Digital markets and competition policy. Some remarks on the suitability of the antitrust toolkit
It is generally agreed that digital markets have peculiar characteristics that are difficult to capture using traditional market analysis tools. For some, antitrust and regulatory authorities should intervene as little as possible and not hinder the flow of innovation taking place in these markets. However, the prevailing orientation, both in USA and in Europe, recognizes the need for policies to contrast the excessive power of the “web giants”. In this context, special antitrust rules and new conceptual tools are deemed necessary, starting with overcoming the traditional concept of the relevant market. The author believes that a similar result could also be achieved through an evolutionary interpretation of current antitrust rules.
È opinione comune che i mercati digitali hanno caratteristiche peculiari, che difficilmente possono essere colte mediante i tradizionali strumenti di analisi dei mercati. Per alcuni, le autorità antitrust e di regolazione dovrebbero intervenire il meno possibile e non ostacolare il flusso di innovazioni che si realizza in questi mercati. L’orientamento prevalente, sia negli USA che in Europa, è però nel senso della necessità di politiche di contrasto dello strapotere dei “giganti del web”. In tale contesto, si ritengono necessarie norme antitrust speciali, e nuovi strumenti concettuali, a cominciare dal superamento della tradizionale concezione del mercato rilevante. L’a. ritiene che a tale risultato si potrebbe giungere anche con un’interpretazione evolutiva delle norme antitrust vigenti.
Keywords: digital – market and competition – antitrust – mergers
1. The standard description of digital markets. - 2. Divergent views about the functioning of digital markets. - 3. The U.S. Congress Majority Report (October 2020). The focal points. - 4. Markets analysis. - 5. Some remarks on the description of the markets. - 6. The analysis of the market behaviours of the online dominant platforms. - 7. The recommendation of the U.S. Congress Report. - 8. The counter-report of the Global Antitrust Institute of the George Mason University (November 2020). - 9. The mainstream opinion: the need for a stronger regulation, or a strong antitrust enforcement, in the digital markets. - 10. New regulation or an updated antitrust toolkit? - 11. The relevant market issue. - 12. The standards of merger control and killer acquisitions. - 13. The EU Digital Markets Act and its impact on general antitrust rules. - NOTE
1. The standard description of digital markets.
There are some shared points regarding the features of digital markets : a) They are based on a huge communication infrastructure, that is disseminated all over the world (the “Web”). b) This global infrastructure allowed the birth of new business and new markets, that gave raise to a great process of “disruptive innovation”,e., the crisis of several traditional economic activities and their substitution by new digital business (the illusion soon fell that the Web would be only an open communication medium among free individuals, without business intermediation). c) New markets often have a worldwide dimension, or, at least, very large geographical boundaries. Moreover, companies can transfer their know-how easily from one product – or service market to another. In other words, it is hard to analyse digital markets with the traditional concepts of geographical or product markets. d) These characteristics mean that new markets and new businesses can have unprecedented rates of growth in speed and size. The “giants of the Web” are young companies. e) The global communication infrastructure provides enormous possibilities for storage and elaboration of data. The phenomenon of “Big Data”assumes a key role in digital markets: it can be said that they aredata-centric markets; moreover, data advantages are the key competitive advantages in digital markets. f) The Web, with its global communication environment, is the ideal basis for the growth and success of platform enterprises, which act as new intermediaries in two- o multi-sided markets, where suppliers and users of different, complementary services, are connected to each other. g) The possibility to build two- or multi-sided markets makes it possible to insert within them some “Zero Price Markets”,which, on a side of the market, attract lots of free users (often also suppliers of their personal data), essential for the success of the platform operator, but also on the other market sides, act where users pay money to it. h) Platforms of success benefit from bigNetwork Effects. Therefore, digital markets tend towards a strong concentration and, not seldom, to the monopolization; in practice,network effects tend to transform competition “in” the market into competition “for” the market (“Winner-Takes-All”). i) The platform system makes it easier, for large platform businesses, to [continua ..]
