵ The Dutch agency therefore limits the use of said standard to environmental damage correction agreements, in which case two additional safeguards would depoliticize its use. One safeguard restricts the application of the citizen welfare standard to agreements that help realize a public sustainability objective.³⁶ The second safeguard provides that the agreements at issue must be as efficient as a potential government measure ensuring the same objective.³⁷

The public objective defence resembles the legitimate objective defence introduced in Wouters.³⁸ While legitimate objective defences motivate the clearance of potentially anticompetitive agreements under Article 101(1), the public objective defence purports to motivate the clearance of anticompetitive agreements under Article 101(3). Slightly confusing is that ACM considers the legitimate objective defence ‘insufficiently clear’ to base green antitrust on,³⁹ but feels nonetheless confident enough to export it to Article 101(3). This poses the following questions. What does the European Court of Justice (ECJ) exactly do in Wouters? How does Wouters relate to state action doctrine? Can the legitimate objective defence be exported to Article 101(3)? Does Urgenda make the answers to the previous questions any different?

Wouters concerns private sector regulation based on which the Dutch Bar prohibited its members to partner up with accountants (Regulation).⁴⁰ Even though the Regulation was held to restrict competition,⁴¹ it was cleared under Article 101(1) because the pertaining anticompetitive effects could be considered inherent in the pursuit of a legitimate objective—ensuring the proper practice of the legal profession as organized in the Netherlands.⁴² To arrive at that conclusion, the ECJ had investigated whether the Regulation was ancillary to the deontological principles that govern the proper practice of the legal profession.⁴³ If so, the pertaining anticompetitive effects were caused by these principles rather than the Regulation. The ancillarity investigation was conducted in two steps. A preliminary investigation to identify the deontological principles that govern the proper practice of the legal profession and thus constitute the reference point for determining potential ancillarity.⁴⁴ And a necessity investigation to determine whether the Regulation was ancillary to the deontological principles identified. The preliminary investigation established that, according to the Dutch Act on Advocates, lawyers must guarantee ‘that all steps taken in a case are taken in the sole interest of their client’.⁴⁵ Accountants, on the other hand, must undertake an objective examination and audit their clients’ accounts to inform other parties on the reliability of those accounts.⁴⁶ The necessity investigation established that the anticompetitive effects were caused by the deontological principles rather than the Regulation. Because the principles that governed the legal and accountancy profession were not compatible and to some extent even incompatible,⁴⁷ members of the Bar might no longer be in a position to live up to their own deontological duties when partnering up with accountants.⁴⁸ Also, the Regulation restricted competition no more than necessary to ensure that lawyers lived up to their deontological duties.⁴⁹

Based thereupon, Wouters effectively adds another branch to EU state action doctrine, which makes sure that antitrust does not extend to anticompetitive effects that are caused by the State. At its core, state action doctrine clears anticompetitive effects that are required by national legislation.⁵⁰ This ‘public norm’ requirement ensures that the State has actually been responsible for the political decision of displacing competition with regulation.⁵¹ The State concerns the legislature because it is this particular institution that represents the citizen, who is the ultimate decision-maker in the political domain.⁵² State action defences work as a double-edged sword: antitrust applies once the public norm does not compel firms to take particular action,⁵³ or leaves room for competitive activity.⁵⁴

The ‘useful effect’ jurisprudence further specifies that for state measures to qualify as public norms they may not enable firms to escape Article 101(1).⁵⁵ Particularly, Member States may not delegate their responsibility for displacing competition with regulation to the private sector. This implies that the State must carry ‘originator responsibility’ (they must specify subject matter, method and extent of regulation as well as the criteria that the private sector must meet) and ‘concluder responsibility’ (they must check whether the private input lives up to these requirements).⁵⁶

Wouters follows a similar approach. Even though the Regulation is not based on a ‘public norm’ that precludes the Bar from using its delegated regulatory power in a manner that restricts competition, ancillarity analysis ascertains that the State is responsible for its anticompetitive effects. The principle that members of the Bar must act in the sole interest of their client originates with the State, while the Regulation restricts competition no more than the original State-defined deontological rule.

Since state action defences focus on who is responsible for restricting competition, they are necessarily limited to Article 101(1). If the State is responsible, the agreement is cleared under Article 101(1), which means that Article 101(3) is no longer relevant. If the parties to the agreement are responsible, the anticompetitive effects are theirs to account for under Article 101(1), but they can also claim the agreements’ potential procompetitive effects under Article 101(3) irrespective of the public status of the objective the agreement helps to promote. This makes a ‘public objective’ defence of no relevance under Article 101(3).

ACM suggests that the Urgenda ruling changes the above because ‘even though ACM is an independent authority […] it is also part of the State’.⁵⁷ In other words, because SCOTN confirmed that courts may order the Dutch State to live up to its international obligations in the context of global warming, it would also instruct ACM to green up antitrust. I do not think that is what Urgenda says. Potentially relevant are SCOTN’s findings regarding the institutional balance between the legislative State and the judiciary. Here, SCOTN first emphasizes that decisions regarding the reduction of greenhouse gasses fall within the political domain of the legislature, and that the role of the judiciary in constitutional democracies is limited to verifying whether the legislature remains within the boundaries of the law when using its discretionary power to weigh relevant political considerations to make value judgements. SCOTN then specifies the extent of said verification duty for situations where the legislature has abused its political power, as was held to be the case in Urgenda.⁵⁸ In so doing, the court held that courts may order the government to live up to its (international) obligations but only insofar this does ‘not amount to an order to take specific legal measures but leaves the State free to choose the measures to be taken in order to achieve a 25% reduction of greenhouse gas emissions by 2020’.⁵⁹ In other words, it is by not instructing the legislature what specific legal measures to take that the lower courts stayed away from political decision-making. So, insofar Urgenda translates to antitrust, it confirms that private collective action can only be cleared if the legislative State is responsible for the pertaining anticompetitive effects.

