The dominance of tech giants does not confront us with an unpalatable choice between laissez-faire and populist interventions. This paper takes stock of available knowledge, considers desirable adaptations of regulation in the digital age, and draws some conclusions for policy reform.
The initial enthusiasm for the ongoing technological revolution has recently given way to a global ‘techlash’. Many academics and policymakers call for taming the large tech platforms, regulating large tech companies as public utilities, breaking them up, using a tougher antitrust enforcement, or engaging in industrial-policy programs in big data and AI. This paper investigates the merits of the various arguments.
Economists’ standard view on what has been happening is that numerous industries are now subject to substantial economies of scale or scope, a winner-take-all scenario, and widespread market power. Incumbents enjoy direct network externalities (our concurrent joining of Facebook or Twitter allows us to interact through these platforms) or indirect ones similar to those associated with urban amenities (I benefit from your using a search engine, an app such as Waze, or a delivery service because that improves their quality). Competition in the market may also be limited by the existence of large fixed costs. For example, designing a first-rate algorithm, web-crawling and indexing (all of which are necessary for a search engine to be effective, especially if it aims to respond satisfactorily to uncommon queries) is onerous; accordingly, there are really only two players in the English-language segment, that is, Google and Bing, with Google overly dominant.
Limited competition stemming from network externalities and fixed costs generates large markups for winners and a concomitant willingness to lose money for a long time to buy some prospect of a future monopoly position. Firms accordingly need deep pockets, as is observed directly (Amazon lost money for a long stretch of time and Uber has engaged in expensive recruiting of drivers through bonuses) and suggested indirectly (firms that have never turned any profit reach phenomenal market caps).
Monopolies always raise concerns about high prices, low innovation and – if the monopoly position may be challenged – the possibility of abuses of dominant position against potential rivals. Tech giants are no exception.
The possibility of consumer harm through high prices is sometimes questioned by platforms on the ground that many services are available for free to consumers. This argument, however, ignores levies on the other side of the market. Advertisers pay hefty fees for advertising on the platforms; these fees raise their cost of doing business, with potential indirect harm to consumers. Similarly, the fees paid by the merchants, so their goods and services be listed and recommended by the platforms, increase consumer prices. The ‘no-consumer-harm’ argument also ignores the theoretical possibility that a zero price may still be too high, as data are extremely valuable to platforms (that is, the flow of payments should be toward consumers; this is discussed further below).
Yet high profits might be the cost to pay for the very existence of the very valuable services. In some way, the consumer must pay for the industry’s investment costs. So, a better posed question is: ‘are platform profits in line with investment costs, or do platforms enjoy “supranormal profits” or “ex-ante rents”?’
Whether the high profits made by Google, Facebook and other dominant platforms really constitute supranormal profits is debated; identifying supranormal returns requires data not only on profits currently made by a dominant firm, but also on the losses it incurred during the shakeout period leading to monopolisation, and on the probability of emerging as the winner of the contest. Needless to say, we have little data on the latter two variables.
Excessive prices are not the only issue with monopolies. As was recognised long ago, a monopoly’s management enjoys an ‘easy life’ and may not keep its costs under control, as it is not spurred by competition. Monopolies also may fail to innovate, as they are loath to cannibalise their own products. They may even fail to adopt minor innovations. A case in point is provided by the taxi monopolies across the world. The very useful ‘innovations’ introduced by ride-hailing companies such as Uber, Didi and Lyft (geolocation, traceability, preregistered credit card, electronic receipt, mutual rating, etc.) were neither new nor rocket science. Yet, they had not been taken on board by traditional taxi monopolies, resulting in suboptimal service. Interestingly, in some cities, the very same taxi monopolists reacted to Uber’s entry by introducing apps, accepting credit cards or offering a fixed price from the airport to the city centre. The virtues of competition in action….
Note, finally, that even if there were no supranormal profits, this would not mean that there is no scope for policy intervention. Firms might be playing dirty tricks in the marketplace, spending money on killer acquisitions or hiring battalions of lobbyists and lawyers to acquire or preserve their dominant position. Contestability does not rule out social waste.
This paper considers desirable adaptations of regulation to the digital age. It considers, in turn, the merits of alternative institutions and policies to regulate the tech sector; data-related issues; the resurgence of industrial policy and trade‐related issues; and institutional innovation. The final section concludes.
Cite this as:
Tirole, J. (2022), ‘Competition and the industrial challenge for the digital age?’, IFS Deaton Review of Inequalities, https://ifs.org.uk/inequality/competition-and-the-industrial-challenge-for-the-digital-age