Chillin'Competition

For the many amongst us who are trying to keep up with Nicolas Petit’s scholarship, Big Tech and the Digital Economy: The Moligopoly Scenario will feel both familiar and new at the same time. In this new book, Prof. Petit treats us to his usual incisive and pathbreaking analysis, based on a solid comparative and inter-disciplinary methodology. At the same time, the monograph format allows him more time and place to venture into more fundamental issues or more extensive research than a journal article allows.

There is much to like in this book, including Chapter II, where Petit further develops the use of 10-k forms as an alternative source of empirical data for research into competitive relationships, and the end of Chapter IV, where he puts together a game-theoretical model that seeks to explain the entry pattern that are observable when established players square off against each other or confront new entrants. No doubt these will prove seminal contributions to academic and policy debates.

As a fellow researcher on matters of innovation and competition law, and given the limited scope of this blog entry, I will focus on three points where the monograph paves the way for further research.

What is the nature of moligopolistic competition?

Petit is absolutely correct to emphasize that competition law overlooks the competitive relationship between the large established Big Tech firms – the MAGNAF, as they are rechristened once Petit adds Netflix to GAFAM. But what is the nature of this competitive relationship? Petit’s moligopoly hypothesis is that Big Tech firms are “simultaneously… monopolists and competitive firms”. They are subject to “a form of pressure… from firms outside of the product and service markets that they serve alone [and] also from indeterminate firms, markets or industries” (at 34). This hypothesis owes much to Baumol’s contestable markets theory, whereby the fear of “hit-and-run” entry is sufficient to discipline even a monopolist. Baumol’s theory has had limited practical significance in competition enforcement, and Petit attempts to enrich it so as to give it more salience. He differentiates between tipped and untipped markets, upon which he superimposes a distinction between direct and indirect entry, to explain the strategic preference of Big Tech firms for indirect entry into untipped markets (at 155-167). However impressive, Petit’s analysis, I would argue, fails to catch the entirety of the competitive interaction amongst Big Tech firms, because it remains stuck in the established analytical framework of competition law (much like other offspring from Baumol’s theory, such as the idea of competition for the market as an alternative to competition in the market).  

Indeed competition lawyers and scholars are used to thinking of markets as exogenously determined (through product characteristics and use or econometric data) and static over time, whereas they are actually one of the main competitive parameters amongst Big Tech firms. There firms compete on the market, like they compete on product features, quality, etc.[1] As Petit alludes to at 88, firms face the risk of becoming competitively irrelevant, i.e. of being pushed to the margins of the overall ecosystem, in spite of their continuing market power on their narrowly defined market of origin. This was the fate of Microsoft when its operating system became a sideshow once the center of action moved online with the Internet. This was also the challenge faced by Google when smartphones, led by the iPhone, threatened to bypass its search engine and other online properties by transferring the bulk of online traffic directly to apps and confining its search engine to the long tail of unusual queries (hence the development of Android to counter the threat). Shifting markets around is the essence of disruptive innovation.[2] Hence, I would argue, indirect and direct entry, as Petit defines them, are closely linked: competition amongst Big Tech is not frontal, but sideways, aiming to push the competitor to the periphery of the ecosystem, and indirect entry could simply be a more sophisticated form of direct entry.

What can competition law achieve on moligopolistic markets?

Secondly, Petit gives a detailed set of guidelines on the role of competition law in moligopoly markets. He correctly advocates more robust merger control against horizontal mergers in tipped markets (with a presumption of harm, at 198-202), and against “killer acquisitions” (at 222-223). In insightful analysis reminiscent of his previous work on traditional oligopolies, he also focuses on applying competition law to “ancillary” conduct on the sides of a multi-sided platform, such as might affect the functioning of the platform and ultimately consumer welfare (at 193-198). Yet when it comes to the behaviour of Big Tech firms on untipped markets, Petit proposes to rely on competitive pressure (i.e. contestability) to discipline firms, and to leave behind the maintenance of rivalry as a goal of competition law (at 184 and 224-225).

At the core of Petit’s analysis lies the proposition that competition amongst Big Tech firms is sufficient to lead to an optimal allocation of resources to invention (at 167), and hence that entry by newcomers does not deserve special protection and could even be discounted (at 206-207). I would argue to the contrary that when it comes to innovation (let alone disruptive innovation), established firms and newcomers behave differently, since the former have something to lose. They loathe cannibalization of existing revenue streams.[3] The ongoing competitive race between Zoom and Microsoft Teams illustrates this dynamic very well: Microsoft’s behaviour is influenced by the need to develop Teams within its existing product suite. Because newcomers and startups bring a different approach to innovation, it would seem justified to devote particular attention to keeping all avenues open for them, so that they have a fair chance to carry out innovative product strategies.

