Chillin'Competition

[There is a great deal of (passionate) discussion around sustainability agreements in antitrust these days. Last week, the Commission issued a brief which touched upon this matter, among others relating to Europe’s green agenda. Coincidentally, we featured a guest contribution by Maurits Dolmans (see here). Maarten Pieter Schinkel, Professor of Economics at the University of Amsterdam, has prepared a respose, which we are delighted to publish and which you will find below]

In a recent blog on Chillin’Competition, Maurits Dolmans names me as one of “a few economists” who warn against the perverse incentives of permitting competitors to make agreements in exemption of the cartel prohibition. But all economists know as first principle that incentives matter! See, for example, Cowen & Tabarrok (2021), chapter 1, page 1. Maurits Dolmans is right that its assumptions make a model, and also that their reasonableness for the model’s message should be scrutinized. However, he misrepresents our argument. In fact, none of the assumptions that Maurits Dolmans says we make – and then criticizes – we actually do make. Our latest academic paper – to which he refers, and I appreciate that – speaks for itself. We laid out our argumentation accessible in a chapter in the Concurrences book on sustainability and competition law and in a ProMarket article. Nevertheless, I feel compelled to reply here briefly, following the order of his bullet points.

Just to be clear, I think we all agree that the central question in this debate is whether allowing competitors to make sustainability agreements induces them to take greater effort to produce more sustainably than when they wouldn’t be able to make such agreements and remain in competition without. After all, more sustainable production is the end (which we share), and allowing anticompetitive sustainability agreement is the proposed means (which is on the table). Only if the answer to this central question is reasonably “yes” does it make sense to see how sustainability agreements may be exempted from cartel law and permitted – which is where the exemption requirements come into play.

We arrive at “no, likely not”, on the basis of a solid analysis – and not in the way Maurits Dolmans seems to understand it:

• Maurits Dolmans writes: “Consumers are assumed to be willing to pay as much as is needed to avoid climate damage” This is not true. We need no conditions on WTP: for either positive, zero or negative WTP, firms invest (weakly) more in green when in competition than when they would be allowed to coordinate their green investments. If WTP is not positive, then firms will not invest in green, whether they compete or coordinate on it – period. Yet whenever there is the smallest WTP for more sustainable products, green is promoted more in competition than cooperation. So the answer to the central question is “no” – i.e. not “yes”. The intuition is not that hard: coordination kills the companies’ drive to steal business from each other by offering a greener product.

He continues: “ … and it is always profitable for firms to meet that demand” This is hardly a heroic assumption: firms choose their investments in greener production methods optimally, by weighing the benefits from being able to charge a higher price – resulting from consumers’ willingness to pay more for the greener product – against their green investment costs. This is no different from normal business decisions.

Maybe Maurits Dolmans has some threshold fixed costs for transition in mind, that couldn’t be overcome in competition. But why would firms coordinating their efforts – and not also their prices, mind you, because that is explicitly not allow under the proposed policy – make such loss-making investments?! Hence, green investments are larger in competition than when competitors coordinate their green investments.;

• Maurits Dolmans states: “Regulation is assumed to offer a fully effective solution” This is not true. Our comparison is between competition and (private) coordination – to see if the latter can improve upon a situation in which regulation is suboptimal, for otherwise there is no need. Our point is that allowing competitors to coordinate on doing less green than they would do in competition will make matters worse, not better. So allowing agreements on sustainability is not a fix of government failure to regulate properly.;

• He writes that: “Firms are assumed to benefit only from (and to seek only) short-term profit maximization,” This is not true. We show that also when firms are driven by intrinsic motivation to invest in green (which may be related to long-run business interests), does coordination erode the incentives to invest in greener production methods.

We would also assume that firms: “always collude to minimize green investment or greenwash if they can get away with it”. Yet this is not an assumption, but a finding.

Maybe Maurits Dolmans means to say that companies should be trusted to invest in green collectively against their own commercial interests. But given what is at stake, shouldn’t we have some commitments at least that this is warranted before we do that?! It is perfectly reasonable to take companies’ business interests as the main driver of their corporate behaviors.

