The dynamic competition school claims that competition authorities, by analyzing firms’ capabilities, can protect what this school calls dynamic competition. Competition authorities, however, cannot adequately analyze firms’ capabilities. This school, therefore, lacks any framework which the authorities can use to analyze what they call dynamic competition. Any framework to protect dynamic competition, or what some call competition to innovate, must instead first identify the future products the competing firms are trying to make, writes Larry Landman.
The authors of “Dynamic Competition Is (Also) a Pro-enforcement Framework” imply in their title that they have developed a framework which allows competition authorities to protect what they call dynamic competition. These authors believe that a competition authority, by analyzing firms’ capabilities—their resources and ability to innovate—can protect what the dynamic competition school calls dynamic competition. The authors believe that an authority can do this even if the authority has not identified the competing products the parties to the transaction are trying to make. Further implying that they have developed this framework, the authors also say at the end of their article that analysts must develop “analytically precise” frameworks which will allow competition authorities to properly analyze this dynamic competition.
But the authors fail to provide any such framework. As the authors say, since the European Commission has made innovation the center of its new draft Merger Guidelines, developing such a new framework is indeed crucially important. If competition authorities are to protect innovation in the modern economy, they must have an analytically precise framework which allows them to protect innovation. But in none of the four cases the authors analyze in depth, their description of other cases, nor in any article the authors link to do they provide any analytically precise framework. This is because, first, as I explain in greater detail below, a competition authority can only block a transaction if the parties to the transaction are both trying to make the same future product and are thus competitors. Second, as I also explain in greater detailbelow, despite what these authors claim, competition authorities cannot adequately analyze firms’ capabilities.
The authors’ article suffers from three fatal flaws. First, it claims competition authorities can protect dynamic competition without the authority identifying a future product the relevant firms are trying to make. Second, the authors note that in many cases, competition authorities have pointed out that the relevant firms were trying to make better versions of the products they already sold. But this in no way shows that the authors, or the competition authorities, have developed an analytically precise framework which allows the authorities to analyze what the authors call dynamic competition. And third, the authors claim that a non-existent framework deserves credit for protecting “dynamic competition” in what is just a standard, run-of-the-mill vertical foreclosure case.
The first case the authors analyze in depth illustrates the first two of the three fatal flaws. Quoting from a press release and an article written by United States Department of Justice officials, the authors claim that in Applied Materials/Tokyo Electron the DOJ “did not predict which future products or services would not be supplied following the merger.” Yet, the authors then say the relevant firms were “the two largest competitors” in what the authors themselves call the market for “high-volume non-lithography semiconductor manufacturing equipment.” Thus, these authors themselves identified the market in which they believe the DOJ protected competition: that for high-volume non-lithography semiconductor manufacturing equipment. If the firms merged, one less firm would make such equipment.
The article the authors cite first identifies the relevant market as that for “leading-edge semiconductor tools for high volume manufacturing.” It then focuses on a specific market for such tools, that for “deposition and etch semiconductor tools.” The article says that the merger would have led to a “significant reduction in competition and possibly even monopolization” of the market for these “deposition and etch semiconductor tools.” The firms obviously competed in the market to sell such currently existing tools, and thus competed in the relevant current market. And since the merged firm may have been able to monopolize this current market when the authors say “From a static competition viewpoint, the deal was unproblematic” they are simply wrong. In fact, DOJ seems to have acted reasonably when it considered blocking this transaction simply to protect competition in this current market.
The article the authors cite also says that the firms were competing to make better versions of the products they already sold, that for better deposition and etch semiconductor tools. The firms were thus also competing in what I call a “Future Market,” a market for products at least some of which do not exist yet. In this case the firms were competing in the Future Market for better deposition and etch semiconductor tools. Again, a merger would have eliminated the tools one of the firms would have made.
The article says that the DOJ found that the firms had assets and capabilities they would use to make these better versions of their currently existing products. But, as I explain in greater detail below, of course the firms have at least some of the skills and resources they need to make better versions of the products they already sell. Any firm trying to make a new product will have at least some of these skills and resources. If it did not, then it would not try to make the new product. The article says only a few firms had the assets and capabilities to make these future products. Thus, it seems, only a few firms competed in the relevant Future Market. The DOJ, therefore, acted reasonably when it also sought to protect competition in this Future Market.
In this case, the DOJ protected competition in a current market and a related Future Market. As this shows, and as I also discuss in greater detail below, a competition authority can only block a transaction so as to protect competition to innovate if the relevant firms are making products which, if they exist, will compete against each other in the future. In other words, contrary to what the authors claim, the competition authority must identify the competing products the firms are trying to make.
I extensively analyzed the second case the authors cite, Visa/Plaid, in my article “Nascent Competition and Transnational Jurisdiction: The Future Markets Model Explains the Authorities’ Actions.” In its complaint in this case, the DOJ did indeed identify the market in which both firms competed, the “online debit market.” Visa already sold a product which competed in this market and Plaid was developing a better version of this product. Plaid’s product would allow customers to make online payments directly from their banks, eliminating the need for middlemen such as Visa. The authors even acknowledge that Plaid was developing a “competing online debit-processing service.” Thus, the authors identified the product both firms made.