Accordingly, the maintenance of rivalry would remain a valid objective for competition law, even in untipped markets. As Petit notes at 226, there is uncertainty on these markets, as on all markets where innovation matters, because of the inherent unpredictability of innovation: invention very often involves serendipity, and furthermore the process of diffusion and adoption that turns an invention into an innovation is itself contingent.[4] Partly because of that, economics offers no methods or tools that could ascertain dynamic market performance with the same level of accuracy as is achievable with static performance on parameters such as price or output.[5] Against that background, it would seem prudent for competition law to step back from letting enforcement be guided by a market performance goal (consumer welfare), and retreat to the maintenance of rivalry as the best that competition law can achieve in a context of uncertainty.[6]

Are all MAGNAF similar?

Amongst the many praiseworthy features of this monograph, Petit makes extensive use of business literature, next to the more standard economics sources that competition scholars are used to work with. Business literature can be very enlightening in any discussion of innovation, because it offers perspectives that are considered somewhat off the beaten path by economists. In particular, one of the most influential strands of business literature on innovation, starting with Christensen’s famous Innovator’s Dilemma,[7] builds on heterodox neo-Schumpeterian economics, such as Nelson and Winter’s evolutionary theory.[8] Using business literature contributes to giving a proper place to neo-Schumpeterian economics in legal scholarship.

One of the lessons from evolutionary theory, which is carried over in the business literature, is that rival firms cannot be assumed to follow a similar production function. Each firm has unique capabilities, which are tested when firms must adjust to changes in their environment by seeking new routines. The most capable firms survive. Indeed Petit notes that MAGNAF is a fairly heterogenous set of firms (at 63), with their respective culture (at 141-142).

Transposed in competition law terms, the uniqueness of each firm would imply a far more differentiated analysis, that takes into account, for instance, the respective SWOT analysis and strategic orientations of each firm. Unfortunately, this would be paired with increased enforcement costs and a loss of legal certainty. The precise contours of a competition law that would pay more attention to the unique capabilities of each firm remain to be drawn.

In particular, remedies could also be more tailor-made, including measures that competition law has so far seldom used, such as obligations to implement open standards or achieve interoperability (briefly mentioned at 204-205) or even governance remedies, whereby a dominant platform could be subject to a specific mode of governance involving stakeholders (a form of governance separation remedy). In this respect, the lessons learned from the application of competition law to standard development organisations could be usefully brought to bear.

---

[1] For more details, see Larouche, “Platforms, Disruptive Innovation, and Competition on the Market”, CPI Antitrust Chronicle, February 2020, which is part of a larger ongoing research project.

[2] J. Gans, The Disruption Dilemma (Cambridge: MIT Press, 2016) provides an overview of the literature, which started with the work of C. Christensen.

[3] With the possible exception of Amazon when it came to moving from printed to e-books, and Netflix when jumping from mailing DVDs to streaming content. For that very reason, these two cases are usually presented as examples of “heroic” management.

[4] See the discussion in A. Butenko and P. Larouche, “Regulation for Innovativeness or Regulation of Innovation?” (2015) 7 Journal of Law, Innovation and Technology 52-82.

[5] Leading to what David S. Evans and Keith N. Hylton, “The Lawful Acquisition and Exercise of Monopoly Power and Its Implications for the Objectives of Antitrust” CPI Journal, Competition Policy International, vol. 4, refer to as a ‘tractability bias’ in favour of mathematical models of static competition (given the complexity of dynamic models), which induces a ‘static competition bias’ in competition law.

[6] As pointed out in P. Larouche and M.P. Schinkel “Continental Drift in the Treatment of Dominant Firms: Article 102 TFEU in Contrast to § 2 Sherman Act” in D. Sokol and R. Blair, eds., Oxford Handbook of International Antitrust Economics – Vol. 2 (Oxford: OUP, 2014) 153-187, this can be seen as a rebooting of ordo-liberalism out of concern for dynamic efficiency.

[7] C. Christensen, The Innovator’s Dilemma (Boston: Harvard Business School Press, 1997)

[8] Richard N. Nelson and Sidney D. Winter, An Evolutionary Theory of Economic Change (Harvard: Belknap Press, 1982).