ACM’s proposal for a regulatory impact assessment under the indispensability condition of Article 101(3) cannot help to correct the democratic deficiency of the public objective defence either. The reason is simple: this condition does not compare anticompetitive agreements with an alternative regulatory measure the State should have taken but with an alternative less restrictive private arrangement.⁶⁰

It follows that the public objective defence does not depoliticize the use of a citizen welfare standard under the consumer benefit condition. Said standard makes antitrust inherently political because it purports to clear anticompetitive effects that the State should have caused but did not cause based on non-consumer benefits that the State should have ensured but did not ensure.

‘Fair share’ according to MasterCard and prior case law

In mainstream antitrust, ‘fair share’ as referred to in the consumer benefit condition means that ‘the pass-on of benefits must at least compensate consumers for any actual or likely negative impact caused to them by the restriction of competition found under Article [101(1)]’.⁶¹ In other words, ‘fair share’ means ‘full consumer compensation’. The focus on consumers and their benefits in the relevant market concurs with the view that consumer welfare is the better standard to enforce competition law because it precludes ‘unfair’ welfare transfers from consumers to producers. ‘Unfair’ refers to those welfare transfers that producers can extract from consumers by way of their market power.⁶² ACM argues that MasterCard and prior case law imply otherwise. Is ACM right?

In MasterCard, the ECJ discussed the matter of compensation in a two-sided market where banks provide payment cards to cardholders (consumers on the issuing side of the market) but also deal with merchants who accept payments with those cards (consumers on the acquiring side of the market). The MasterCard payment system posed a problem under Article 101(1) because the multilateral fallback interchange fees in the payment system (MIF) restricted competition. As for the application of Article 101(3), the ECJ observed that, ‘should the General Court have found that there were appreciable objective advantages flowing from the MIF for merchants, even if those advantages did not in themselves prove sufficient to compensate for the restrictive effects identified [under Article 101(1)]’, the advantages for cardholders ‘could, if necessary, have justified the MIF if, taken together, those advantages were of such character as to compensate for the restrictive effects of those fees’.⁶³

One cannot derive from this consideration that ‘fair share’ can also include benefits that occur outside the relevant market because the ECJ did not discuss the consumer benefit condition. When answering the question as to which markets may be regarded as generating the objective advantages that may be taken into account for the purposes of the analysis under the efficiency condition,⁶⁴ the court stuck to the question asked and discussed only the first exception condition, not the second.⁶⁵ Meanwhile, the General Court (GC) and AG Mengozzi (AG) had addressed the meaning of ‘fair share’.⁶⁶ The GC had emphasized that, while the advantages under the efficiency condition may arise ‘not only for the relevant market but also for every other market on which the agreement […] might have beneficial effects’ as well as ‘for any service the quality or efficiency of which might be improved by the […] agreement’, the ‘very existence of the [consumer benefit] condition necessarily means that the existence of […] advantages attributable to the MIF must also be established in regard to [merchants, i.e. those consumers that suffer anticompetitive harm]’.⁶⁷ The AG further specified that the consumers that suffer anticompetitive harm must be ‘compensated in full’.⁶⁸ This is because taking account of benefits enjoyed by consumers other than those suffering anticompetitive harm from the agreement ‘would amount to allowing [one] category of consumers to be favoured to the detriment of [another]’.⁶⁹ According to the AG, that kind of ‘distributive logic’ has ‘no connection with the practical scope of competition law’.⁷⁰ The Supreme Court of the United Kingdom underwrites this reasoning in Sainsbury’s Supermarkets v Visa/MasterCard, a case that concerned a similar two-sided payment card system as in MasterCard.⁷¹ Stressing that the consumer benefit condition is only useful when it has something to add to the efficiency condition, the court underscores that these conditions are ‘essentially different’ in that the consumer benefit condition adds a different kind of ‘fairness’—one that precludes aggregate efficiency gains across different markets from motivating ‘fair share’.⁷²

In other words, while the ECJ did not address the issue, the GC, AG, and UKSC all agreed that ‘fair share’ really does mean ‘full consumer compensation’.

Prior case law confirms that finding. One strand of cases regards the interpretation of ‘appreciable objective advantages’ under the efficiency condition: Consten and Grundig,⁷³  Landewyck,⁷⁴ and Compagnie Générale Maritime.⁷⁵ Another strand regards the interpretation of ‘fair share’ under the consumer benefit condition: Shaw⁷⁶ and Asnef-Equifax.⁷⁷

In Consten and Grundig, the ECJ noted that ‘an improvement in the production o[r] distribution of goods in question’ must show ‘appreciable objective advantages of such a character as to compensate for the disadvantages which they cause in the field of competition’.⁷⁸ Here the court describes the balancing mechanism that Article 101 entails. While Article 101(1) targets the costs of restrictive agreements in terms of allocative inefficiency, Article 101(3) looks into potential countervailing benefits in terms of productive efficiency and/or economic progress.⁷⁹ In Landewyck, the ECJ further clarified that ‘the requirements for the maintenance of workable competition may be reconciled with the safeguarding of objectives of a different nature and that to do this certain restrictions of competition may be permissible’.⁸⁰ In other words, ‘appreciable economic advantages’ may concern efficiencies at large. This position is confirmed in Compagnie Générale Maritime, where the Court of First Instance (CFI) underlined that ‘appreciable objective advantages’ regard ‘not only the […] relevant market’ but also ‘every other market on which the agreement in question might have beneficial effects’ as well as ‘in a more general sense, for any [product or] service the quality or efficiency of which might be improved by the existence of that agreement’.⁸¹