The authors say that since Plaid made a better product, its market share would probably grow. The authors thus claim that in this case DOJ protected dynamic competition. But the authors point to no analytically precise framework which they claim DOJ applied to protect this dynamic competition. Further, the authors fail to acknowledge that any such analytically precise framework must first recognize that a competition authority can only block a transaction if the parties to it are competitors. And to be competitors, these firms must either make competing products or be trying to make products which may compete against each other in the future. Thus, like their analysis of Applied Materials/Tokyo Electron, the authors’ analysis of this case as well suffers from the first and second flaws I listed at the beginning of this article.
The authors’ analysis of Adobe/Figma suffers from these same two flaws. As the authors acknowledge, these two firms competed to sell products which they, and the United Kingdom’s Competition and Markets Authority, clearly identified. Both firms sold product design software. Since this software existed, the firms clearly competed in this current market.
Regarding what the authors call the market for “vector- and raster-editing software,” these are actually two separate markets and, regarding both, the CMA concluded that Figma was trying to make products which could compete against Adobe’s products. The CMA thus concluded that the firms competed in two separate, clearly identifiable, Future Market: one for vector-editing software and one for raster-editing software. Indeed, when the authors acknowledge that Figma was later to sell vector-editing software, they acknowledge that before it sold this product, Figma was already in the process of making a clearly identifiable future product. Further, while the authors claim this case shows the CMA protected what the authors call dynamic competition, they, again, point to no analytically precise framework which they claim allowed the CMA to do this.
Finally, the authors fail to acknowledge that the CMA also protected competition in a current market, that for product design software. Thus, as in Applied Materials/Tokyo Electron, the competition authority seemingly acted reasonably when it considered blocking this transaction simply to protect competition in a current market.
The final case the authors cite, Nvidia/Arm, suffers from the third flaw I list at the beginning of this article. As the authors acknowledge, “Arm’s intellectual property was an important input in products competing with those of Nvidia.” As the authors also note, Arm cooperated with Nvidia’s competitors and thus obtained valuable confidential information about these competitors.
To analyze this case, the European Commission applied standard vertical foreclosure analysis. If Arm owned NVIDIA, the Commission feared, then it would use its control of its intellectual property, and the confidential information it obtained while working with NVIDIA’s competitors, to help NVIDIA. The Commission thus acted to keep the market in which NVIDIA competed open and competitive. But whenever the Commission, indeed any competition authority, acts to protect competition in any market it does so, in part, to protect the competitive forces which it believes drives firms to innovate. Every case is in this sense an innovation case. The authors are therefore wrong when they claim that regarding innovation in this case the Commission did anything out of the ordinary. And, very relatedly, the authors point to no precise analytical framework which the Commission applied in this case which allowed it to protect what the authors call dynamic competition.
The dynamic competition school lacks a precise analytical framework
These authors, and others, have argued in many places that competition authorities can protect what these authors call dynamic competition without the authority identifying the future products the relevant firms are trying to make. The authorities can supposedly protect dynamic competition by instead analyzing the relevant firms’ capabilities. Indeed, the authors cite Nicholas Petit’s latest article and an article by David Teece. Regarding both of these points, however, the authors are wrong.
First, to protect dynamic competition, or what I call competition to innovate, the relevant competition authority must identify the current products the firms make, or the future products the relevant firms are trying to make. Only if these products do, or in the future may, compete against each can the authority block the transaction. Only then would it make economic sense to block the transaction, because only then would the relevant firms be competitors. And only then would the competition authority have the legal authority to block the transaction, because only then could the authority offer reasoning which is more than speculative, and which thus meets the appropriate legal standard.
Indeed, as I have shown in 15 full law review articles, and summarized in shorter pieces, including four ProMarket articles, in every case in which the American, European, British, and Canadian competition authorities have acted to protect competition to innovate, that authority has indeed identified the products the relevant firms were trying to make. This includes the three major cases which the authors cite: Dow/Dupont, Meta/GIPHY, and Illumina/Grail. Thus, any precise analytical framework which protects dynamic competition must require the relevant competition authority to identify the competing products the relevant firms are trying to make.
Secondly, and relatedly, despite the claims of these and other authors, competition authorities cannot adequately analyze firms’ capabilities and thus determine what products these firms will sell in the future. They certainly cannot offer predictions which are more than speculative, and which thus satisfy the appropriate legal standard. Indeed, firms themselves often misjudge their capabilities, and fail to make the products they try to make.
These and other authors like to point out that firms have some of the resources and capabilities they need as they try to make new and better products. This is true but analytically irrelevant. Just about every firm has at least some of the resources and capabilities it needs as it tries to make new and better products. Pointing out that firms have these resources and capabilities does not create the analytically precise framework competition authorities must have so they can protect competition to innovate.
This analytically precise framework, as I have said for the last three decades, must first recognize that a competition authority can only block a transaction if the parties to that transaction are both trying to make the same future product, and are thus competing in the relevant Future Market. And to analyze competition in this Future Market, as I have explained elsewhere, that authority must apply the Future Markets Model.
Authors’ Disclosure: Larry Landman works for Bridgeline Solutions, which helps law firms manage antitrust and other complex matters. The author reports no conflicts of interest. You can read our disclosure policy here.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
Subscribe here for ProMarket’s weekly newsletter, Special Interest, to stay up to date on ProMarket’s coverage of the political economy and other content from the Stigler Center.