• His final bullet point reads: “It is assumed that consumers must be fully compensated for any price increase. Out-of-market benefits or improved access to non-market goods (say, clean air or a safe environment) supposedly do not count as compensation.” – after which he goes on to discuss these (legal) issues at length. Yet this is not true. Our argument does not rely on full (or any) consumer compensation at all. At the end of our paper, we do have a quick look at out-of-market benefits towards satisfying the compensation requirement and suggest a way to assess them – only to conclude that those too are less under coordination of green investments.

Maurits Dolmans concludes his blog with his title promise to offer three criteria to distinguish ‘beneficial cooperation from greenwashing’. Again I am featuring: “Prof. Schinkel has shown that where consumers are willing to pay for green products, firms may have an incentive to collude on greenwashing.” Well, properly understood what we show is that if investments in more sustainable production methods are costly – which no one denies is the case – firms that are allowed to coordinate those investments have incentives to reduce them below their levels in competition, in order to save on those costs. Collaborators will find it optimal to do a binding minimum needed for the exemption – i.e. wherever consumer compensation is required. The policy that Maurits Dolmans advocates, in other words, sets the firms up for greenwashing.

Our advice: drop the policy. But Maurits Dolmans want to keep pursuing it, and somehow curtail its perverse incentive effects: “So, how to distinguish a legitimate sustainability agreement from a cartel?” Alas, his first criterion is no more than “there should be a market failure” – and subsequently the same false logic just debunked. His second criterion is a loose reference to legal conditions for exemption that have nothing to do with our warning for perverse incentives – as said also. The third criterion has some substance. Reference is made to a forthcoming Oxera paper, which constructs some theoretical examples of conditions under which collaboration can increase sustainability investments. The core idea behind those examples is that firms would gain privately, through ‘spillover benefits’ from their rivals’ investments in more sustainable production. That is, firm A would see its profits increase (and/or intrinsically like it a lot) if firm B offered consumers a more sustainable product. If such positive spillovers are large enough to overcome the negative effect on sustainability investments from eliminating business-stealing by offering the greener product, they may lead to higher sustainability levels in collaboration than in competition.

But what would such sizeable spillovers be? Of course we had thought of this possibility too – which is standard in the related R&D literature. In the present matter, however, it is not obvious that there are any such spillovers. Because how exactly would it benefit Procter & Gamble if Unilever reduced its emissions from transportation by offering compressed deodorants first? And would E.ON really accept a profit hit for the warm glow of knowing that Vattenfall poaches its customers by offering green electricity? Rather the opposite is the case: one firm cleaning up its act makes the other firms only look dirtier. They are in product quality competition: consumers increasingly prefer more sustainable products.

Spillover benefits from R&D are that one firm’s inventions give others free ideas on how to improve themselves as well. Yet the green antitrust debate is about transitioning to known cleaner or fairer production methods. That is a private cost-benefit analysis. Few spillovers there. The examples so far presented of green cartels that might work are rather constructed therefore. For claiming sufficiently large spillover effects to justify collaboration, it certainly does not suffice to remark that firms also suffer from the cost of climate change, nor to loosely refer, as Maurits Dolmans does, to corporates supposedly believing today “that it is worthwhile to pursue longer-term survival of our environment and, therefore, themselves, our economy, and society as a whole.” Again, the corporate incentives involved need to be well understood, before proceeding to such drastic measures as permitting the elimination of competition.

There is a lot at stake in this debate. There is a real risk of cartel greenwashing attached to what Maurits Dolmans and other proponents advocate. I share their grave concerns about climate change. Yet if we get this wrong, we risk to worsen, instead of curbing it. I submit that we should be wary of corporate incentives to weaken the cartel prohibition. We contribute: Competition is likely a stronger driver of sustainability investments than cooperation. The assumptions and logic in our analyses are valid and transparent. Maurits Dolmans and I have discussed and corresponded about them extensively. Still I need to correct him here on what we actually say. I am open to further discussion – certainly about spill-overs, which is where Oxera’s Agenda is contributing some fringe examples in which coordination may sometimes work. But a prerequisite to a(ny) fruitful discussion is representing the other’s arguments properly. Only then can we advance our knowledge on these important issues. For now, our findings stand: Green collaboration reduces green efforts under the circumstances at hand.