The cases regarding the interpretation of ‘fair share’ under the consumer benefit condition, Shaw and Asnef-Equifax, are generally known for their finding that consumer compensation can be based on average consumer benefit.⁸² But both judgments also show something else. In Shaw, the CFI determined that ‘it is not material that the benefits produced by the agreements do not entirely compensate the price differential by a particular lessee if the average lessee does enjoy that compensation’.⁸³ In Asnef-Equifax, the ECJ reiterated that ‘the overall benefit of consumers in the relevant market must be favourable’.⁸⁴ In other words, to receive a ‘fair share’, the average consumer must at least be fully compensated.

Prior case law thus agrees with GC, AG, and UKSC that ‘appreciable objective advantages’ and ‘fair share’ are different notions, and that ‘fair share’ is congruent with full consumer compensation. So, what explains the narrowing down to consumers and their benefits in the relevant market? And how does the average consumer fit in?

The explanatory factor is that the institutional position of consumers and non-consumers (all other persons than consumers in the relevant market) inherently differs in antitrust. In market democracies, individuals have dual identities corresponding with their position in, on the one hand, the marketplace and, on the other, the legislative state. The marketplace is a system within which individuals in their capacity as consumers and producers can act freely and make their own choices. The legislative state provides a vehicle for individuals of a polity to determine whether and if so, how one should limit their individual autonomy and contractual freedom to promote the common good. In the commercial arena, consumers, not producers, are the ultimate decision-makers. Adam Smith and Ludwig von Mises, succinctly summarized by Vanberg, have already explained why: ‘because we only produce in order to consume’, it is ‘only as a consumer’ that ‘man […] can command’.⁸⁵ In the political arena, citizens (consumers and non-consumers together) are the ultimate decision-makers since the limitation of said liberties requires acquiescence by all (which is safeguarded in the democratic decision-making process). It is in their capacity of decision-maker in the political arena that citizens determine how much decision-making power their consuming alter ego’s have. In short, the consumer benefit condition targets consumers in the relevant market because they have been appointed ‘sovereigns’ in that particular marketplace.⁸⁶

Consumer sovereignty also limits the use of ‘average consumer benefit’. If consumers are the ultimate decision-makers in the marketplace, average consumer benefit can only be used as a rough yardstick to measure full consumer compensation on the condition that competition is not eliminated. The reason for this is that market competition constitutes the ‘institutional embodiment of voluntary exchange processes’,⁸⁷ where transactions between producers and consumers take place based on ‘mutual benefit’.⁸⁸ This necessarily implies that, ultimately, it is the voluntary agreement to an exchange that evidences consumer benefit and thus full consumer compensation.⁸⁹

The ‘polluter pays’ principle and antitrust

A last point to consider is whether all of the above might be too short-sighted after all because the ‘polluter pays’ principle laid down in Article 191(2) TFEU mandates competition agencies to take account of non-consumer benefits when determining ‘fair share’.⁹⁰ It goes without saying that said principle, which holds that polluters rather than society must bear the costs of pollution, is an important principle in combatting climate change. Also the European Climate Law, for example, underscores that the Union’s and Member States’ actions must be guided by, amongst other, this principle.⁹¹ However, the relevant question is whether the ‘polluter pays’ principle instructs competition agencies to green up antitrust. Green antitrust proponents believe it does. In situations of inefficient regulation, market power would constitute a useful, alternative means for firms to internalize environmental costs. That this results in consumers paying anticompetitively higher prices would not pose a problem but instead concur with the ‘polluter pays’ principle because it would make them pay for the pollution they cause when consuming products that harm the environment.⁹² I disagree.

Article 191(2) TFEU does not address competition agencies when enforcing the competition rules that apply to undertakings. Article 191(2) TFEU is part of EU environmental policy, one of the various so-called cross-sectional policies that specify rules for the making of public policy within the EU and its Member States. Cross-sectional policy instructions do not touch on antitrust policy.⁹³ This is because antitrust policy does not concern the making of public policy but the enforcement of public policy already made: competition law constitutes the public policy and the institutional expectation is that competition agencies stick to the making of objective law enforcement policy, which precludes all but the protection of consumers against market power given the market regulation that is.⁹⁴ This institutional set-up concurs with the ways in which the ‘polluter pays’ principle is implemented.⁹⁵ First, through command and control regulation that makes polluters bear the costs of pollution by setting environmental standards including control systems. Second, through market-based regulation that uses financial incentives to influence production and consumption behaviour by internalizing environmental costs and benefits into the budgets of producers and consumers. Third, through voluntary approaches that encourage the production and consumption of less polluting products.⁹⁶ The two regulatory tracks most often command or incentivize firms to take measures that reduce pollution.⁹⁷ Efficient regulation fulfils the ‘polluter pays’ principle in that it ensures that firms' operational costs also include external costs that thereupon will be included as a cost item in the price. The fair result thereof is that, while pollution decreases, producers and consumers carry the same burden: producers bear higher production costs and consumers pay higher prices resulting from those higher production costs. If legislative government fails to produce efficient regulation, the ‘polluter pays’ principle still applies and instructs the legislature to do better. The third, voluntary track encourages less polluting production and consumption through voluntary agreements like, for example, ecolabels. Clearly, mainstream antitrust does not stand in the way of voluntary agreements.

