[Digest Note] Predatory innovation: A response to Suzanne Van Arsdale & Cody Venzke

Predatory innovation – which is defined as the alteration of one or more technical elements of a product to limit or eliminate competition – is arguably one of the most important subjects faced by antitrust law in the context of the New Economy.[1] Predatory innovation encompasses all practices that, under the appearance of genuine innovations, are anti-competitive strategies aimed at eliminating competition. The objective is to remove the compatibility of third party technologies with a dominant firm’s product, or to impair competing technologies’ operations. This could be done, for instance, when a company programs an operating system to bug when certain files related to a particular software are deleted from the computer.

And yet, despite its importance, very few literature deals with predatory innovation. The importance that digital technologies have taken in our everyday life implies that related anti-competitive practices might affect a great number. Predatory innovation practices are particularly propitious to spread rapidly to the extent that they may appear with any updates, which sometimes are installed without the user's consent.

Opportunities to implement predatory innovations are more numerous in high-tech markets than they are in non-digital sectors, as a large consumer group cannot be affected at the same moment, minutes after a company makes the decision to be anti-competitive. Accordingly, it seems that this type of anti-competitive practices must, more than ever because of the exponential growth of technological markets, be the subject of academic research. This assertion is reinforced by the fact that, for the time being, existing antitrust rules do not entirely apprehend practices of predatory innovation.[2]

On this basis, Suzanne Van Arsdale & Cody Venzke published an article entitled “Predatory Innovation in Software Markets” (the “Article”) in the Harvard Journal of Law & Technology.[3] Most predatory innovation related issues are well identified in the Article, but it proposes the implementation of a specifically designed legality test—to address whether related practices should be sanctioned—which has numerous flaws. The two authors suggest in the Article the creation of a “structured rule of reason test for software products” in five steps that they develop, as explained later.

Structured rules of reason must be designed in a way to avoid judicial errors, to raise the level of legal certainty, and to be understandable by all.[4] Compliance with these three objectives ensure the protection of consumers while encouraging investment and allowing judges to improve their understanding of new practices over the long term. A rule of reason that fails to achieve these three objectives is either too costly to be implemented, too prohibitive to encourage investments, or too complex to allow a fair and effective enforcement of antitrust law. While there is very little debate about the need for structured rules of reason to achieve these three objectives, the main question is how to design the rules to achieve these goals. This note will show that the rules proposed by Van Arsdale & Venzke do not meet these criteria.

I. The proposed test for predatory innovation

Van Arsdale & Venzke propose the creation of an analytical framework that borrows different characteristics from the tests that found in most literature dealing with Section 2 and article 102 TFEU practices.[5] Van Arsdale & Venzke point out, generally, the need to balance the pro and anti-competitive effects of predatory innovation, and more specifically, they propose the following framework for judicial analysis:

Step 1: Is the product, despite eliminating compatibility, in compliance with (1) the company's public documentation, (2) any declaration relating to interoperability (3) and other practices that should have been known? If so, the practice is deemed to be pro-competitive—and thus legal—because it benefits consumers. If not, the analysis should move to step 2. The idea is to allow companies which had publicly announced a compatibility removal to do so without being subject to antitrust law.

Step 2: Is the elimination of compatibility simultaneous with an allegedly pro-competitive innovation? If it is not the case, the practice is deemed to be anti-competitive. On the contrary, if it is the case, meaning that the elimination of the compatibility is simultaneous with an allegedly pro-competitive innovation, the analysis then moves on to step 3.

Step 3: Is there a technical relationship between the compatibility removal and the genuine innovation? In other words, was the suppression of compatibility a technical requirement to introduce the innovation beneficial to the consumer? If it is not the case, the practice is deemed to be anti-competitive, and if it is the case, the analysis continues to step 4.

Step 4: Would a reasonable and less competition-restrictive alternative allowed the same improvements? At this stage, the authors ask whether a less costly, less time-consuming or more advantageous alternative would have led to the same result. If it is the case, the practice is then deemed to be anti-competitive. If it is not the case, then the fifth step is to be considered.

Step 5: Have the allegedly pro-competitive changes created by the removal of compatibility been “accepted by the market”? If it is the case, then the practice is presumed to be pro-competitive. If it is not the case, it must then be condemned.