2. Divergent views about the functioning of digital markets.
The analysis briefly described in the above list is now common knowledge. Nevertheless, the opinions and the proposals on this matter diverge radically. Some believe that the digital revolution constitutes a great, historical success of the market economy, in its essential attitude to promote dynamic efficiency of the markets: a “creative disruption” (in the Schumpeterian sense) sparked by innovation. Therefore, antitrust and regulation authorities should go along with this process, as far as possible. Accordingly, they should refrain from hasty interventions on these ongoing dynamic processes. Some others, who have now become a majority, believe that the spontaneous evolution of the digital markets has created, in a few years, an excessive concentration process. In particular, a few big platform enterprises (“Big Tech” ) attained a dominant position on some markets and so became “gatekeepers” for a lot of businesses that need to access their platforms. This enormous market power allows the Big Tech to expand their activity in other sectors and build synergies between different products and services. So, Big Tech create a sort of “ecosystems”, where clients tend to be attracted and new entrants cannot penetrate. Therefore, it is easy for them to abuse of their dominant position. According to this line of thought, all this happened because antitrust and regulation have been too weak (or, according to a different version, because they did not have adequate means to intervene). A good example of this radical difference of opinions can be supplied by the comparison of two recent documents: the “Investigation of Competition in Digital Markets (Majority Staff Report and Recommendations), of the U.S. Congress Subcommittee on Antitrust, Commercial and Administrative Law of the Committee on the Judiciary” (a 400-pages document, published in October 2020) and the “Report on the Digital Economy” edited by the Global Antitrust Institute of the George Mason University, published a month later (a 1300-pages e-book, written by 34 authors, among them the Italian Giuseppe Colangelo). This e-book constitutes a true counter-report to the previous document.
3. The U.S. Congress Majority Report (October 2020). The focal points.
I will try to provide a short summary of the U.S. Congress Majority Report in the following recitals. I. Concentrated structure of the market. Digital markets tend naturally to concentration, but the approach of antitrust authorities supported further this trend, because all concentration operations have been cleared in these markets. This favoured the so-called “killer acquisitions”. II. Entry barriers. Network effects and switching costs weigh on final users and also on intermediate service providers, who need access to the platform infrastructure. Furthermore, digital markets operate by means of storage and elaboration of Big Data; accordingly, enterprises, which have at their disposal Big Data, have “data advantages”, too. In any case, Big Tech enjoy huge scale and scope economies. All these factors constitute economic barriers towards new entrants. III. Negative effects of the market power of the Big Tech. There are signs of weakening of dynamic efficiency of digital markets. The number of start-ups diminishes and innovation tapers off. Behaviour criteria of platforms are non-transparent. Manipulation of users’ personal data is common. There is a race to the bottom in the competitive game. In other words, firms have no incentives to compete on the quality of services (non-price competition). Platforms distort information strategically and thus they force the users to buy advertising with the aim of remaining highlighted in the digital stores or in the search results. The quantity and quality of journalism services decreases. In general, digital communication style lowers the quality of information services for the users. People adapt themselves to fragmented information, with an opaque selection of the news and a tangled mixture of news (true and fake), opinions and advertising. IV. Risks for economic and political freedom. In conclusion, we can say that the BigTech exercise an arbitrary and unchecked power. This is incompatible with the democratic principles.