In fact, greening up antitrust undermines the aforementioned fair result that the ‘polluter pays’ principle implicitly warrants. Crucial is that consumers no longer pollute in the situations that green antitrust would cover. For example, through setting a higher environmental minimum standard, private collective action ensures that consumers can no longer purchase products that cause the pollution the standard sets out to correct. This observation has two consequences for the application of the ‘polluter pays’ principle. One is that the principle, if applicable as claimed, has served its purpose: consumers no longer cause pollution for which they would have to pay. The other consequence is that, insofar as consumer benefits do not outweigh anticompetitive harm, producers apparently charge a higher price than necessary to produce the greener product. The unfair result of using the ‘polluter pays’ principle as proposed would thus be that producers over-profit while consumers overpay. Antitrust is intended precisely to prevent this outcome.

In short, while the ‘polluter pays’ principle points at the legislative government to act in case of market failures in terms of environmental costs, it does not instruct competition agencies to green up antitrust in case the government fails to act.

III. GREEN ANTITRUST ACCORDING TO THE COMMISSION

The above analysis confirms that the Commission is right to not introduce a citizen welfare standard for interpreting the ‘fair share’ requirement in the consumer benefit condition. Still, the Commission does agree that antitrust must green up to effectively supervise market situations in which individual production and consumption decisions cause negative externalities that the legislative government failed to address, or to fully address.⁹⁸ The EU Guidelines therefore design a new competition policy for all agreements between competitors that pursue one or more sustainability objectives.⁹⁹ The new policy thus includes but is not limited to climate change.¹⁰⁰ The greening up of Article 101(3) is shaped as follows. The efficiency condition is considered to include sustainability efficiencies since it encompasses ‘increases in product variety and quality, improvements in production or distribution processes, and increases in innovation’.¹⁰¹ As mentioned earlier, that interpretation of the efficiency condition is not up for debate. The consumer benefit condition is updated to broaden the scope for private collective action aiming to promote green. Even though ‘the benefits deriving from the agreement [must still] outweigh the harm caused by the agreement’,¹⁰² anticompetitive harm may also be compensated by consumer benefits that occur ‘irrespective of the consumers’ individual appreciation of the product’.¹⁰³ The indispensability condition is tasked to legitimize private collective action by way of filtering out greenwashing.¹⁰⁴ Bottomline is that market failures must necessitate producers to collectively overcome ‘so-called first mover disadvantages’.¹⁰⁵ The residual competition condition is trusted to facilitate private collective action as long as it does not eliminate competition on price or innovation. It is considered fulfilled ‘even if the [restriction] covers the entire industry, as long as [its participants] continue to compete vigorously on at least one important aspect of competition’.¹⁰⁶ Lastly, the EU Guidelines list the 1999 Commission decision in the CECED case as the prototype of a common minimum standard that meets the conditions of Article 101(3).¹⁰⁷ Does the new policy befit Article 101(3)?

The consumer benefit condition and collective consumer benefits

Mainstream antitrust hinders private collective action aiming to improve green since individual consumer benefits will not outweigh anticompetitive harm if consumers are not willing to pay the higher prices that accompany more costly greener products. Greener antitrust does not meet the ‘fair share’ requirement because employing a citizen welfare standard, it determines consumer compensation based on sustainability benefits that consumers do not enjoy. In an attempt to meet greener antitrust proponents ‘halfway’, the Commission expands individual consumer welfare analysis with collective consumer welfare analysis. ‘Individual consumer benefits’ include ‘individual use value benefits’ that ‘result from the use of the product and directly improve the consumers’ experience with the product in question’,¹⁰⁸ and ‘individual non-use value benefits’ that ‘result from the consumers’ appreciation of the impact of their sustainable consumption on others’.¹⁰⁹ ‘Collective consumer benefits’ occur ‘irrespective of the consumers’ individual appreciation of the product’ but are considered to nonetheless ‘objectively accrue to the consumers in the relevant market if the latter are part of the larger group of beneficiaries’.¹¹⁰ In short, while individual consumer welfare analysis builds on the preferences of individual consumers in the relevant market, collective consumer welfare analysis takes account of out-of-market benefits that consumers in the relevant market will experience as a result of the private collective action.¹¹¹ To illustrate, the EU Guidelines mention the example of car drivers who purchase less polluting gasoline.¹¹² Since car drivers are also citizens, they as well benefit from the cleaner air that results from the use of less polluting gasoline. This then would imply that, insofar as the group of driving consumers substantially overlaps with the group of benefitting citizens, these out-of-market benefits can be taken into account to demonstrate full consumer compensation.

But can out-of-market benefits motivate full consumer compensation even if limited to consumers in the relevant market? The EU Guidelines suggest that they can because the ‘substantial overlap between consumers and citizens’ requirement would trace back to the judgements in Compagnie Générale Maritime, GlaxoSmithKline Services, and MasterCard.¹¹³ This appeal to judicial authority seems somewhat out of place. In Paragraph 43 of the Guidelines on Article 101(3), the Commission posits that ‘where two markets are related, efficiencies achieved on separate markets can be taken into account [to establish ‘fair share’] provided that the groups of consumers affected by the restriction and benefitting from the efficiency gains are substantially the same’. The footnote to this paragraph informs us that this position is based on Compagnie Générale Maritime.¹¹⁴ In that case, however, the CFI discussed out-of-market benefits under the first rather than the second exemption condition without any reference to substantially overlapping consumer groups.¹¹⁵ In GlaxoSmithKline Services, the CFI simply referred to Compagnie Générale Maritime when discussing ‘appreciable objective advantages’ under the first exemption condition.¹¹⁶ In MasterCard, the ECJ did make a statement as referenced but again under the first exemption condition and in relation to other consumers in a two-sided market rather than to citizens in a non-market.

