It is a joke that sounds like hubris, and there is indeed plenty of that to be seen. But in this context global ambitions on the part of the only-just-begun are oddly reasonable. If your idea for a service or product can be scaled up to cover the world, why would you not plan to do just that? And if your idea cannot be scaled up that way, should you not find one that can? After all, capturing a significant, even dominant share of the world market more or less straight out of the box is clearly possible. It has been crucial for the internet’s biggest successes: Amazon (About half of America’s book market, more than that in e-books); Alibaba (about 80% of e-commerce in China); Facebook (which claims 1.3 billion active members); and Google (68% of online searches in America, more than 90% in Europe).
Regulators worry that such dominance lays consumers and competitors open to all sorts of abuse. A European Parliament resolution calling for the European Commission to consider the break-up of Google by splitting its search engine from “other commercial services”, which was to be voted on shortly after The Economist went to press, is a bit of cheap political grandstanding. But the commission’s concerns about Google are real. (Disclosure: Google’s executive chairman, Eric Schmidt, sits on the board of the company which owns this newspaper.)
And it is not the only digital giant which has come under scrutiny. Many writers have called for Amazon to be investigated on antitrust grounds, though the proximate cause for concern, a dispute between Amazon and Hachette, a publisher, has now been settled. Worries about Facebook tend to focus more on privacy than on market dominance—but the two concerns are not unrelated. Both speak to alarm at the sheer size of the companies concerned.
Standard issues
Something about the internet clearly favours such mushrooming quasi-monopolies. But rather than being a regrettable by-product of internet commerce and in need of strict oversight, might they instead be an integral part of what makes it so prone to rapid growth and life-improving innovation—not a bug, as the phrase has it, but a feature? Peter Thiel, one of the founders of PayPal (now the market leader in online and mobile payments), and an early investor in Facebook, argues that monopolies add “entirely new categories of abundance to the world”. If he is right, then treating monopolies in the digital realm just like their bricks-and-mortar—or oil-well-and-pipeline—predecessors would be very bad for innovation and growth.
In America, and in most of the rest of the world, private monopolies are treated with deep distrust. The law, though, does not deem them bad in themselves. In the 1911 decision which cut Standard Oil into 34 different companies, America’s Supreme Court held that the fate of a dominant company should depend on whether it abused its dominance to the detriment of consumers. This seemed, at that time, to be the case for Standard Oil. But the court also heard that the company—which had dominated the oil business since the 1880s—had previously done much for consumers and society at large. It produced kerosene that could be relied on to burn brightly while not exploding, as rival products often did (hence the name “Standard”). It spent heavily on infrastructure and research, developing new products such as lubricants.
Those early years provide fodder for Mr Thiel’s argument that monopolies can be agents of progress. The promise of years of monopoly rents, he points out, provides a powerful incentive to innovate. And the profits monopolies can bring allow bold long-term plans and ambitious research projects that “firms locked in competition can’t dream of”. In a recent book, “Zero to One”, he argues that economists’ “obsession” with competition—the poor saps believe it generates incentives for companies to serve customers better—is “a relic of history”. Competition, he says, explains failure, not success; success comes from providing a unique solution, and thus naturally tends to the monopolistic.
This will not necessarily come as a surprise to economists. They have understood the desire to avoid the rigours of market competition since Adam Smith; Joseph Schumpeter was writing about the innovation that a dominant position makes possible in the 1930s. But Mr Thiel does provide an up-to-date playbook for those seeking to build a business this way.
A clever startup does not try to compete directly with an incumbent. It picks a seemingly unimportant market which it can monopolise. The classic example is Facebook, which began with Ivy League student bodies, but today’s other giants also focused on things others cared little about. The size of the search market seems obvious now, but it wasn’t when Google got started; the big player, AltaVista, was a loss leader for a struggling equipment-maker. When Amazon aimed at books it was not, admittedly, that small a niche. But it was a deliberately circumscribed subset of the market the company intended to take on eventually (Amazon really has been megalomaniac from the beginning).
Once it has gained a foothold, the firm raises venture capital and tries to enter bigger markets and grow as quickly as possible. “It’s about building the next big platform,” says Reid Hoffman, the founder of LinkedIn (the world’s biggest social network for professionals, with 300m registered users), another noted Silicon Valley venture capitalist. The latest example of this is Uber, which is seeking to dominate the world of taxis (and perhaps, one day, urban logistics more broadly). When the company launched in 2010 in San Francisco it offered only the services of drivers with full-sized luxury cars in and around San Francisco. It went on to raise $1.5 billion in venture capital, giving it a valuation of more than $17 billion. With this it has expanded to 230 cities in 50 countries. It is adding a new city almost every week, and its next fundraising round may value it at $40 billion.