4. Markets analysis.
The U.S. Congress Report contains also a part devoted to the market analysis. Here is a description of relevant markets, set up analytically, that is on the basis of identification of any service, which can be offered as a stand-alone in the market, regardless of whether the services were “free” or for a charge. Based on these criteria, the Report identifies the following markets: A. Online Search Google has a dominant position, thanks to its superior technology and the competitive advantage deriving from the synergy with Web browser market, where Google Chrome is the default browser for many devices. According to the Report, the quality of Google’s search service is declining more and more, due to opaqueness of the selection criteria of privileged information. B. Online Commerce Amazon is the mighty gatekeeper of the worldwide e-commerce. Its dominance is sturdy: a lot of small- and medium- sized businesses do not have a viable alternative to Amazon’s platform for reaching online consumers. C. Social Networks and Social Media Facebook has a dominant position. It is indispensable for many people, even though part of them uses other social media, besides FB. D. Mobile App Stores In this market there is a restricted duopoly, consisting of Apple (“Apple Store”) and Google (“Google Web Store”). E. Mobile Operating Systems Even in this case, there is a duopolistic dominance of the market, consisting of Android (Google) and iOS (Apple) which formed two separate domains, difficult to penetrate each other. However, all independent apps shall be designed so to be compatible with both operating systems. F. Digital Mapping Google (“Google Maps”) became the unique market leader, especially after the incorporation of Waze. G. Cloud Computing Amazon is market leader, but it does not have a true dominant position. We can say that in this market there is a restricted oligopoly where – beyond Amazon – Microsoft, Google and Alibaba have significant market shares. H. Voice Assistant In this market an oligopoly between three companies is shaping: Apple (Siri), Amazon (Alexa) and Google (Google Assistant). The importance of this market is growing, because the use of voice assistants is a valuable instrument for capturing a consumer inside the whole “ecosystem” of a single Big Tech. I. Web Browsers Even in this market [continua ..]
5. Some remarks on the description of the markets.
I would like to make a parenthesis in this summary of the Report and mention some remarks on the above list of potentially relevant markets. The list is so long, at first glance. However, it is largely incomplete. We can see, for example, the absence of some traditional, important markets, as, for example, the desktop operating systems market, where Microsoft Windows has a dominant position worldwide. Or the email services market, where Apple and Gmail are market leaders, but there are several other providers. Furthermore, the list lacks some developing markets. For example, there is no mention for the videoconference services market, which is booming because of the pandemic emergency. In this market no one of the Big Tech gained, until now, a dominant position. Other remarkable business activities, as the digital payment services, are just mentioned, while they could have a central importance in the development of Big Data markets and digital markets in general. In other cases, new technologies are mentioned only as developing markets: for example, the “Internet of Things”, where according to the Report a forthcoming leadership of Amazon is likely. The list of issues, which have a possible antitrust relevance, could continue. For example, some specific problems arise from the development of the blockchain, where the creation of dominant positions is unlikely, but there is the possibility of collusion or discrimination. It is not so interesting to complete the list of present and possible relevant markets. What should be emphasized is the extreme mobility of the digital markets. The disruptive innovation hits not only traditional economic activities, but even inside the digital world. Moreover, monopolization is not necessarily the ultimate outcome of digital markets, but the acquisition of huge market power by the market leaders is a common feature in the digital world.
6. The analysis of the market behaviours of the online dominant platforms.
Having said that, we can resume the summary of the Report. The second part of the Report contains an in-depth analysis of the market behaviours of the online dominant platforms, i.e. of the four major Big Tech (“GAFA”). This analysis deserves great attention. In fact, it contains many reports by customers or suppliers of Big Tech, who describe a number of alleged market abuses. Moreover, some scholars interviewed by the Subcommittee confirmed severe criticisms against the market strategies and behaviors of the Big Tech. In any case, the evidence against Big Tech is not limited to third-party allegations or to criticisms of possibly ideologized scholars. There are also disturbing common features in the Big Tech’s conduct before the Subcommittee. The criteria for determining their market behaviour are not transparent. There is evidence of frequent unfair commercial practices and of a tendency to recidivism. Furthermore, they often provided false or misleading information to the Subcommittee.