The question therefore remains whether out-of-market benefits can be taken into account to determine consumer compensation because they ‘objectively accrue’ to consumers in the relevant market when consumers and benefitting citizens substantially overlap.¹¹⁷ I believe they cannot. The reason is that collective consumer welfare antitrust overlooks that consumer benefits must pass the ‘agreement test’ for consumers to be compensated fully (see Section II).¹¹⁸ In a system of voluntary exchange processes, voluntary agreement is essential because consumers only get to be ultimate decision-makers if they are able to individually determine whether the benefits of more costly, greener products outweigh the accompanying higher price. Benefits therefore only ‘objectively accrue’ to consumers insofar as the latter value these benefits for themselves or for others. This means that the internalization of out-of-market costs through market competition and thus antitrust is necessarily limited to individual consumer benefits. Collective consumer benefits fail the agreement test obviously. To correct for consumers who are unwilling to pay for higher priced greener products, collective consumer welfare analysis resorts to benefits that private collective action forces upon consumers. This is inherent political decision-making: collective consumer welfare antitrust limits the choices of consumers in the marketplace, while citizens overall had agreed that these were legitimate choices to make.

It follows that collective consumer benefits cannot motivate ‘fair share’. Considering that the internalization of out-of-market effects is inherently limited to voluntary exchange in the market realm and voluntary agreement is essential for evidencing full consumer compensation, ‘fair share’ is necessarily limited to individual consumer benefits.

The indispensability condition and greenwashing

The indispensability condition has a straightforward task to fulfil under Article 101(3). It must rule out that an efficiency improvement claimed under the first exemption condition can be attained by a less restrictive agreement than the one up for clearance.¹¹⁹ In other words, to maximize the efficiency improvement claimed, the indispensability condition minimizes anticompetitive harm. In addition, the EU Guidelines task the indispensability condition to ascertain that private collective action is necessary to correct inefficiencies that result from the existing regulatory structure.¹²⁰ The baseline is that private collective action is not necessary where ‘[p]ublic policy and regulations have taken care of negative externalities’ or ‘where there is demand for sustainable products’.¹²¹ Necessity only arises if ‘the free interplay of market forces’ brings along that ‘sustainability benefits cannot be achieved’ because ‘consumers are unwilling to pay a higher price for [greener products]’.¹²² In these circumstances, private collective action may be used to jointly overcome ‘so-called first mover disadvantages’.¹²³ Phasing out agreements fulfil the new necessity test if necessary to discontinue polluting technology;¹²⁴ common minimum standards do as long as they leave ‘room for participating undertakings to individually apply a higher sustainability standard’.¹²⁵

In brief, the new necessity test is tasked to identify whether private collective action purports to improve green or pursue cartel overcharges. As such, it serves a two-fold goal. On the one hand, it legitimizes private correction of negative externalities that public policies and regulations failed to correct. On the other hand, it protects against greenwashing. To do so, it investigates whether the first mover disadvantage to be corrected is genuine or fake. First mover disadvantages are considered ‘genuine’ if consumers are unwilling to pay a higher price for greener products.¹²⁶ First mover disadvantages are considered ‘fake’ when overall consumer demand for sustainable products shows that market competition works.¹²⁷

This addition to the indispensability condition is puzzling. For one because the new necessity test is inapt to distinguish between genuine and fake first mover disadvantages. Necessity analysis compares the effectiveness of more restrictive arrangements to achieve an efficiency improvement with that of less restrictive arrangements. This method does not work here. Since the market mechanism is based on voluntary exchange it is inherently imperfect in internalizing external costs.¹²⁸ The result is that both genuine and fake first mover disadvantages require collective action to be corrected. The correction of a genuine first mover disadvantage requires collective action because consumers generally are unwilling to pay for green. The correction of a fake first mover disadvantage requires collective action because some consumers are unwilling to pay for green. Differently put, a first mover disadvantage claim may be fake in terms of overall consumer demand but collective action is nonetheless necessary to correct the external costs at hand.

The aforementioned inaptness relates to the institutional fact that the indispensability condition is not meant to legitimize private collective action. The reason for this is that constitutional market democracies operate on two efficiency levels.¹²⁹ A ‘regulatory efficiency’ level that is determined by the regulatory structure that only the legislative government is democratically legitimized to define and redefine. And an ‘exchange efficiency’ level that results from free exchange contracting within that regulatory structure. Market competition improves on regulatory (in)efficiency by improving exchange efficiency. Objective competition enforcement follows suit in that it protects competition in terms of voluntary exchange.

Moreover, first mover disadvantages are a ‘rare phenomenon’ as Schinkel and Treuren point out.¹³⁰ Empirical literature review shows that ‘firms can differentiate their products as more sustainable’, while ‘consumers do, in general and increasingly, have a willingness to pay for [sustainable products] that is great enough to make unilateral sustainability investments profitable’. They add that even if lack of consumer willingness to purchase green would lead to a deadlock preventing more sustainable production, sustainability agreements do not break that impasse other than by imposing prices above and beyond mere cost recoupment.

As a matter of fact, phasing out agreements and common minimum standards, both classified as genuine first mover disadvantage corrections in the EU Guidelines, actually concern fake first mover disadvantage corrections as cases like CECED, Chicken of Tomorrow, and Ancient Power Plants show.¹³¹ In CECED, 90 per cent of consumers purchased higher energy-efficient washing machines.¹³² In Chicken of Tomorrow, consumers did value higher quality chicken meat and market dynamics in the supermarket sector showed a trend of differentiation regarding animal welfare and sustainability.¹³³ In Ancient Power Plants, the five coal-fired power plants listed to be closed based on their high pollution rate covered about 10 per cent of Dutch power production capacity.¹³⁴ Differently put, it appears that the real first movers in these cases had been all but disadvantaged when advancing green.