No amount of money, though, could make such growth possible without the digital world’s increasingly mature and ubiquitous infrastructure. Uber’s clients all have phones on which they can download apps and credit-card details with which they can pay; its drivers’ phones have navigation software; the company has algorithms for pricing and predicting travel time that can be applied more or less anywhere, and can be scaled up from thousands of customers to millions just by buying more computing power in the cloud. If Uber can convince local regulators that its service is legal—sometimes a big If—it and its backers see few barriers to continued rapid growth.
Once growth has begun, another monopoly-enabling aspect of internet businesses kicks in. “Network effects” are the modern equivalent of a principle first enunciated in St Matthew’s Gospel: “For whoever has will be given more, and they will have in abundance. Whoever does not have, even what they have will be taken from them.” The most obvious example of the phenomenon is the rise of Facebook. As the social network attracted new members it offered more people for all those members to interact with, which increased the benefits of membership, which attracted more members.
All the big internet success stories except Apple—which is mainly a hardware-maker—have relied on network effects. In eBay’s case more buyers meant a better market for sellers, while more sellers made it a more interesting marketplace for buyers. There is a useful distinction to be made between this and the boost Facebook got from network effects. There the benefit was direct: the addition of new members helped all members. But in the eBay case the effect was indirect. The addition of new buyers did not directly help existing buyers—indeed, it may sometimes have worked to their detriment—but by attracting more sellers it helped them indirectly. These indirect network effects are more subtle than their direct counterparts. Indeed they are subtle enough to have attracted a lot of interest from academic economists. The Nobel prize awarded to Jean Tirole of the University of Toulouse this year was in part a recognition of his work on the topic.
The need for speed
Again, Uber provides an example of the phenomenon at work. As the firm expands the number of drivers it has in a market, the time it takes for a car to get to a customer shortens, which attracts more passengers, which in turn begets more drivers. As its business grows, drivers also have less downtime, meaning the firm can lower prices, which again attracts more users.
Not that Uber is leaving its success in the hands of purely economic drivers. In August Lyft, Uber’s main internet competitor, claimed that the firm had used dodgy methods against it, with Uber “brand ambassadors” requesting Lyft rides and then either cancelling or pressing drivers to defect to Uber; the company has responded that it never uses “marketing tactics that prevent a driver from making their living”. Vanity Fair quoted the firm’s boss, Travis Kalanick, saying he had intervened in his rival’s fundraising efforts by warning investors that Uber would soon raise more capital itself, pushing them to invest with him. One of the company’s executives has suggested the firm should consider spending $1m to dig up dirt on its media critics (he later apologised). The robber barons of days of yore would look on in admiration.
This goes too far even for Mr Thiel, who has called Uber “the most ethically challenged company in Silicon Valley”. But he is happy for would-be monopolists to push the envelope of established practice on the basis that the end result is a force for good. Some academics agree, at least with the latter part of the analysis. Digital monopolies, they have argued, are essentially harmless; Google, which in 2010 was the first of the new monopolies to come under close antitrust scrutiny, has provided some of them with a worked example.
Their first point is that digital monopolies are often not real monopolies, either because there is no selling going on, or because the bit of the market they dominate is not the whole story. Thus Google may dominate the world of search and the associated sale of advertisements, but it gets no direct revenue from the first and in the second it competes with a range of other people in the larger overall market for online advertising (see chart 2).
The second argument is that network effects and advantages of scale do not in themselves amount to barriers to entry. Facebook was not the first social media company, and did not have unique access to network effects. In many markets it faced competitors with many more members, such as Myspace or Orkut. Google, which operated Orkut, closed it down in September. MySpace is barely a shadow of its former self. There are plenty of firms both large and small trying to take on Google in any number of ways. Whatever barriers Google may enjoy, they are clearly not sufficient to put off all comers.
And, crucially, all this competition is only a click away. “That consumers can switch to substitute search engines instantaneously and at zero cost constrains Google’s ability and incentive to act anti-competitively,” wrote the late Robert Bork, a conservative judge, in a 2012 paper commissioned by Google that he wrote with Gregory Sidak, an antitrust expert. The paper had a wide influence; in 2013 the Federal Trade Commission (FTC), which had been looking into action against Google, decided not to proceed.
Similar arguments can be made about the other digital giants. EBay may dominate online auctions, but theirs is a small and shrinking piece of a growing e-commerce pie. Facebook may benefit from strong network effects, but many users “multi-home”, meaning that they are also members of other online clubs, such as Pinterest and Twitter. Amazon has an even stronger incentive than Google to keep customers happy and prices low: people tend to shop around for the products it offers at all sorts of other sites (a tendency Google is more than happy to encourage).