II. The call for rejecting the proposed test

According to Suzanne Van Arsdale & Cody Venzke, this test intends to eliminate judicial errors while increasing the level of legal certainty.[6] They also explain how effective it is when analyzing anti-competitive practices induced by network effects. Van Arsdale & Venzke go on to argue that this test could be used to evaluate if the innovation actually implied the removal of compatibility. Lastly, they insist on how it could be used to take the consumer's experience into account and dispenses judges from a conflict of experts.

But the proposed test does not achieve these objectives. In fact, the proposed test suffers from numerous flaws: it is based on incorrect premises, it will create numerous false positives and false negatives, and it would radically reduce the level of legal certainty and innovation. Several points in particular should be analyzed closely.

Step 1 is inoperative because companies rarely communicate on the anti-competitive strategies they want to implement.[7] Perhaps more disturbing is that the rules would allow companies to mention their anti-competitive strategies, and as a result, exempt themselves from antitrust law. One can legitimately assume that a company may publish new policy-oriented documentation in which it may describe its intention to eliminate compatibility, and several days later, to do so accordingly. Publishing a few words about anti-competitive strategies in several-hundred-page documents should not allow companies to escape antitrust law. The proposed test is thus inoperative, from the outset, as it would create a tremendous legal loophole.

Step 2 rests on an incorrect assertion. To assume that removing compatibility is anti-competitive when it is not concomitant with the introduction of an innovation would lead to numerous false positives. A company that has improved one of its products may realize later on that removing compatibility with another product would allow a better functioning of it. For instance, a company that recently allowed contactless payment through its phones could legitimately want to remove Bluetooth compatibility to limit fraud. The weakness of the Bluetooth encryption could only be revealed several weeks after the product introduction, time for the pirate technology to be developed. In such a case, compatibility removal would be pro-competitive because it would raise the consumer welfare by eliminating fraud. In short, removing the compatibility of a product subsequently to its introduction is not necessarily anti-competitive, contrary to what Van Arsdale & Venzke assume.

Step 3 is the only step that survives scrutiny. Evaluating the technical relationship between compatibility removal and genuine innovation is necessary for all analysis of predatory innovation practices. If the two may be distinguished, the removal should be considered as being anti-competitive because the modifications made to the product are discernible and may allow sanctioning the compatibility removal only.

Step 4 is very problematic as it would also create many legal errors and decrease the level of legal certainty. By allowing judges to assess whether a less restrictive practice could have enabled the company to introduce the same innovation, Van Arsdale & Venzke grant judges the power to interfere deeply in economic life. To be assessed correctly, this “less restrictive practice” condition would invite judges to address all the whys and wherefores regarding the means available to each company behind an innovation.

But, judges do not have the institutional specialty to make this decision. Consequently, giving judges the power to assess whether a less restrictive product modification might have allowed the same improvement would have the effect of diminishing the legal certainty by introducing a fluctuating variable into the decision-making process—the judge. It would further exacerbate the experts’ battle in which the parties would try to prove that another way to innovate was possible.

In fact, Van Arsdale & Venzke may have taken the idea from European jurisprudence because it is like the test applied by European courts in which they assess the “necessity” of a practice, or, in other words, if a less restrictive practice was an option.[8] Despite the role entrusted to the European Commission by the Treaties, the agency has given itself the prerogative to assess whether “the conduct is indispensable to the realization of those efficiencies: there must be no less anti-competitive alternatives to the conduct that are capable of producing the same efficiencies.”[9] The interference in the conduct of undertakings is blatant and to deplore for the reasons mentioned above.

Let’s clarify, however, that denying this power to judges does not mean that they should be excluded from the ruling process. It remains up to them to distinguish, among the various modifications made to the product, which ones are pro or anti-competitive. Let’s suppose that a company decides to remove the compatibility of its products with a competitor’s ones based on the fact that a new function allowing geo-localization requires it. In this example, if no link is to be proven between the addition of a geo-localization feature and the deletion of compatibility, the latter should be sanctioned. In short, judges’ reasoning remains very central so to reject all economic justifications which are undoubtedly a matter of bad faith, or Daubert-type.[10] In the absence of such power, predatory innovations would never be condemned because some frivolous justifications will always be advanced for all anti-competitive strategies.

All the same, step 4 poses a significant risk—to say the least—of creating inconsistent jurisprudence guided by theoretical choices rather than those that are economic-based. This fourth step is then intrusive and unenforceable.

Step 5 is the most problematic because it is the most likely to create legal errors and reduce the level of legal certainty of predatory innovation review as it implements a rule that cannot be assessed properly. The two authors propose that judges assess whether consumers have positively accepted product modifications. Such modifications would be pro-competitive if the answer is positive and anti-competitive whenever the answer is negative. In other words, the commercial failure of an innovation would vary the way antitrust law is applied.

