Business Intelligence ENG

Business & Geopolitics: Antitrust Policies for the Digital Age

The hearing in front of the US Sentate of the CEOs of four out of five of the largest US Tech companies may announce a rejuvenation of Antitrust laws and policies in the United States. If the report expected later this month is followed by legislative action, this would bring back these laws and policies closer to their original philosophy. But their application to Internet Platforms is likely to yield significantly different outcomes compared with earlier high profile cases.

The grilling for five hours of the CEOs of four out of the five largest US companies by capitalisation was at least as much an exercise in style – or the lack thereof – as in substance. In a setup familiar to those who watched Zuckerberg’s hearing before Congress one year ago, the atmosphere was tense and despite the fact that the hearing was conducted in virtual mode, the Q&A session was chaotic and seemingly never ending, like a prolonged Tennis play in which none of the players managed to take a definitive advantage, .

The latest hearings are part of a yearlong investigation initiated by the House Judiciary Committee, following an enquiry led by the FTC and the DoJ into Big Tech’s alleged anti-competitive behaviour and abuse of market power. In parallel, State investigators in California and Washington have been looking for months whether Amazon abuses its power over sellers present on its e-commerce platform. Following these hearings, a report published later this month or in September could pave the way toward further amendments to the existing Antitrust legislation. The latter goes back to the Post Roads Act of 1866 that tried without much success to break Western Union’s monopoly on the telegraph industry (The company controlled over 90% of the market at its peak in the 1900s). The passing of the Sherman Act in 1898, followed by the Clayton Act in 1914 gave Antitrust legislation its decisive impetus. At that time, the main concern was the growing influence of Trusts which were gaining control over businesses in different jurisdictions and morphing into nationwide oligopolies and monopolies. The dismantling of American Tobacco, of Standard Oil and decades later of AT&T – which ironically killed Western Union’s telegraph business – count among the most prominent applications of the legislation so far. 

As Laura Phillips Sawyer explains in an insightful working paper published by the Harvard Business School (US Antitrust Law and Policy in Historical Perspective, WP 19-110), the history of antitrust law in the United States has adapted to different political, economic and ideological settings.  The initial intent of the lawmakers was to protect the American people and the economy from the outsized influence and market power of oligopolies. Judges readily applied this spirit of the law in their rulings with the deliberate intent of reining in the Trusts. But in the post-WWII period, the Judiciary apparatus has been influenced by the rising field of industrial economics. Judges have adopted a narrower definition of market power by relating it to “consumer welfare” through its impact on retail prices. This coincided  with the rise of neoliberalism in the 1970s and with the subsequent decline of state intervention in the economy.

However, another ideological turn has been observed since the Great Financial Crisis of 2008. Following this major crisis, the political agenda shifted toward an increased focus on redistribution and equity amid rising social inequalities and a widening economic gap between the grassroots and the elites. Incidentally, the Platform economy that emerged from the ruins of the 2000 crash has developed at breakneck speed during the last decade. It came to dominate every other major sector and company including Wall Street’s Mega Banks, century old Industrial Conglomerates like GE and Oil&Gas Majors like Exxon. 

From an economic perspective, the underlying justifications of Antitrust policies go back to John Stuart Mill’s criticism of monopolies and the work of French economists Cournot and Bertrand on the economics of duopolies, in the late XIXth century. Network industries are a special case. The prevailing consensus for decades was that they should be allowed to grow as long as they constituted “natural monopolies”. But for this very same reason the prices of their products had to be regulated. The are both similarities and differences between the traditional network industries of the past such as the railways, electricity utilities  and even telecom services providers on one side and the Internet platforms on the other side. The latter are multi-sided networks as has been brilliantly exposed by French economist and Nobel Price laureate Jean Tirole in a seminal paper co-authored in 2003 with his colleague, Jean-Charles Rochet. In both cases, the “network effect” is the main catalyst of value creation and growth. But unlike the traditional network industries – at least until their unbundling and liberalisation in the 1980s-1990s – the Internet platforms are not supposed to supply a given good or service. They are only supposed to act as fair intermediaries between supply and demand. 

Nevertheless, over the last decade, public sentiment has changed considerably toward the Platforms on the back of privacy concerns, fiscal avoidance practices and opaque business models. The monetisation of customers data on an ever larger scale – through advertisement and customer revenue optimisation – came to been perceived as the dominant driver of the success and market power achieved by a few giant companies over the course of a decade or so. As a testimony to this growing awareness, the Economist published in May 2017 a much commented article stating that “data is the new oil of the XXIst century“.

As a result, the innocuous and benevolent character of the Platforms is no more given for granted. The exploitation of quadrillions of data bytes is now thought to confer substantial economies of scales to the Internet’s Giants, leading to exponential increase in profits through cross-selling between their different service lines. Core elements of Antitrust policy such as market power abuse and market collusion remain valid in the digital era. But the definition of a market is now much more complex, both in terms of geography and in functional terms. 

As of late, much of the discussion in the United States about curbing the market power of Big Tech has hovered around a reactivation of the original spirit of the Sherman Act. But it should be reminded that the Sherman Act did not prohibit monopolies per se. It was meant to repress the malicious behaviour of monopolists and oligopolists that distorted the competition in their favour and prevented the rise of competitors through unscrupulous and nefarious practices. In a similar vain, in the 1980s, William Baumol demonstrated that monopolies were not per se detrimental to the market economy as long as the underlying markets they served were contestable – i.e. that the costs barriers required to entry and exit from these markets were not prohibitive.  

There are also principles that are proper to the digital age. Here is an overview of these principles:

  • Data ownership and data privacy : users should be able to control the amount and type of personal data used by a Platform / Website and for which purposes this data is used.
  • Open access: in the case of platforms that play the role of intermediaries between consumers and providers of a goods or services, the platform should avoid any discrimination in terms of access to the platform for either the buyers or the suppliers of goods and services..
  • Data portability / Multihoming options: the buyers and the providers of goods and services that operate on a Platform should be able to switch easily to a competing Platform – if there is one.

Taking into account these principles, there are a few criteria that could be applied to assert if an Internet Platform abuses its market position:

  • Pricing schemes: i.e. through rampant cross-subsidies between different services
  • Abnormal switching costs that could prevent customers from multi-homing and exiting any platform
  • Blatant conflicts of interests

Finally, if there is a strong evidence of a violation of Antitrust laws and policies the regulator could go for the “nuclear options”. In order to curb any potential or proven market abuse, they could force the platforms to implement the following measures:

  • Divestment from subsidiaries and controlled firms that cause conflicts of interest by providing the same goods and services that are provided by third parties operating on the Platform
  • Horizontal functional unbundling of different services and applications when a platform is simultaneously present in different markets that are used for cross-selling and cross-marketing products – although this sort of cross-selling is not exclusively found in the Internet world 
  • Horizontal geographic unbundling –  This seems largely irrelevant in the Internet Age within the same country. It could be easier to apply on an international basis. 
  • Vertical upstream unbundling of Internet Platforms from physical Infrastructures that might be used to gain an unfair advantage (e.g. AWS Cloud Business)
  • Vertical downstream unbundling from physical distribution networks (e.g. Apple stores).

Alexandre Kateb