Meanwhile, fake first mover disadvantage corrections all but improve exchange efficiency. CECED and Chicken of Tomorrow show that phasing out agreements and common minimum standards are prone to greenwashing since the participating firms primarily aim to collectively overcome ‘last mover competition’. In CECED, competitive pressure of cheaper, lower energy-efficient washing machines motivated manufacturers of more expensive, higher energy-efficient washing machines to phase out lower energy-efficient machines (more on this in the last paragraph of this section). In Chicken of Tomorrow, broiler farms of low-quality chicken meat and supermarkets selling low-quality chicken meat wanted to avoid competition when pressurized by animal rights activists to green up.¹³⁵

Last, not missing out on fake first mover disadvantage does not imply missing out on correcting regulatory inefficiency altogether. Ancient Power Plants shows that when ACM did not clear the agreement to discontinue coal-fired power plants because emission reduction benefits did not outweigh anticompetitive harm in terms of higher electricity prices, the Dutch government did take over and greened up Dutch energy supply by increasing public energy-efficiency norms.¹³⁶

Summing up, the indispensability condition cannot be customized as claimed. The market mechanism improves on regulatory (in)efficiency by way of improving exchange efficiency. Antitrust contributes to improving exchange efficiency by way of protecting competition in terms of voluntary exchange. In an attempt to bypass this institutional setting, the new necessity test fails to produce effective competition policy. Inapt to distinguish between genuine and fake first mover disadvantages, it does not add to the efficiency condition that already investigates whether private collective action is fit to correct external costs.¹³⁷ Allowing fake first mover disadvantage corrections to piggyback on the indispensability condition, the new test does not prevent but accommodate greenwashing.

The residual competition condition and protecting competition

The residual competition condition stipulates that an agreement may not ‘afford [the participating] undertakings the possibility of eliminating competition in respect of a substantial part of the products in question’. The interpretation of the last exemption condition is crucial for green antitrust: if it prohibits private collective action, corrections of negative externalities are a no-go to begin with. The EU Guidelines state that the condition ‘ensures that some degree of residual competition will always remain on the market concerned by the agreement, regardless of the extent of its benefits’.¹³⁸ Yet they also note that the condition does not protect competition in all its expressions but only in its most important ones, which means that the condition is still satisfied if private collective action eliminates competition ‘on quality or variety, [as long as] competition on price is also an important parameter for competition in the industry concerned and is not restricted’.¹³⁹ Likewise, the elimination of competition ‘for one or more of the variants of the product does not necessarily mean that competition in the relevant market is eliminated’.¹⁴⁰ This interpretation traces back to CECED and the Guidelines on Article 101(3). In CECED, the Commission introduced the principle of ‘compensatory competition’, which implies that a ‘restriction in one product dimension, energy consumption, may increase competition on other product characteristics, including price’.¹⁴¹ In that case, the Commission also held that the phasing out of lower energy-efficient washing machines did not eliminate competition since those product variants accounted for 10 per cent of the market only.¹⁴² In Paragraph 110, the Guidelines on Article 101(3) codify these findings in stating that to find out whether an agreement eliminates residual competition it is ‘relevant to examine [the agreement’s] influence on the various parameters of competition (italics added)’.¹⁴³ In so doing, the Guidelines on Article 101(3) attenuate the baseline expressed in Paragraph 105, namely that ‘[u]ltimately the protection of […] the competitive process [must be] given priority over potentially pro-competitive efficiency gains which could result from restrictive agreements’. The question now is whether attenuation is the right way to go. It would appear not.

The Commission’s position that the residual competition condition protects competition partially rather than fully poses a two-sided problem. On the one hand, partial protection disregards the principle of citizen sovereignty that implies that only the legislature has decision-making power to recalibrate the extent of voluntary exchange between producers and consumers that market regulation provides for. On the other hand, partial protection disrespects the principle of consumer sovereignty that instructs competition agencies to protect consumer choice regarding all competition dimensions the legislature allowed competition on. In short, both citizen and consumer sovereignty command that the duty to not eliminate competition applies fully.

Full protection of competition precludes the clearance of agreements that eliminate competition on quality or variety based on ‘compensatory competition’. Basically because the text of the residual competition condition enforces the principles of citizen and consumer sovereignty by way of prohibiting agreements once they allow the participating firms ‘the possibility of eliminating competition’ rather than limiting that prohibition to ‘the possibility of eliminating competition in its most important expressions’. In other words, the text of the residual competition condition stipulates that the degree of residual competition is investigated per competition dimension.

Full protection of competition also precludes the clearance of agreements that eliminate one or more product variants because they would lack substantial market power when all product variants are taken into account. Again, in line with the principles of citizen and consumer sovereignty, the text of the residual competition condition connects the ‘possibility of eliminating competition’ to a ‘substantial part of the products in question (italics added)’. So, if an agreement eliminates competition from one or more product variants, this implies that the ‘products in question’ concern the product variants to be eliminated. Insofar the participating undertakings produce and/or sell a ‘substantial part’ of these product variants, the agreement cannot be cleared because it allows ‘undertakings the possibility of eliminating competition in respect of a substantial part of the products in question’. Based hereupon, the discontinuance of low energy-efficient washing machines in CECED should not have been cleared. Since the washing machines to be eliminated covered 90 per cent of all low energy-efficient washing machines on the market,¹⁴⁴ the agreement obviously ‘afford[ed] undertakings the possibility of eliminating competition in respect of a substantial part of the products in question’.