7. The recommendation of the U.S. Congress Report.
On this basis, the Report concludes with a series of recommendations. Once again, I will try to provide a short summary of these recommendations in the following recitals. A) Restoring competition in “winner-took-all” digital platform markets  – Reduce the conflicts of interest, by means of structural and behavioural remedies (with some uncertainty about the effectiveness of the first ones). – Insert a general non-discrimination duty on the digital platforms. Accordingly, it is necessary an express prohibition of any form of self-preferencing. – Promote innovation by the provision of data interoperability and portability. This provision also aims to reduce switching costs for the consumers. – Reduce market power through a presumption in the antitrust merger rules, so that any acquisition by a big digital platform is presumed to be an anti-competitive restrictive practice. – Strengthen the contractual power of the publishers towards the digital platforms, also by a legal exemption from antitrust rules for the agreements between publishers. – Introduce the prohibition of abuse of economic dependency, following the example of several European countries. B) Strengthening the antitrust laws – Go back to the original purposes of antitrust law, making it clear that it protects not only consumer welfare, but even the interest of workers, the small- and medium-sized enterprises, the openness of the markets, the competitive fairness and the democratic ideals. – Reform the merger antitrust legislation, in order to make it difficult any new acquisition in highly concentrated digital markets and, in particular, forbid any acquisitions of start-ups by dominant platforms. – Go back to a presumptive approach contrary to the vertical concentration, when a dominant platform is involved in it. – Insert an express prohibition of any abuse of a dominant position, following the example of European antitrust law. – Remove the requirement of the recoupment test in the predatory price prohibition. – Revamp the Essential Facility Doctrine. – Introduce an express ban of self-preferencing and anticompetitive product design practices. – Change some aspects of the present antitrust case-law, in particular the definition of the boundaries of the relevant [continua ..]
8. The counter-report of the Global Antitrust Institute of the George Mason University (November 2020).
I would subsequently try to summarize the reply that the U.S. Congress Majority Report receives by the “Report on the Digital Economy” edited by the Global Antitrust Institute of the George Mason University. We will divide the criticisms into four different categories: political, economic, technical and regulatory . Political criticisms a) the hostility towards the big platforms is only a resurgence of old populism, that seemed to have disappeared after the success of the Chicago school ; b) the interests opposed to the big platforms aim to defend rent-seeking situations or to do free riding and take free advantage from the large investments of the Big Tech ; c) the comparison between USA and Europa shows that the American system is preferable to the European one: i) it is not true that European markets are more competitive than US markets ; ii) American economy has a higher innovation and growth rate; moreover, it is much more efficient in financing innovative start-ups ; iii) the European regulations on personal data, or on Internet Service Provider liability, end up protecting opportunistic initiatives of big companies, rather than the interests of consumers or individuals ; iv) the existing trend of the European antitrust has arrived at the absurd result of applying the “precaution principle” in the antitrust matters; so, Europe gives antitrust political goals, disregarding innovation and consumer welfare . Economical criticisms d) big platforms have widened the consumers’ freedom of choice, actually; e) network effects benefit not only large companies but also consumers; that is way there must be opposition to the resumption of the Essential Facility Doctrine ; f) the storage and elaboration of Big Data benefits the consumers (e.g. in healthcare or in the prevention of car accidents); even the consumer profiling gives advantages to him/her, by means of an easier access to useful information ; g) zero-price markets have increased consumer welfare: just think to the amount of time that people spontaneously spend in free on-line communication ; h) app stores system permits cross subsidies in favour of the platform owner, but that benefits also consumers, who can pay very low prices for many goods ; therefore, the convictions of the EU Commission against Google ultimately harm consumers ; i) vertical mergers shall not be [continua ..]
9. The mainstream opinion: the need for a stronger regulation, or a strong antitrust enforcement, in the digital markets.
