When Mark Zuckerberg appeared before Congress earlier this year to discuss how the now-defunct political-data company Cambridge Analytica acquired data of up to 87 million Facebook users without their knowledge or consent, one of the few pointed questions came from Lindsey Graham, a Republican senator from South Carolina. “Who’s your biggest competitor?” Graham demanded. After Zuckerberg replied that Google, Apple, Amazon, and Microsoft all had some overlap with various Facebook products, Graham chafed at the answer.
“If I buy a Ford and it doesn’t work well and I don’t like it,” pressed the senator, “I can buy a Chevy. If I’m upset with Facebook, what’s the equivalent product I can go sign up for?” A bit later, the senator came back to the theme when he asked if the social network’s CEO thought Facebook was a monopoly. “It certainly doesn’t feel like that to me,” said Zuckerberg.
But to many people, it does. With over two billion users, the company is the colossus of social networking, dwarfing rivals like Twitter and Snapchat. Along with Amazon and Google, which is owned by holding company Alphabet, it dominates the internet landscape. Apple and Microsoft are often mentioned in the same breath as these tech giants, but their business lines are more varied and less internet-centric—enterprise software in Microsoft’s case, phones and other devices in Apple’s.
There’s another key difference too. Facebook, Google, and Amazon all have business models that require them to scoop up large amounts of data about people to power their algorithms, and they derive their power from this information. It’s the sheer scale and sophistication of the data-collection empires they’ve built that make them so distinctive.
For the past decade or so, these three firms have had a relatively smooth ride to the top. Their cornucopia of services, often provided for free, made them immensely popular and turned them into some of the most valuable businesses in the world. Their combined market capitalization of some $2 trillion at the end of May was roughly equal to the GDP of Italy. Now, however, debates are in full swing on both sides of the Atlantic about how to deal with their dominance.
The history of technology has seen singularly powerful corporations before—think of IBM and its reign in mainframes, and Microsoft, the undisputed heavyweight in the PC era. What’s different this time is the enormous influence the big firms have over so many parts of daily life, and the troubling issues this raises.
The Cambridge Analytica affair is just the latest in a long line of data scandals that have dogged Facebook. In 2009 it made information about users public without their permission; a few years later, Facebook researchers deliberately manipulated News Feed posts seen by almost 700,000 people to test whether they could influence users’ moods without their knowing. (Yes, was the disturbing answer.) Google, too, has had privacy run-ins, and in 2012 it was fined by regulators in America for circumventing default settings in Apple’s Safari web browser to place ad tracking software on people’s computers without their knowledge.
These incidents may appear isolated, but they fit into a bigger picture. Like the oil barons at the turn of the 20th century, the data barons are determined to extract as much as possible of a resource that’s central to the economy of their time. The more information they can get to feed the algorithms that power their ad-targeting machines and product-recommendation engines, the better. In the absence of serious competition or (until Europe’s recently introduced General Data Protection Regulation) serious legal constraints on the handling of personal data, they are going to keep undermining privacy in their push to know as much about their users as they possibly can.
Their dominance is allowing them to play a dangerous and outsize role in our politics and culture. The web giants have helped undermine confidence in democracy by underestimating the threat posed by Russian trolls, Macedonian fake-news farms, and other purveyors of propaganda. Zuckerberg at first dismissed claims that disinformation on Facebook had influenced the 2016 election as “pretty crazy.” But Facebook itself now says that between June 2015 and August 2017, as many as 126 million people may have seen content on the network that was created by a Russian troll farm.
Facebook and Google have built new tools for identifying disinformation and vetting advertisers, but it’s not yet clear how effective these will be. Even with news that is not clearly fake, researchers have shown that Facebook’s content-recommendation algorithms tend to serve up stuff that reinforces people’s prejudices. This might be happening even if the social-media industry were more fragmented. But the immense reach of platforms like Facebook has undoubtedly magnified the impact: according to a Pew Research Center study published last year, 45 percent of American adults now get at least some of their news from Facebook.
Then there’s the considerable market power they’ve built up, which has created turmoil in some industries and stifled innovation in the areas they dominate. Facebook and Google now amount to a digital advertising duopoly: they pocket three out of every four dollars spent on digital advertising in America, and they control 84 percent of global spending on such ads, excluding China. Google controls almost 80 percent of search advertising revenue in America and has a huge share in many other countries.
Amazon, meanwhile, accounts for over 83 percent of e-book sales in the US and nearly 90 percent of online print sales. The companies’ dominance has plunged the media and book publishing industries into turmoil: between 2006 and 2016, ad spending in US newspapers fell by almost two-thirds, and much of that money ended up in Facebook’s and Google’s hands. Amazon has also become a powerful online gatekeeper for many other kinds of online sales, and it handled around 44 percent of all e-commerce transactions in the US last year.
Their platforms give the data barons an unprecedented amount of control over what we see, read, and buy. Jonathan Taplin, the director emeritus of the Annenberg Innovation Lab at the University of Southern California, argues in Move Fast and Break Things, his book about the power of the internet giants, that rebel artists have long had to deal with “suits” who control distribution of their work. But the rise of companies like Facebook and Amazon has increased the stakes immeasurably. “The concentration of profits in the making of arts and news,” he writes, “has made more than just artists and journalists vulnerable: it has made all those who seek to profit from the free exchange of ideas and culture vulnerable to the power of a small group of … patrons.”
The data barons like to say that claims about their dominance are overblown. During his congressional testimony, Facebook’s Zuckerberg noted that the average American uses eight different communication and social apps. What he neglected to mention was that the social network owns several of the most popular ones, like its Messenger service and Instagram. Google argues that companies like Amazon and Facebook effectively compete with it in search by helping people find information, but its real competitors are dedicated search engines like DuckDuckGo and Microsoft’s Bing, which have relatively small market shares. Amazon can point to the fact that there are lots of companies offering e-commerce services, and that it competes with brick-and-mortar retailers, but its dominance in areas like book publishing is impossible to ignore.
The data barons’ power makes startups extremely reluctant to challenge them, and makes venture capitalists wary of backing the few mavericks that do. Speaking at an antitrust conference earlier this year, Albert Wenger, a managing partner at Union Square Ventures, said that one of founders’ top priorities these days is to avoid the internet giants’ “kill zones,” the areas in which they are capable of crushing any competition. And those zones are only going to grow as the web companies plunge into more businesses. Breakthrough ideas often come from startups rather than from large firms, so this could be depriving us of important innovations.
It wasn’t supposed to be this way. By lowering barriers to entry and making it easy for consumers to switch services with a few clicks of a mouse, the internet in its early days seemed designed to ensure that digital empires would promptly be besieged by fleets of rebel startups. So why didn’t that happen?
Part of the answer involves one of Silicon Valley’s favorite buzz phrases: “network effects.” Many online products and services become more valuable as more people use them. Buyers flock to Amazon because they know they’ll find lots of sellers—and hence lots of choices; people join Facebook because their friends are there. The US internet giants have been particularly skilled at harnessing these effects, as have Chinese firms like Alibaba and Tencent, which have become similarly dominant in their home market.
Thanks to network effects, Facebook, Google, and Amazon have been able to harvest oceans of data, which they use to continually refine their products and services. That, in turn, wins them even more users, which yields even more data, and so on. When other businesses show signs of succeeding in their markets, the data barons have often swooped in to buy them using their high-priced shares or vast cash reserves. Facebook bought Instagram and WhatsApp; Amazon picked up Zappos and Quidsi, two fast-growing online retailers; and Google acquired Waze, which was on the road to becoming a serious competitor to Google Maps. Sometimes consumers don't even notice these deals: after the Cambridge Analytica scandal broke, some Facebook users posted that they intended to move to Instagram in protest, clearly unaware it belonged to Facebook.
The reason the data barons have been so aggressive is that they are all too aware of how network effects can be turned against them by rivals and used to threaten their data-driven monopoly power.
Why haven’t antitrust regulators blocked deals to promote competition? It’s mainly because of a change in US antitrust philosophy in the 1980s, inspired by neoclassical economists and legal scholars at the University of Chicago. Before the shift, antitrust enforcers were wary of any deals that reinforced a company’s dominant position. After it, they became more tolerant of such combinations, as long as prices for consumers didn’t rise. This was just fine with internet companies, since most of their services were free anyway. Critics say trustbusters exercised too little scrutiny. “Just because the web companies offer products for free doesn’t mean they should get a free pass,” says Jonathan Kanter, an antitrust lawyer at Paul Weiss.
Another reason antitrust officials have struggled with the internet giants’ power is that they haven’t really appreciated how network effects can breed dominant market positions. At least Europe’s watchdogs have been tougher on anticompetitive behavior. Last year the European Union’s antitrust authority fined Google 2.4 billion euros ($2.7 billion) for unfairly favoring its own price--comparison shopping service in search results, depriving rivals of traffic. (The firm says it did nothing wrong and is appealing the ruling in court.) The EU is also investigating claims from rivals that Google uses its Android mobile operating system and AdSense advertising service to unfairly suppress competition.
In the US, the big web companies had lobbying clout and close links to the Obama administration, which may have given them an easier ride. But their standing with the government could be about to change: Steven Mnuchin, the US Treasury secretary, has urged the Department of Justice to take a hard look at the market power of big tech firms, and Joseph Simons, the new chairman of the Federal Trade Commission, which also has antitrust powers, said in his Senate confirmation hearing that he’d watch “big and influential” companies in Silicon Valley carefully. “I’m very optimistic that by the end of the year we’ll have a major investigation or two out there,” predicts Luther Lowe, the head of public policy at Yelp, a service that collects local reviews about things like restaurants and repair shops. Yelp has been locked in a long-running war of words with Google, which it says unfairly favors its own reviews in search results. Google rejects the charge.
If Lowe is right, the big internet firms could end up spending more time in American courts. But thanks to their vast wealth, fining them for any transgressions won’t diminish their power.
One radical solution would be to break them up, just as the US government splintered the dominant Standard Oil monopoly in the early 1900s. Some progressive advocacy groups in the US have been running online campaigns with slogans like “Facebook has too much power over our lives and democracy. It's time for us to take that power back,” and calling on the FTC to force the social network to sell Instagram, WhatsApp, and Messenger to create competition.
Facebook isn’t the only company in their sights. Earlier this year, Lina Khan, a researcher at the Open Markets Institute, one of the organizations behind the Facebook campaign, argued in a paper that because Amazon has become so dominant in e-commerce, it should be regulated and made to choose between being a seller of goods itself and running the digital platform it and other merchants use to reach customers. If it chose to be a platform, it would, among other things, have to spin off Whole Foods, the US supermarket chain it bought last year.
Making the legal case for breakups will be hard, though, because the internet giants don’t fit the stereotype of rapacious monopolists that raise prices and squeeze investment. They manipulate markets in a different and seemingly more benevolent way. They’ve become so dominant by developing products and services that many of us want to use. And they gain their immense power through collecting data about our online activity.
Still, just the threat of corporate dismemberment can have a salutary effect. In the 1990s the Department of Justice tried to force Microsoft to stop bundling its Internet Explorer web browser with its dominant Windows operating system, saying it was giving the browser an unfair advantage over Netscape. The government ultimately failed to get Microsoft broken up, but the bruising battle made the company more cautious about wielding its power to block small firms in emerging markets like online search—a fact that helped Google to flourish.
So how to curb the power of the data barons? Rather than waiting for legal battles that may or may not foster more competition, we urgently need to find ways to bolster rivals. That means reducing the vast chasm between the amounts of information held by the web giants and the rest. Regulation can help here: Europe’s new data privacy regime requires companies to hold people’s data in machine-readable form and let them move it easily to other businesses if they want to. This “data portability” rule will allow startups to get hold of more data quickly.
If one or another of the data barons is found guilty of anti-competitive behavior, a settlement should include a requirement that they have to share some of their data with rivals. Google, for instance, could be forced to hand over some search data that other firms working on search engines could use to train them, and Facebook could be made to share some of its “social graph” data on people’s online relationships. The best way to do this (while protecting people’s privacy) would need to be carefully thought out, but it would have a bigger impact than large fines, which the internet companies can easily pay.
Some argue that we need to think much more boldly—and not just with the big internet companies in mind. Viktor Mayer-Schönberger, a professor at the University of Oxford, has proposed what he calls a “progressive data-sharing mandate” that would apply to all businesses. This would require a company that has passed a certain level of market share (say, 10 percent) to share some data with other firms in its industry that ask for it. The data would be chosen at random and stripped of all personal identifiers. Intuitively, the idea makes sense: the closer a company gets to dominating its market, the more data it would have to share, making it easier for rivals to compete by building a better product.
Mayer-Schönberger’s suggestion may be hard to make work, but tackling the power of the internet giants will require novel approaches. It will also require a more muscular merger policy, one that looks beyond the narrow test of whether a proposed acquisition would raise prices to consider what it would do to future competition. We’re going to need to block not just large deals that would cement the web giants’ dominance but also smaller ones capable of eliminating competitors that might go on to challenge them. Carl Shapiro, an antitrust expert at the University of California, Berkeley, has said this could lead to some “false positives,” blocking acquisitions of young companies that never do turn into real threats to a Google or a Facebook. But he says this might be a price worth paying in order to stimulate more competition.
The need for such moves is even more pressing now that we’re heading deeper into the artificial-intelligence era. AI feeds on massive amounts of data to gain its power. So the data barons’ vast data reserves give them a head start in training AIs that will run all kinds of devices and systems, from driverless cars to software that decides whether or not you should get a loan. That’s going to make it harder than ever for other firms to catch up.
The prominent AI investor Kai-Fu Lee noted this in an article in the New York Times last year: “The more data you have, the better your product; the better your product, the more data you can collect; the more data you can collect, the more talent you can attract; the more talent you can attract, the better your product.” It’s no coincidence that Facebook, Google, and Amazon are intent on gaining as much data as possible and securing some of the brightest AI minds on the planet to work for them.
Increasingly, the AI-powered, voice-activated assistants these firms are building will be in our cars, homes, and offices, as well as on our phones. We’ll expect them to deliver “the” answer to questions, rather than the smorgasbord of suggestions that often gets served up today. The companies whose algorithms decide what those answers will be will have an even more powerful influence over us and over the global economy. And to ensure that they hold on to their dominance, Facebook, Google, and Amazon will soon be vacuuming up even more data about us.
During his congressional testimony, Zuckerberg accepted that new rules are going to be needed to govern his company and others. “So I think the internet is becoming increasingly important in people’s lives,” he said, “and I think we need to have a full conversation about what is the right regulation, not whether it should or shouldn’t be.” To create those rules, we urgently need to focus on the source of the internet giants’ power and the dangers that it entails. The sooner we find smart ways to diminish the firms’ dominance of our data, the better.