It turns out that Paragraph 105 of the Guidelines on Article 101(3) hits the mark: the rationale of the residual competition condition is to protect the competitive process at all times. The residual competition condition is the ultimate guardian of the competitive process because competition law contributes to promoting efficiency by way of protecting competition in terms of voluntary exchange. Based thereupon, the residual competition condition precludes all private collective action. In so doing, it protects against ordinary cartels like price cartels that necessarily cover a substantial part of the products in question in order to be effective. But it also protects against green cartels that necessarily cover a substantial part of the products in question to overcome claimed first mover disadvantages. Precluding all private collective action, the residual competition condition rules out green antitrust.

CECED: greenwashing avant la lettre

Based on the above, it appears that the concept of green antitrust is not well thought out. This begs the question whether CECED set a good example to begin with. To reduce energy consumption and, based thereupon, pollution from electricity generation (CO₂, SO₂, and N₂O emissions),¹⁴⁵ the participating undertakings introduced a higher minimum standard to phase out less energy-efficient washing machines (cat. D to G machines).¹⁴⁶ The agreement thus pursued two objectives. One was to restrict competition by ‘controlling one important product-characteristic on which there is competition in the relevant market’.¹⁴⁷ The other was to reduce energy consumption and pollution. The Commission then had to sort out whether the reduction of energy consumption and pollution constituted the primary purpose of the new standard or covered for greenwashing. It concluded that there was no greenwashing: consumers would not be hurt because individual consumer benefits outweighed anticompetitive harm that would be smaller than anticipated under Article 101(1).¹⁴⁸ But is that so?

Individual consumer benefits only motivate consumer compensation if the pertaining restriction is indispensable to attain those benefits.¹⁴⁹ According to the Commission, the new standard was indispensable indeed because less restrictive alternatives were less effective. Information campaigns, for example, were considered ‘plainly less effective than a standard’ since ‘the provision of information [was] not sufficient to realize the agreement’s environmental benefits to their fullest extent’ because ‘external costs [were] not fully reflected in consumer’s calculations when contemplating a purchase’.¹⁵⁰

This reasoning appears to be flawed. Savings on electricity bills allowed buyers to recoup increased costs of upgraded, more expensive washing machines within nine to 40 months, depending mainly on the frequency of use and electricity prices.¹⁵¹ Considering that washing machines have an average life time of 12 years,¹⁵² that return and pay-back period was considered fair and reasonable.¹⁵³ This means that timely savings on electricity bills incentivized consumers to purchase energy-efficient machines, and in so doing, helped realize the environmental benefits claimed provided consumers were adequately informed. Considering that consumers could easily access information on cost-effectiveness of choice amongst different energy categories through the EC energy label and 90 per cent of consumers purchased higher energy-efficient machines,¹⁵⁴ it appears that consumers were adequately informed indeed.¹⁵⁵

It follows that the new standard targeted the remaining 10 per cent of consumers that did not buy higher energy-efficient machines even though cost-effectiveness information clarified that higher purchase prices would pay off timely. Still, these circumstances cannot fulfil the indispensability condition. Since the internalization of external costs through market competition is inherently based on voluntary exchange, antitrust cannot, in protecting that process, allow producers to force consumers to make more rational decisions—ultimate decision-maker-ship necessarily implies that consumers may make irrational choices as well.¹⁵⁶ Also, Viscusi and Gayer point out that irrational choices may proof to be less irrational when considering that ‘differences in preferences and income generate different levels of consumer demand for products’.¹⁵⁷ By analogy, it may well have been the case that consumers of lower energy-efficient machines had too little income to pay the higher price upfront. If so, the way to get these consumers on board in a voluntary manner would have been to set up a payment system that linked instalment payments to the period within which the savings on electricity bills allowed consumers to recoup the higher purchase costs.

The finding that anticompetitive harm would be smaller than anticipated poses a problem as well. For the Commission did not evidence its primary presumption that ‘the restriction of competition regarding one product dimension, energy consumption, may increase competition on other product characteristics, including price’ (italics added).¹⁵⁸ In lieu of substantiating that the elimination of actual competitive pressure of lower-priced lower energy-efficient machines would be compensated by a potential increase of competitive pressure amongst higher-priced higher energy-efficient machines, it merely stated that ‘it cannot be ruled out that [higher energy-efficient machines] may become available at a lower price’ and that ‘were these effects to take place (italics added)’, the narrowing of the price range and the increase in average selling prices would be ‘less pronounced than would otherwise be foreseeable’.¹⁵⁹

At the same time, the Commission neglected that manufacturers of higher energy-efficient machines faced ‘a stagnant market with strong competitors and research of sales opportunities’,¹⁶⁰ in which the new standard enabled them to control ‘purchasers [with] buying power’ that could otherwise ‘focus their orders on machines below category C’.¹⁶¹ Differently put, manufacturers discontinued lower energy-efficient machines to improve their profitability in the higher energy-efficiency segment, which, absent the restriction of competition, would have required intensified price rivalry. Again, the ECJ condemns this kind of behaviour because, even if it does not limit total output, it does change the structure of the market, in this case by eliminating a few product variants, which means that it hurts consumers per se.¹⁶²

It follows that the CECED decision condoned greenwashing avant la lettre. The agreement was not indispensable since overall consumer demand showed that market competition worked. By consequence, the individual consumer benefits referred to could not motivate consumer compensation under the consumer benefit condition. The agreement did not meet the residual competition condition either. Since the undertakings that agreed to phase out lower energy-efficient washing machines cover around 90 per cent of the relevant market, the agreement eliminated competition ‘in respect of a substantial part of the products in question’. So, in fact, the promotion of green covered up the real purpose of foreclosing consumers’ opportunity to purchase lower energy-efficient washing machines: to increase profitability through cartel overcharges.