A space of one’s own
Yet this analysis paints too rosy a picture. It is a platitude of the online world that everything competes with everything—for attention, if nothing else—but the internet giants rarely compete strenuously on each other’s home turf. Instead they scuffle for position in markets that are less important to them while trying to build up defences around their core businesses.
And in those cores they do have singular advantages; all the big internet companies “own a space”, even if the nature of that space is not always obvious to outsiders. Google, for instance, has a huge advantage in knowing what people want to buy, and thus a dominant position when it comes to auctioning off these leads to would-be sellers. Facebook knows more than anyone without security clearances about people’s friendships, their networks of acquaintance and their areas of interest. (Interestingly, though, for many purposes Facebook’s trove of finely nuanced information does not provide a must-have edge; most social-media apps can learn as much about their users’ contacts as they need to get started just by tapping into their address books.)
There will still always be would-be competitors, established and startup alike. Some of them will be too ambitious, big or unprepossessing to be taken over by the incumbents (a time-honoured strategy). And the really big ones can afford to build global networks of data centres and dedicated fibre-optic cables, just like the incumbents have, in order to do battle. But they are likely to struggle to fill them up with data and algorithms.
Amazon and Google have amassed huge piles of data that keep their recommendation, search and other algorithms sharp. Google boasts a remarkable concentration of data scientists capable of using the information to make search results and ads more relevant. Network effects bring it ever more data; continued success and smart strategy ensure it keeps its edge in brain power. Though Microsoft has been able to build, in Bing, a broadly Google-equivalent search engine, it has not managed to make it a profitable business, despite billions of dollars spent trying.
This is not just because of Google’s data advantage. It is also because, even if competition is always only a click away, most users do not seem interested in taking the step. People establish habits and routines to make their lives easier, and once services such as Google’s search engine and, even more importantly, mail and map apps become incorporated in those routines they stick. The switching costs in Google’s other market are even higher. Running ad campaigns on Google is complex, so though big firms use many platforms, small firms that have learned how to use Google tend to stick with just the one.
It is on this point that one of the more troubling worries about Google rests. One of the things that economists find interesting about indirect network effects is the scope they provide for subsidising one class of network member in order to profit from another. Thus, for example, the makers of games consoles keep prices relatively low to get lots of users and then, having spread their format widely through subsidy, charge the people who produce games a substantial fee to make use of it.
Google provides free search so that companies will pay top dollar for a prime position in the list of ads next to the results. Prices in auctions for key words, which trigger ads when searched for, have risen as high as $50 in some cases, though the average price is $0.50 (and only paid if the ad is clicked on). “Google takes more than any other actor in the chain,” says Ben Edelman of Harvard Business School, who has done work for Microsoft, Google’s upstart competitor in the field (yes, there is an irony there). Google says the customers are simply paying for an efficient way to reach global audiences.
Some advertisers have complained about Google making it difficult to pull data about customer responses to advertising out of the system in an automatic way. When the FTC decided not to press on with its case against Google in 2013, the company said it would fix the problem. Microsoft complains that the fix has not gone far enough, and that advertisers are still having difficulty managing campaigns across platforms.
The hardest play is defence
Google and the other internet giants have all built up peculiar market power, albeit to differing degrees, and they have done so remarkably quickly. Others, perhaps even one of the seedling firms from 500 Startups, may yet do the same. Whether they actually abuse it, however, is an entirely different question. It may well be that monopoly powers may be achieved more easily online than off without any need for foul play. It does not follow that they are particularly easily kept once they have been built—or that foul play will not ensue in the struggle to keep them.
Indeed, it is possible that the difficulties of hanging on to them could be the trigger for bad behaviour. As the first flush of heady growth wears off and the technological or business edge that made it possible gets copied or ground down, the need to keep satisfying market expectations gets stronger and once insurgent, even idealistic cultures turn more corporate. Pressures mount, and it may become ever harder not to cross the line into the abuse of a hitherto well-deserved pre-eminence.
Part of what makes Amazon remarkable is that it has managed to keep itself in something like the early upswing of this trajectory for a peculiarly long time. It has never stopped investing heavily to reach scale, instead preferring to expand ever further the scale it aspires to reach. For the time being its investors remain willing to expect only growth, not returns—a patience which explains how Amazon has been able to build up an infrastructure, in terms of data centres and warehouses, that a rival would be very hard put to match. That said, even if there is no need to show profits, there is clearly a call for money made in the established bits of the business to fund growth in the newer parts. And some suspect the bully-boy tactics the company used against Hachette may be evidence that its shareholders are starting to get impatient.