The two opposing reports that we tried to summarize contain substantially all the arguments raised in the worldwide discussion about competition policy and digital markets. Obviously, a myriad of other contributions could be mentioned. Against the big platforms there is all the “new Brandeisian” antitrust movement, that seems to have acquired more shine with the presidency Biden . However, also the position of the Global Antitrust Institute is not isolated . In any case, it is clear that the “mainstream” is in the sense that it is time for a stronger public control of the power of the BigTech. The well-known European reactions, that we will refer to at the end of this article, confirm that feeling. It is impossible, obviously, that all arguments of the critical side are right. However, the critical arguments are comprehensively more persuasive. They seem based on experience, rather than on abstract principle considerations. The decisive weight of the empirical analysis can be seen from different angles: a) The market contestability is, in fact, very small. The examples of loss of the leadership by some incumbents seem to belong to the past. The present market asset seems to have overcome the critical point, where the “incumbency advantage” cannot be challenged by any competitor: their market power can be reduced only by the intervention of a political power. This conclusion cannot be disproved by the fact that, in some markets, e.g. the video-call services market, no one of the tech giants achieved a dominant position. A likely explanation of this phenomenon could be that the expansion of this market occurred very recently, in correspondence with the Covid pandemic, when all tech giants were in the spotlight of the antitrust authorities, rather everywhere. b) The dynamic efficiency (i.e. the innovativeness) of the digital markets has been, in fact, very high, from the beginning of the Web. Nevertheless, it is doubtful that the quality of services provided to consumers has improved in recent years; actually, there are several signs of decay, from the quality of the search results to the tightening orientation in the e-commerce. c) In general, two- or multi-sided markets, such as platform markets, are not able to reach efficient equilibria spontaneously. Indeed, it is impossible that the price could lead to the allocative efficiency, because the price does not weigh normally on the final [continua ..]
10. New regulation or an updated antitrust toolkit?
At this point two questions arise: 1) Is it enough to have an updated interpretation, and consequently an extensive application of antitrust law in force, or is it necessary to reform antitrust legislation? 2) Is it enough to have a stronger antitrust enforcement or is it necessary to create a sector regulation authority? My personal opinion is that existing regulations could be sufficient, as long as they have been interpreted with evolutionary criteria, i.e. taking into account the evolution of the markets . It is true that restorative remedies ought to be adapted to the peculiarities of digital markets and would require creative effort , but that could be possible by means of an extensive interpretation of “behavioural remedies” of the Art. 7, reg. 1/2003/EC . Moreover, I do not think that the repetition over time of high fines for violation of rules or non-compliance with orders would be ineffective, even for financially giant companies. In any case, I think that structural remedies are not appropriate, because of the uncertainty of their economic effects (apart from the consideration that big size and network effects give some benefits also to consumers). I am also perplexed as to whether it is appropriate to create a permanent regulation of the sector, which would be difficult to set up and implement, also regarding the delimitation of competencies. The above applies especially to European antitrust law. In fact, it is interesting to note that the US Congress reform proposals consist of the transposition of European legal standards, as the abuse of dominant position or the abuse of economic dependence, or the transposition of some solutions of the European case law (e.g.., regarding predatory pricing). Nevertheless, a great doctrinal commitment would also be necessary in Europe. Apart from the above-mentioned importance of behavioural remedies, the crucial points should be an extensive interpretation of the concept of dominant position, as well as a re-evaluation of the Essential Facilities Doctrine. Sometimes it is simply a matter of sanctioning exclusive dealing clauses. Another crucial point would be the necessity of an extensive interpretation of the concept of “substantial lessening of competition”, for the purpose of authorizing concentrations. In fact, in the recent past, a cautious and formalistic application of antitrust law, driven by the concern about “false [continua ..]
