‘Social media’, market power and the health of democracy

With the whistle blown on Facebook, Congress must allocate ownership of personal data to the person—not the platform—to allow competitive providers to emerge.

According to its former employee Frances Haugen, Facebook algorithms consciously amplify dangerous misinformation and privilege the most divisive content posted on the network. Such content is more frequently shared by users and foregrounding it maximises traffic on the platform—and so turnover.

This modus operandi, which became still more aggressive from 2018, is generating perverse incentives pushing even relatively moderate users to sharpen and polarise their content to obtain visibility. It is a Darwinian struggle for prominence which, given the rules of the game, leads to the survival of those users most fit for division and risks skewing public opinion and altering political outcomes. A recent working paper I co-authored shows that exposure to political information through ‘social media’ has been closely associated with the diffusion of divisive ideas in Europe in the last decade.

Haugen also revealed that, as the volume of divisive content circulating on the platform grows, it becomes more difficult, and more expensive, to monitor, especially in marginal areas where the economic return is not sufficient to justify the associated expense. This is a very dangerous short-circuit, especially in times when co-ordination via web platforms could issue in last January’s siege of Capitol Hill in Washington.

Digital monopolies

What is worse, and what dramatically exposes democratic societies to the consequences of the algorithms deployed in Menlo Park in California, is that Facebook and its Big Tech peers, mostly also located in Silicon Valley, occupy dominant positions in extremely concentrated markets. A few firms control the information and communications technologies sector, acounting for increasing shares of physical assets, revenues and market capitalisation. Apple, Microsoft, Amazon and Google’s Alphabet (with Saudi Arabia’s Aramco) lead the ranks of the top 100 companies in the world.

As for ‘social media’ specifically, a few platforms account for most of the traffic of opinions and information they colonise on the web. Facebook, which took over Instagram and WhatsApp, is certainly the biggest, coming next in the PwC ranking.

Why such an impressive push for scale in the digital world? Recognise first that public goods are those that are non-rival and non-exclusive (such as the air we breathe). Unlike most private goods and services, data are non-rivalrous and can be reproduced at no or minimal cost—as with ideas and knowledge more generally. But they are excludable and can thus be a source of monopoly.

An expanding system could facilitate the entry of new participants. But firms involved in the production of non-rivalrous goods will tend to seek ways to build fences around them, to engender scarcity artificially—and, in the process, generate rents from the assets they own.

In contrast to true public goods, exclusion is possible in the digital ecosystem through a combination of scale effects, strengthened property rights, first-mover advantages and other anti-competitive practices. The ‘network effects’ through which everyone gains by sharing the use of a service or resource—nowhere more evident than on content platforms—have given rise to demand-side economies of scale, which allow the largest firm in an industry to increase and lock in its attractiveness to consumers and gain market share. This makes it almost impossible for competitors to become attractive and challenge market dominance.

Digital mononopolies are made more dangerous by the fact that most people do not see them as a problem. The perceived price for using a platform such as Facebook or the services provided by Google is zero, even if of course this is not the case. Operating on multi-sided markets, these giants can cross-subsidise, sacrificing profit by constraining one side to enhance the attractiveness of (and recoup losses on) the other.

Google and Facebook offer their products free in exchange for personal data, which makes them more attractive to advertisers. Ultimately, Facebook’s or Google’s market power in advertising increases and so does the average cost of advertising, which will be eventually reflected in the price of goods.

Restoring competition

One way to address monopoly in a digital world and pave the way for a more pluralist and efficient market would be to break up the large firms responsible for market concentration. This takes literally the frequent comparison between, respectively, oil in the analogue and data in the digital economies. Standard Oil, which controlled 95 per cent of US refineries and had deals with the railways which restricted the ability of others to compete, was broken up in 1911 and required by law to split into many pieces.

The tendency of the market to generate monoplies, however, would make the new configuration inherently unstable. Another approach would be to change the structure of the market in a more profound way, so to avoid the risk of any future such agglomeration.

In this sense the proposal advanced a few years ago by Luigi Zingales and Guy Rolnik, to reconfigure data ownership, is today more relevant than ever. In a nutshell, the University of Chicago economists propose a legislative reallocation of property rights akin to what has been done on the mobile market, where some countries have established that a phone number belongs to the customer—not the provider. This redefinition of property rights, or ‘number portability’, has made it easier to switch provider and so has fostered competition.

Along the same lines, in the social-network space, it would suffice to reassign to each customer the ownership of all the digital connections they created, a ‘social graph’. This way customers could sign into a Facebook competitor and instantly reroute all their Facebook friends’ messages to the new platform. By guaranteeing the latter access to new customers’ data and contacts, ‘social graph portability’ would reduce the positive network externalities favouring the existing platforms and ensure the benefits of competition.

The US domination of ‘social media’ and other content platforms—with the top seven such firms all originating there—is evident. Any solution will therefore require legislation by Congress. The White House knows that the momentum generated by the Facebook scandal will fade and that the window of popular support for major changes to the technology landscape will close. The time for action is now.