Google is further along the trajectory. It is highly profitable, boasting a profit margin of nearly 20% and net income of $13 billion, in large part from its advertising business. Since the firm went public in 2004 it has regularly beaten market expectations, with average annual revenue growth of more than 30%. It has helped itself along by reserving top spots on its search-result pages for links to its own services, such as Google Shopping and Google Maps. It has offered more and more information directly on its search pages, often doing away with the need to visit other sites.
These proclivities may have led to interest from the FTC and the European Commission, but they have not given Google a “browser moment”—so-called in memory of Microsoft’s decision to bundle Internet Explorer, its web browser, with its Windows operating system in an attempt to kill off Netscape, which was enjoying huge success in the browser market. The closest Google has got is with Android, its operating system for phones. Although it is open source, meaning that device-makers get it free and with permission to tweak it, the official version comes with strings attached. If device manufacturers want their users to have access to Google Play—Android’s app store, which adds various useful functions to the operating system—they need to give Google’s mobile apps a particular prominence. This strikes some as being an unfair use of the company’s position; Google says such rules are necessary to keep the Android experience from falling apart.
Such actions are changing the perception of Google in Silicon Valley. One sign of the shift is the recent decision by Mozilla, the organisation behind Firefox, a popular browser, to no longer use Google as the default search engine for its software; instead it will use Yahoo, which is powered by Bing. Google has been doing as much as it can to promote its own browser, Chrome.
Google defends its actions by saying that they are meant to make life easier for consumers: showing information directly saves clicks; keeping handset-makers on a short leash helps to avoid confusion because of different interfaces. But it is hard to get around the fact that the same measures also cement and expand Google’s dominant position. And in the world of indirect effects, helping the user is not all that has to be done to play fair.
The FTC decided not to go after Google in 2013 because the company offered assurances that it would change some practices, such as letting advertisers get at their data. The commission reckoned that any harm done to competitors was outweighed by consumer benefits. It declined to “second-guess a firm’s product design decisions where plausible pro-competitive justifications have been offered.” In Europe, where the interests of competitors play more of a role in antitrust cases, regulators have taken a more critical view. “The Commission takes the preliminary view that Google is abusing its dominant position in the relevant markets for horizontal web-search services,” they wrote in a 2013 memo to the firm. Since then Google has offered changes in some of its practices, such as giving less prominence to its own offerings in advertising next to search results.
If data were everything...
This is the sort of thing that gives people like Mr Thiel conniptions. Governments should not be messing around with a dynamic success which promises only ever higher levels of innovation. “The dynamism of new monopolies itself explains why old monopolies don’t strangle innovation,” Mr Thiel writes in his book. The rise of mobile computing, driven by Apple, killed off Microsoft’s regulator-attracting desktop dominance. Before that Microsoft ended IBM’s mainframe monopoly.
Google too will be disrupted one day, quite possibly through a similar “platform shift” changing the dominant form of computing. This may already be happening. Android has made Google a force in the mobile world, but it cannot change the fact that that world operates by different rules than the world where Google grew up and makes money. Computing on mobiles is much more fragmented. People buy goods and services using apps, rather than by searching the web or looking at adverts. What is more, clicks on smartphones are not as profitable as those on a personal computer. If growth in Google’s search-advertising revenue has slowed recently, analysts say, it is partly because more clicks come from mobile phones.
Yet some still worry that Google could prove to be the ultimate digital monopoly. They do not think that its reason for being is primarily online search or the advertising business; they see it as being in the business of mining any and all data it can accumulate for new profit streams. The data hunger such a goal demands is the main reason, they argue, why Google is entering markets as diverse as self-driving cars, smart homes, robotics and health care. “Google is trying to leverage the advantage it has in one area into many others,” says Nathan Newman, a lawyer and technology activist. The idea is that Google could use its assets—its data, its unparalleled ability to exploit those data, its brilliant employees and knack for managing them—to take control of other industries.
For such a data-centric conglomerate to get ever more dominant seems against the flow of history and intuitively unlikely. But intuitive views of the direction of internet competition have been wrong before, as the existence of giants like Google, Amazon and Facebook bears witness. And should it show signs of coming to pass, the current antitrust skirmishes will give way to an epic battle on the scale of the one against Standard Oil. “If we will not endure a king as a political power,” said John Sherman, the senator who gave his name to America’s original antitrust law, “we should not endure a king over the production, transportation and sale of any of the necessaries of life.” Even one that makes things very, very easy.
More about:
