The Internet, it used to be said, would let a thousand flowers bloom. The low cost of market entry would enable thousands of application and content developers to reach out to users. There’d be more diversity, more choice, more freedom for users to exercise that choice.
It hasn’t quite worked out that way. Sure, it’s easier for individuals and small companies to bring something to market, but it’s proved hard for most of those to make a living. There’ve been a few big winners, but they’ve been just a few. The market for the biggest Internet services – search, social media, online retail, Internet-enabled services – is dominated by a few big firms.
Instead of choosing between many different platforms, we’re dependent on a few. So what went wrong (or right, depending on your point of view)?
Telecoms and Internet: a tale of two modalities
Last week, I revisited the process of liberalisation that turned state monopolies in telecoms into competitive private sectors. What similarities and differences are there between the market dominance we once experienced with telcos and that which is evident online today? And what similarities and differences are there between the Internet today and other concentrated global markets like those for oil, cars and pharmaceuticals?
I’ll pick out two similarities, one concerned with each of those scenarios, and rather more differences, before suggesting implications for policy on both competition and the Internet.
The first similarity’s just the existence and degree of dominance. Pre-liberalisation (fixed) telecoms markets were (usually) monopolies. Mobile markets have (usually) been competitive from the start but many show high degrees of market dominance.
And that market dominance has emerged in online platforms too. Google’s share of online search worldwide is over 80 per cent. Facebook is similarly dominant in social media and messaging, YouTube in video, Twitter in microblogging, Amazon in online retail, Wikimedia in reference. This isn’t universal – alternatives have equal dominance in certain markets, notably in China – but it’s widespread.
The second similarity’s concerned with efforts to perpetuate or grow that market dominance. Market leaders in most sectors buy up potential competitors and strategic innovators in order to maintain and diversify their market share. Big tech companies have been no different. Facebook acquired Instagram and WhatsApp to protect its market against potential adversaries. Google bought YouTube to diversify its offering. Amazon’s the world-leading cloud computing business. Technology companies are reaching keenly into other, newer digital markets, including autonomous vehicles and artificial intelligence. They’re no longer Internet firms, but engaged across the range of new technologies.
… but also differences
These similarities matter. Big tech firms are following in the footsteps of other business sectors. But there are also major differences. Understanding these is important in working out the implications of market power for public policy. I’ll focus most on differences between telecoms of yesteryear and the platforms of today.
How did we get here?
First, we got where we are today in very different ways. Historic telecoms monopolies were created and enforced by governments. Today’s platforms won dominance through private enterprise, attracting users by offering services they valued highly as additions to their daily lives. Instead of having market dominance given them, they’ve earned it.
Second, they have different business models. Phone networks, like other retail services, charge users money for the time and goods/services (here, data volumes) they consume. Their users are their customers. It’s different for those leading online platforms that offer free services to users. Their real customers are advertisers. Their users provide them with the assets (user data) that maximise their value to those advertisers. Their users and their customers are different.
Third, network effects (like Metcalfe’s Law) are stronger in these online markets than they are in telecoms. Larger telecom networks have more value to their users than smaller ones because they enable users to reach more other users. Larger platforms have the same advantage but it’s exacerbated by access to data. Larger platforms can provide not just more contacts but more accurately targeted search results or online content – more accurately targeted for both users and advertisers – because they’ve access to much bigger data sets.
And fourth, I’d say, they’ve benefited from different attitudes because they’ve grown so quickly from start-up to market dominance.
Users welcomed the benefits of greater social interaction and wider access to content that they like that platforms have provided. Membership of Facebook, Instagram, Snapchat communities felt at first like being part of an insider group, out there at the forefront of the Internet. Google and Wikipedia users felt likewise back in the day.
Governments (rightly) welcomed innovation and new services like these as positive contributors to economic prosperity and social welfare. It’s only recently that both governments and users have started to have doubts about the concentration of market power and the impact of platforms’ business models on politics and privacy.
Where are we now?
Three further differences, then, concerned with the way platforms have grown in recent years.
First, their geographic reach. Those old telecoms monopolies were national businesses (as well as being run by governments). Today’s mobile telecoms providers are often multinational, based in different regions, but there’s a fair degree of competition between them in most countries; and they follow national rules and regulations.
The leading online platforms are far more global in their reach and dominance (though China’s an exception), and principally headquartered in America. This makes it extremely difficult for any individual government to regulate them in the way that they regulate other communications media or other economic sectors.
Second, their scope. Platforms often claim they’re merely intermediaries – channelling information to end-users (Google, Facebook, Twitter), or connecting buyers and sellers (Amazon, Uber, Airbnb). They compare themselves, accordingly, to post offices and telecommunications networks.
In practice, though, they’re much more than mere pipelines for their users’ content. Their algorithms that choose content for individuals are exercising editorial functions which are curated by the platforms. Increasingly, they’re also offering content of their own creation.
Market dominance in media contexts has more than merely economic significance. It also poses risks to plurality in public discourse, which have been a separate strand in regulatory practice. Initially, it’s likely that new platforms increased the diversity of content to which users were exposed. These days, there’s more concern that algorithms are narrowing the range of content to what they know their users like. More diversity for users overall may have led to less diversity for individuals.
Third, their potential impact on the future. As already noted, online businesses are leveraging their financial clout and economies of scope and scale to build positions in other high-tech areas, such as data management and artificial intelligence. They’re bound to do so in order to secure their future. High capital costs mean, too, that the scope for start-ups in these areas is limited.
These future markets may, as they approach new markets from different directions, be less dominated by single firms than online market segments now. An oligopoly with several major players sharing dominance, of the kind familiar in other capital-intensive sectors with large economies of scale such as oil, automobiles and pharmaceuticals, may be more likely.
So what are the implications?
The paragraphs above are meant to offer a description, not a criticism. I’ll make four final observations.
There are aspects to the digital sector that make it susceptible to market dominance – particularly the powerful network effects which are evident in key parts of the sector, and the economies of scope and scale that give large and global businesses advantages over the small and local.
These will continue, but that doesn’t mean that any particular instance of market dominance we see today is going to be permanent. The pace of change in markets and services means there’s rapid turnover in which market segments are most significant. Dominance is likely to continue but there’ll be change in where it matters most, and in how dominant individual firms are in the segments that are most significant tomorrow.
The extent of dominance by leading online platforms was not anticipated by any major stakeholder (including the platforms themselves). It’s come about quickly, in an emergent sector with an innovative ethos, which was expected to be competitive and has no experience of competition law. There’s been little time to come to terms with this. Internet businesses still see themselves as different from those in other sectors, and many governments still share that view. While it’s clear that national competition law (anti-trust in the United States) would not tolerate such dominance in other sectors, the global nature of the Internet means that few governments have the authority to intervene effectively.
Traditional approaches to competition law are unlikely, therefore, to be effective. There are potential similarities, it’s true, in the effects of market dominance, but there are big differences too in the nature of these markets and how they’re changing. If more competition’s deemed desirable, old regulatory mechanisms aren’t likely to achieve it. New thinking will be needed, by governments and online businesses alike, about what it means and how to foster it.
Image credit: rawpixel.com