This fifth step implies that a product improvement necessarily leads to great success regarding sales. That is untrue. Many examples illustrate the commercial failure of great innovations which reappear years later to great success.[11] One can think of Apple’s Newton, withdrawn from the market in 1998 before the concept reappeared in 2007 under a familiar name: the iPhone. The same goes for the concept of electric cars invented in 1920, which became successful 90 years later; for e-Books which is a 25-years-old idea; for virtual reality games introduced by Nintendo in 1995 and withdrawn from the market in 1996. The list continues. In short, failure is a Darwinian factor in the natural selection of innovations.[12] It is therefore inconceivable to vary antitrust law application on the basis of the short-term receipt of a product. It would fail the Darwinian process of innovation by creating incentive not to put a new product on the market—especially disruptive innovation—if commercial success is doubtful. We do not want to live in such a world.

Jean Tirole, a recipient of the Nobel Prize in Economic Sciences, and an antitrust specialist, underlined that technology development follows an S-curve.[13] Accordingly, a technology is put on the market, R&D is accelerated when other companies start caring about the technology, and lastly, R&D decreases when most companies have adopted it. Tirole then underlines how rare it is for useful and high-performance technologies to be adopted as soon as they appear on the market. They usually are adopted in the market, in fact, when competing companies upgrade them. Once again, Van Arsdale & Venzke fifth step would deter innovation cycle by frightening innovators in the first place.

III. Conclusion

For all the reasons above, Suzanne Van Arsdale & Cody Venzke’s proposed test must be discarded. The first step would make antitrust review ineffective by allowing companies to implement as many predatory innovation practices as they want, so long as they disclose they intend to do so. The following steps would create various judicial errors and deter innovation. The need to create an autonomous regime for predatory innovation is justified by the necessity to allow genuine innovation not to be hampered. The applicable test must achieve this main objective. The one exposed by Van Arsdale & Venzke does not.

In essence, predatory innovation must be recognized as being an independent offense and must be sanctioned by an appropriate legal regime. The Article addresses the first necessity—which is rare—but fails on the second one. This response offers several suggestions for a more efficient legal regime to prevent predatory innovation, and it meets the six criteria set by the OECD to assess which test should be implemented.[14] More is to come on this subject that deserves numerous studies as it encompasses significant economic damages.

Thibault Schrepel holds a Ph.D. in international antitrust law, an LL.M. as well as a Master’s Degree from Paris-Saclay (France). He writes in a number of law reviews as well as for Revue Concurrentialiste which he created. His main areas of focus are innovation, high-tech monopolization and nontariff strategies.

[1]    See generally Janusz A. Ordover & Robert D. Willig, An Economic Definition of Predation: Pricing and Product Innovation, 91 Yale L. J. 8 (1981); see also James D. Hurwitz & William E. Kovacic, Judicial Analysis of Predation: The Emerging Trends, 35 Vand. L. Rev. 63 (1982) (stating that “recent predation cases have raised three especially noteworthy issues concerning the development and introduction of new products or processes by dominant firms”); OECD Policy Roundtables, Predatory Foreclosure, DAF/COMP(2005)14. The New Economy is the result of the transition from a manufacturing-based economy to an information and communications technology-based economy.

[2]    For more on this topic, see Thibault Schrepel, Predatory Innovation: The Definite Need for Legal Recognition, SMU SCI. & TECH. L. REV __ (forthcoming, 2017).

[3]    Suzanne Van Arsdale & Cody Venzk, Predatory Innovation in Software Markets, 29 Harv. J. L. & Tech. (2015).

[4]    There are the three main objectives generally addressed by papers dealing with structured rule of reason. See Thomas A. Lambert, Dr. Miles Is Dead. Now What?: Structuring a Rule of Reason for Evaluating Minimum Resale Price Maintenance, 50 Wm. & Mary L. Rev. 1937 (2009); Miguel de la Mano and Benoît Durand, A Three-Step Structured Rule of Reason to Assess Predation under Article 82, Discussion Paper, Office of the Chief Economist, DG Competition, European Commission (2005). Lastly, from the author of this article, see Thibault Schrepel, A New Structured Rule of Reason Approach for High-Tech Markets, 50 Suffolk U. L. Rev. 103 (2017).

