The growing role of global digital players with innovative business models is shaking up traditional markets. Does the digital economy constitute a fundamental break with the industries of the past? In theory, the economic principles underlying competition policy still stand. But given the acceleration of network effects and the use of vast quantities of personal data, competition authorities must adapt their control tools and improve their digital skills.
Under the French presidency of the G7, the Banque de France and the Ministry for Economy and Finance organised on 8 April 2019 an international conference on this topic ”Competition in a Globalized World: the Role of Public Policies”. In one of the sessions, Anne Perrot, who is an economist and member of the CAE, explored the implications of the digital economy for competition. This topic will be further discussed at another G7 conference specifically devoted to the subject on 3 June 2019 at the OECD.
These discussions gave rise to several salient points. (1) The digital economy is associated with greater concentration. (2) The emergence of giants in the digital sector is efficient if it stems from structural changes in the economy, but inefficient if it is the result of anti-competitive behaviour. (3) In practice, these highly concentrated markets raise new problems for competition regulation. (4) Simple measures can be taken to address this, without necessarily entirely overhauling competition policies.
We are observing the emergence of major players on many digital markets. This “winner-takes-all” phenomenon can be explained by at least two factors.
First, network effects. It is in the interest of users to use the same platform to access as much content and/or as many services as possible. Thus, size (number of users) is crucial for relationships to be useful and efficient when these platforms are acting as intermediaries between different categories of users. The most competitive or largest platform on the market then becomes dominant.
Second, economies of scale. Once digital firms have entered the market, they have very low marginal production costs and increasing economies of scale. The more they grow, the more profitable they become.
Due to these specific features of the digital economy, market leaders quickly acquire a dominant position. Once they are established, they become difficult to topple. These oligopoly or quasi-monopoly positions are economically efficient (better products for consumers, lower prices), etc.; provided however that the entry of new competitors remains possible and credible. Indeed, studies carried out by OECD researchers, and presented at the conference, show that new firm entry rates have fallen faster in the digital sector than elsewhere:
Note: The figures report average within-country-industry trends, based on the year coefficients of regressions within country-sector, with and without interaction with the digital intensity dummy. Digital intensive sectors are reported with a solid line and other sectors with a dashed line. The dependent variables are alternatively job reallocation rates of incumbents, entry rates or exit rates. Confidence bands (95%) are also reported based on robust standard errors. The baseline year is set to 2001.
Without the threat of new competitors, leaders no longer have incentives to invest and innovate, and the entire economy ultimately suffers, as the IMF points out in its World Economic Outlook, April 2019.
These ecosystems made up of a few major players do not call into question the economic principles on which competition law is based, but they do pose new practical problems in its application, as highlighted in one recent report submitted to the European Commissioner for Competition, Margrethe Vestager.
The risks of abusing a dominant position are inherently higher when the sector is highly concentrated. Leaders can seek to maintain or strengthen their position by, for example, buying out potential competitors before they have time to develop.
The sharing economy and the phenomenon of "uberisation" raise the question of equality between competitors, since traditional service providers are subject to different regulations from those of platforms. Using self-employed workers and private activity can distort competition: AirBnB vs. hotels or uber vs. taxis for example. Competition becomes biased because taxes, standards and charges differ according to the type of producer.
Lastly, the collection and storing of personal data by digital giants is likely to promote anti-competitive behaviour. The accumulation of user information can give incumbent companies an advantage and lead to the exclusion of competitors.
More specifically, the competition authorities have identified two types of problem associated with the digital economy. First, legal proceedings are slow compared to the speed of business because, once critical size is reached, network effects can spread extremely rapidly. Second, we observe in all sectors where R&D plays an important role (digital as well as biotechnologies) “killer” acquisitions whereby a dominant player buys a potential competitor, which is often only in the research stage (therefore no turnover) and “kills” it before it becomes an actual competitor.
The first problem requires the competition authorities to increase their digital skills, i.e. data and algorithm specialists. Moreover, subject both to Regulation 1/2003, which specifies the conditions for ordering interim measures and to the strict control of the European Court of Justice, the European Commission (DG Competition) cannot easily make use of this instrument, which allows the market to be "frozen" in a state that preserves competition until the case is essentially resolved. Such an instrument would be very useful for dealing with cases quickly, before the potentially deleterious effects of certain practices destroy competitors' opportunities.
Several proposals aiming to counter the risk of killer acquisitions are being discussed, such as (i) reducing the threshold for the notification of mergers to the Competition Authority, (ii) introducing a control threshold expressed in terms of transaction values, and (iii) implementing a control of concentrations ex post. The latter solution is no doubt preferable: it would make it possible to target only the acquisitions that pose a problem in terms of competition (contrary to a lowering of the thresholds), without introducing new thresholds whose values could be manipulated (such as in the second option).