This article, available here, argues, contrary to the arguments made in the piece above, that the Supreme Court decided the Ohio v American Express case correctly. Multisided platforms have distinct and critical features that set them apart from single-sided markets. Any prima facie antitrust assessment of competitive harm must incorporate the impact on consumers in all sides of a market regardless of market definition, and output effects should be the primary emphasis of any such competitive effects analysis.
The paper is structured as follows:
Section 2 identifies two broad schools of thought on market definition and competitive effects for multisided platforms.
There is a divide among antitrust practitioners, courts, and economists regarding how multisided platforms should be assessed in antitrust investigations.
A first school advocates for a separate effects and markets’ approach. Because users on different sides of a platform have different economic interests, it is inappropriate to view platform competition as being for a single-product offered at a single (i.e., net, two-sided) price. While acknowledging the relevance of cross-market effects in market delineation, this approach rejects incorporating both market sides in the competitive effects analysis. A core justification for a separate-markets approach is the notion that the two sides are not interchangeable substitutes and do not include the same market participants (i.e. customers). Given separate markets, the analysis of competitive effects must also be restricted to each of the separate markets. Further, cross-market effects should be not be incorporated as an efficiency defence since these are “out-of-market benefits” and balancing must occur “within the relevant market”. Variations of this approach include acknowledging a degree of interrelationship between the various sides of a platform but arguing that legal or pragmatic considerations prohibit or counsel against incorporating that reality into antitrust analysis; or acknowledging the potential role of cross-group effects, but only as an efficiency defence to a prima facie finding of harm to a group of consumers on one side of a multisided market.
A second school adopts an integrated-effects approach. The central rationale is that a relevant market must be sufficiently broad that, if it were monopolized in fact or by hypothesis, the monopolist of the conjectured market would be able to profitably exercise market power within that market. Defining separate markets for each side will lead courts to view market power and anti-competitive conduct within the four corners of that market. While there is some divergence in views regarding the need to define one or two relevant product markets, this second school is generally consistent in the view that an assessment of whether there is anticompetitive harm must be incorporated at the prima facie burden stage of a rule-of-reason analysis.
Section 3 argues for the superiority of the integrated effects approach.
The economic literature has clearly established the interrelationship between two sides of a platform as regards profit maximization. The definition of the exercise of monopoly power—the reduction of market-wide output and increase in the market price—cannot be satisfied by evidence of a price effect on only one side of a given platform. Consequently, one cannot assess market power by looking only at half of the profit maximization equation as relevant evidence.
Within the three-step burden-shifting framework for assessing antitrust cases involving platforms, the initial burden of finding anticompetitive harm must necessarily involve an assessment of both sides of a platform. A price change on one side of a platform can imply an increase, decrease, or neutral change in market-wide welfare. What matters is the structure of the interrelated relative prices – not the price levels themselves. It follows that demonstrating harm to one side of a two-sided platform is insufficient to establish that market-wide consumer welfare decreased.
To that end, given the importance of the relative structure of prices, how should the competitive effects analysis actually be conducted and measured? Justice Thomas’s majority opinion properly focuses upon output as the key metric in the competitive effects analysis because “[m]arket power is the ability to raise price profitably by restricting output”. Measuring output effects – and more specifically, the impact of business conduct on market output – is the central purpose and ultimate aim of welfare analysis. Consequently, measuring quantity differences, rather than price, should take primacy in evaluating platform effects. It is true that a common approach to measuring changes in consumer welfare is to test whether business conduct results in a change in market prices. For single-side markets, there is good reason to adopt this approach since price can serve as a reliable predictor for changes in output and consumer welfare. For platforms, however, focusing solely on prices is complicated by the fact that there are two prices that ultimately determine output. Consequently, output, rather than price, should be the touchstone for measuring differences in consumer welfare, for markets that involve platforms.
Section 4 demonstrates why, even if courts define separate product markets, they should adopt an integrated-effects approach.
There are two reasons that courts which adopt a separate-markets approach can and should still adopt an integrated-effects approach to competitive effects analysis. Firstly, even for markets that are not as closely interrelated as two sides of a platform, courts and agencies already utilize the concepts of (1) out-of-market efficiencies and (2) cluster markets to combine separate product markets into a single competitive effects analysis. The approach to out-of-market efficiencies that is reflected in the US Horizontal Merger Guidelines allows for two separate product markets but a single competitive effects analysis when the markets are sufficiently linked. Secondly, separate markets are also consistent with an integrated effects analysis because this approach allows courts to avoid differential treatment of cross-group effects across the market definition and competitive effects stages of analysis.
Section 5 concludes.
I am not able to provide a valuable opinion on a matter that has given rise to opposing views from so many distinguished and expert economists. This is why, against my common practice, I did not comment on the first piece reviewed today. I would nonetheless like to make a few points.
Firstly, it is interesting that the academic debate about this issue seems to take place mainly in the US. I think this has to do with the way competition law is enforced there – before the courts, in the context of a legal framework that requires evidence of prima facie anticompetitive harm. When competition agencies take decisions in administrative systems, they do not need to concern themselves with whether such prima facie evidence is available and can look at competitive effects in the round. Of course, related questions, such as whether the anticompetitive effect must encompass the two sides of the platform, are also relevant in administrative systems – this was one of the issues at stake in the credit card litigation in the UK. Regardless, it may be that administrative systems are better placed to deal with the fact that one needs to take into account of cross-platform effects in multi-sided markets, regardless of whether an integrated or separate markets’ approach is adopted.
The second point is a common refrain of mine – two-sided markets bring to the forefront questions concerning the basic underpinnings of competition law, such as what it is that it protects (e.g. consumer welfare, total output, competitive process), in a way that most traditional competition enforcement actions do not. I was particularly taken by how supporters of an integrated approach to platform markets have found it necessary to argue – correctly – that just finding harm to a group of identifiable consumers is not sufficient to establish competitive harm under competition law (e.g. monopoly pricing is legal). What is necessary is an anticompetitive act that creates such harm. Of course, what this does is bring the analysis all the way back to the question of what types of conduct are anticompetitive, and why – regardless of their impact on consumer welfare, supposedly the linchpin of effects-based competition law.
A third point is a practical one. The authors of this latter paper say that output should be the measure of whether a practice pursued by a two-sided platform is anticompetitive. However, this begs the question of output related to what. After all, if output grows more in the counter-factual world than in the current world, the practice would still be anticompetitive. However, the authors do not really engage with the fact that the US Supreme Court seemed to find that any output increase is evidence that a practice is not anticompetitive – a controversial point, to say the least.