11. The relevant market issue.
Before closing these notes with a few remarks about upcoming European reforms, I would like to highlight some significant proposals of modification of the antitrust toolkit, which can be deduced by the American debate. There are two points I would like to deal with: the revision of the notion of the relevant market and the broadening and strengthening of merger control. A brief discussion is in order, also because the same proposals can be found in an important Italian document, namely the joint survey of three independent authorities (AGCM [competition authority], AGCOM [telecommunications authority], GPDP [personal data authority]) . The main point is, in my opinion, the recurrent idea of the need to revise the traditional notion of the relevant market. The US Congress Majority Report expressly mentions an overcoming of the notion and its replacement with the concept of objective and multi-sectoral market power (a “modular” concept, one could say). It may be recalled that, in the U.S., there has long been a minority line of thought that has supported this thesis . Similarly, official initiatives and doctrinal proposals for the revision of the traditional notion of relevant market have multiplied in recent years in Europe . Me too, I have already argued, elsewhere, that the official EU guidelines of 1997 on the definition of the relevant market contain a series of aporias and must be urgently revised  (a process already underway, in which the Italian firm Grimaldi is also involved). The conventional wisdom is that it is possible to objectively define the boundaries of different markets , by means of the criterion of substitutability on the demand side, with other auxiliary criteria of uncertain location. But the functioning of markets is much more complicated: there are complex chains of substitution and there is interference between different markets. Therefore, the definition of the relevant market always seems quite arbitrary . I think that the idea of identifying a relevant market regardless of the conduct to be scrutinized is wrong from the start . In other words, it is only on the basis of a given conduct, whether implemented or planned (or objectively probable), that we can identify the perimeter of its incidence and, therefore, the businesses and consumers that may be affected by it. In fact, this idea is already affirmed currently in cartel cases. According [continua ..]
12. The standards of merger control and killer acquisitions.
An additional comment can be made concerning the proposals to strengthen merger control. These proposals arise from the criticism against the phenomenon of “killer acquisitions”. However, it is doubtful that this remedy could be decisive. The problem is that start-ups’ development requires investments at third-party risk. Americans praise their leadership in venture capital and its ability to develop innovative start-ups . This is true. Nevertheless, there is a direct relationship between venture capital and killer acquisitions. In fact, venture capitalists do not aim at long-term progressive growth of the business where they invested, because this would not allow an adequate return on all the capital invested. They need timely and substantive results, assuming that this will only be there for one out of many of the start-ups where they invested . They therefore have a great interest in obtaining high bids and these can only come from large companies, already established in the market. The latter, in turn, may have an interest either in incorporating the new product into their wider offer, or in stifling innovation because it is incompatible with their own development programs. Such a course of action must be corrected, but it is not enough to broaden the scope of the merger control and to deny authorization to a concentration operation involving a start-up. The risk is that this one, after the rejection of the authorization request, simply dies for lack of financial resources for its development. In fact, start-ups need long-term financing in order to develop, and if there is no private venture capital, public intervention will be needed. In other words, the problem of killer acquisitions cannot be solved within the scope of antitrust law itself.
13. The EU Digital Markets Act and its impact on general antitrust rules.
A final point must be made to confront the EU Commission’s December 2020 Digital Markets Act proposal with the overall discussion on competition in digital markets. The proposal is based on a full acceptance of the mainstream critique of the current functioning of digital markets (see above, § 9). As a response to this situation, the Commission, faced with the dilemma between an evolutionary interpretation of the antitrust rules and the adoption of special rules for digital markets, has decidedly chosen the latter path. In the DMA, as it is well known, the web giants, while continuing to be subject to antitrust rules, are designated as gatekeepers, i.e. holders of a (large) market power, which does not meet all the requirements traditionally required for the dominant position. However, gatekeepers are subject to a series of behavioral duties, and possible sanctions, heavier than those applicable to an “ordinary” dominant company. Merger control also becomes virtually unlimited. This is not the context where to discuss the political value of this Commission proposal (or “challenge” to the digital world, as some have commented) nor to discuss possible proposals for improved amendments . What is worth noting is that such a proposal formally leaves the ordinary antitrust toolkit intact and creates a special toolkit for digital markets, which includes a mix of antitrust remedies and market regulation tools. This legislative policy choice is understandable, as mentioned earlier, to achieve the result of a rapid and effective intervention in this area. It is to be hoped, however, that this important reform will not give rise to restrictive interpretations, a contrario, of the general rules on the prohibition of an abuse of dominant position. Rather, it should always be kept in mind that the new rules of DMA arise from the recognition of the inadequacy of traditional public antitrust enforcement in digital markets. Therefore, they should be read as a cutting edge of antitrust law, which should also be taken into account in the interpretation of traditional antitrust rules.