IV. CONCLUSION

The EU Guidelines are based on three presumptions. That antitrust can green up to correct negative externalities by accommodating market power. That the last three conditions of Article 101(3) can be customized to clear green cartels. That CECED exemplifies a bona fide green cartel. The above analysis shows that all three presumptions are wrong.

Antitrust cannot green up because private collective action inherently limits the scope of voluntary exchange between producers and consumers that the existing regulatory structure offers. In market democracies like the EU and its Member States, only the legislature is democratically legitimized to define and redefine the scope for voluntary exchange. This means that markets operate on two efficiency levels. A ‘regulatory efficiency’ level that is determined by the legislative government. And an ‘exchange efficiency’ level that results from free exchange contracting within that regulatory structure. Protecting competition in terms of voluntary exchange, antitrust ensures that market competition improves on regulatory (in)efficiency by way of improving exchange efficiency.

By consequence, stretching the last three conditions of Article 101(3) to accommodate green cartels does not work. While the efficiency condition connects competition analysis to the goal of market competition—improve economic efficiency, the other three conditions connect competition analysis to market competition as a means—voluntary exchange processes between producers and consumers.

The consumer benefit condition connects the ‘fair share’ requirement to consumers because they have ultimate decision-maker-ship in the marketplace. In case of external costs, consumers have ultimate decision-maker-ship because citizens, represented by legislative government, chose market competition over market regulation to internalize those costs. Ultimate decision-maker-ship implies that it is up to the consumer to determine whether the efficiency improvement that an anticompetitive agreement brings outweighs the pertaining anticompetitive harm in terms of price increase. This means that ‘fair share’ translates into full consumer compensation based on benefits that consumers value for themselves or for others—individual consumer benefits.

Collective consumer benefits (Commission) cannot motivate the ‘fair share’ requirement because they do not ‘objectively accrue’ to consumers in terms of voluntary agreement. Other than claimed, case law does not permit the use of out-of-market benefits insofar as the groups of consumers and benefitting citizens are substantially the same. Compagnie Générale Maritime and GlaxoSmithKline do not mention the requirement of substantially similar consumer groups. MasterCard does but under the first rather than the second exemption condition and in relation to other consumers in a two-sided market rather than citizens in a non-market.

Non-consumer benefits (ACM) cannot motivate the ‘fair share’ requirement because they do not accrue to consumers in the first place. Case law confirms that finding. While ‘appreciable objective advantages’ under the efficiency condition may include non-consumer benefits (Consten and Grundig, Landewyck, and Compagnie Générale Maritime), ‘fair share’ under the consumer benefit condition is limited to consumer benefits (Shaw, Asnef-Equifax, MasterCard—GC/AG, Sainsbury’s Supermarkets—UKSC). The newly proposed public objective defence appears ill-conceived. The very existence of external costs evidences that the State chose to not compel private firms to internalize those costs by way of restricting competition. The ‘polluter pays’ principle does not mandate green antitrust. For one because its implementation through voluntary agreements requires protecting competition and thus mainstream antitrust. But also because green cartels do not make consumers pay for the pollution they cause since the private collective action at issue inherently implies that the products consumers purchase no longer produce the pertaining external costs. This implies that a price increase that does not meet the full consumer compensation standard constitutes an ordinary cartel surcharge.

The indispensability condition protects against restrictions of competition that are unnecessary to realize the efficiency improvement claimed. This condition is not meant to legitimize private collective action aiming to correct negative externalities because antitrust protects against exchange inefficiency, not regulatory inefficiency. Based thereupon, the necessity test operates within the free exchange spectrum that makes an investigation into a potentially less restrictive alternative practically relevant. This does not apply to the new necessity test. Because negative externalities require collective action to be corrected, an investigation into a potentially less restrictive alternative is otiose. The new test confirms this: both genuine and fake first mover disadvantages require collective action to be corrected. Fake first mover disadvantages like phasing out agreements and common minimum standards are prototypes of greenwashing. Freeriding on the fact that market competition internalizes external costs imperfectly, these arrangements do correct residual external costs but against cartel prices that outmatch the residual benefits. Accommodating phasing out agreements and common minimum standards, the new necessity test accommodates rather than prevents greenwashing.

The residual competition condition protects against producers eliminating competition (in respect of a substantial part of the products in question). The principles of citizen and consumer sovereignty ordain that this exemption condition cannot allow producers and competition agencies to regulate competition respectively eliminate legitimate consumer choice. This has two practical consequences. First, the duty to protect competition applies to all competition conditions rather than important competition conditions only. Second, residual competition must be ascertained against the products/product variants that an agreement eliminates competition on. The result is that the residual competition condition prohibits all private collective action and thus all cartels. As such, it rules out green antitrust.

A revisit of CECED shows that green antitrust is based on a false narrative. Because market competition worked, the higher minimum standard was all but indispensable to promote green. Meanwhile, the promotion of green covered up that the elimination of competition from lower energy-efficient washing machines allowed industry to raise prices above competitive levels.

So, even though the EU Guidelines fare better than the Dutch Guidelines in their attempt to stick to the consumer welfare standard, the Commission is well-advised to revise Chapter 9 of the EU Guidelines. Antitrust is sufficiently green as it is.

Footnotes

ACKNOWLEDGEMENTS

The author thanks Jan Kees Winters, Maarten Pieter Schinkel, and two anonymous reviewers for their helpful comments. Any remaining errors are obviously mine.