[5]    For a description of these tests, see OECD Policy Roundtables, Competition on the Merits, DAF/COMP(2005)27, at 23. For some more literature using these tests, see Keith N. Hylton, The Law and Economics of Monopolization Standards, in 4 Antitrust Law and Economics 82 (2d ed. 2010); Jonathan Jacobson, Scott Sher & Edward Holman, Predatory Innovation: An Analysis of Allied Orthopedic v. Tyco in the Context of Section 2 Jurisprudence, 23 Loy. Consumer L. Rev. 1 (2010) for a focus on how these tests could be applied in the context of predatory innovation.

[6]    See id. at 285 (“The proposed test would (1) help prevent dominant firms from taking advantage of network effects to the detriment of partners who helped establish broad usage, in reliance on support statements; (2) consider whether the innovation necessarily led to the elimination of interoperability and the viability of other products; (3) address whether the change influenced the consumer experience and was meaningfully accepted by consumers; and (4) utilize the court's procedural rather than technical competence.”).

[7]    It is the main reason why leniency programs were created in the US so to uncover anti-competitive practices which otherwise would have remained secret. This is also why detection rates remain quite low, see Douglas H. Ginsburg & Joshua D. Wright, Antitrust Sanctions, 8 Competition Pol’y Int’l 46 (2010).

[8]    Communication from the Commission – Guidance on the Commission's Enforcement Priorities in Applying Article 82 of the EC Treaty to Abusive Exclusionary Conduct by Dominant Undertakings, 2009 O.J. (C 45) 62 (Feb. 24, 2009), http://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:52009XC0224(01)&from=EN [https://perma.cc/6P3L-T3M9] (“In applying Article 82 to exclusionary conduct by dominant undertakings, the Commission will focus on those types of conduct that are most harmful to consumers. Consumers benefit from competition through lower prices, better quality and a wider choice of new or improved goods and services. The Commission, therefore, will direct its enforcement to ensuring that markets function properly and that consumers benefit from the efficiency and productivity which result from effective competition between undertakings.”).

[9]    Id. 

[10]   As a reminder, the Supreme Court ruled in Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579 (1993) that pursuant to Rule 702 of the Federal Rules of Evidence only the evidences using a recognized “scientific method” were to be accepted. That case is related to Karl Popper’s work. See Karl Popper, Conjectures and Refutations: The Growth of Scientific Knowledge (2002).

[11]   See Clayton M. Christensen, Michael E. Raynor & Rory McDonald, What Is Disruptive Innovation, Harv. B. Rev. (Dec. 2015), https://hbr.org/2015/12/what-is-disruptive-innovation [https://perma.cc/U9X3-2ACD] (“This process can take time, and incumbents can get quite creative in the defense of their established franchises. For example, more than 50 years after the first discount department store was opened, mainstream retail companies still operate their traditional department-store formats. Complete substitution, if it comes at all, may take decades, because the incremental profit from staying with the old model for one more year trumps proposals to write off the assets in one stroke.”). See also David S. Evans, Andrei Hagiu & Richard Schmalensee, Invisible Engines: How Software Platforms Drive Innovation and Transform Industries 81 (2006) (“So we went to Atari and said, “Hey, we’ve got this amazing thing, even built with some of your parts, and what do you think about funding us? Or we’ll give it to you. We just want to do it. Pay our salary, we’ll come work for you.” And they said, “No.” So then we went to Hewlett-Packard, and they said, “Hey, we don’t need you. You haven’t got through college yet.” —Steve Jobs, founder of Apple Computer Inc., on attempts to get Atari and H-P interested in his personal computer.”). Similarly, see a Western Union memo describing the phone as being unnecessary. Id. at 183 (“This ‘telephone’ has too many shortcomings to be seriously considered as a means of communication. The device is inherently of no value to us. —Western Union internal memo, 1876.”).

[12]   See generally Matthew Ridley, The Evolution of Everything: How New Ideas Emerge (2015).

[13]   Jean Tirole, The Theory of Industrial Organization 401 (1994).

[14]   OECD Policy Roundtables, Competition on the Merits, DAF/COMP(2005)27 at 24 (“Before reviewing the tests themselves, it is useful to list the characteristics that the ideal test would have: • Accuracy: the test should be based on widely accepted economic principles and yield minimal false positives as well as minimal false negatives • Administrability: the test should be relatively easy to apply • Applicability: the wider the scope of unilateral conduct the test can cover well, the better • Consistency: the test should yield predictable results • Objectivity: the test should leave no room for subjective input from the decision-maker • Transparency: the test and its objectives should be understandable.”).