Breaking Big Tech Up Isn’t the Only Approach for Antitrust

Last week, Andrew Yang, the dark-horse presidential candidate who has positioned himself as at once the most tech-savvy and the most technophobic candidate in the race, released a plan for regulating Big Tech. His ideas range from the mundane but wise (revive the Office of Technology Assessment) to the head-scratching (working directly with companies on new algorithms). But, given the ongoing antitrust investigations being conducted by Congress, federal agencies, and state attorneys general, it’s what Yang has to say about antitrust that most deserves our attention.

In a section titled “A modern approach to antitrust and regulation requires a 21st century framework,” Yang writes that “we must ensure that control over the most powerful technologies in history don’t accrue in the hands of a few. However, we must recognize that 20th century frameworks of breaking up companies just based on size or pricing impact on consumers won’t be effective. Network effects will always ensue, as a dominant player invariably emerges. And no one wants to use the fourth best search engine.”

Yang’s take on antitrust exhibits a lot of the confusion swirling around antitrust and Big Tech. It’s worth unpacking.

There’s more to antitrust than just breaking companies up.

We can partly blame Elizabeth Warren for this one. Back in March, the senator explained her plan to “break up Big Tech” in a Medium post. Since then, the discussion around what to do about the power of the Silicon Valley giants has largely been reduced to the question of whether they should be split up. (The irony is that Warren, who gave a long speech about competition and monopolization in 2016, may have a more sophisticated understanding of competition policy than any other major candidate.)

Antitrust law has in fact always been about more than breaking up companies because they’re too big. Everyone remembers the 1984 breakup of the Bell telephone monopoly, but an earlier antitrust suit, resolved in 1956 without breaking up the company, was perhaps even more significant. In that case, Bell signed a consent decree forcing it to share its patents, most notably the transistor, for free, leading to an explosion in innovation in the budding computer industry.

“The ‘56 consent decree was pretty important,” said Richard John, a historian at Columbia University and the author of Network Nation: Inventing American Telecommunications Licensing. “It’s not ‘break them up,’ it’s just opening the vault.”

The federal antitrust statutes are broadly worded, designed to give the government the power to punish companies for anticompetitive conduct—or, as the original federal antitrust statute put it in 1890, any action “in restraint of trade.” When it comes to Big Tech, there are plenty of alleged restraints to choose from, whether it’s Google manipulating its search algorithm to favor big advertisers or punish competitors (a charge Google denies) or Amazon undercutting third-party merchants by using their sales data to market its own imitative private-label products. Under existing law, antitrust enforcers in the Department of Justice and the Federal Trade Commission have the authority to bring legal action against companies for that kind of conduct.

The problem, from the perspective of critics like Warren, is that since the late 1970s, the government has hardly been using its power. Over the past decade, a loose movement of academics, lawyers, and journalists—sometimes referred to as the New Brandeis school or, more derisively, as “hipster antitrust”—has argued that the move away from antitrust enforcement has been a disaster. (Warren is one of several leading Democrats influenced by the movement, including rival presidential candidates Amy Klobuchar and Cory Booker.) On Monday, Columbia law professor Tim Wu, a prominent member of that group, published a brief manifesto laying out specific principles for revived antitrust enforcement. Notably, many of the items aren’t about company size or breakups; they’re about the need to crack down on certain types of corporate conduct, like “predatory pricing,” or charging below cost to drive competitors out of business and corner the market. Since the early 1990s, the legal standard for predatory pricing has been almost impossible to prove—which Wu and others argue has enabled the rise of companies like Amazon, Uber, and, until its spectacular flameout, WeWork.

“It shouldn’t be a business model that you can lose money for years, flood the market with investment money, and sort of emerge at the end of it,” Wu said. “Which is a very big business model right now, especially for Silicon Valley firms.”

The tech giants didn’t get so dominant just because of “network effects.”

There’s a grain of truth to the network effects argument that Yang alludes to, at least when it comes to Facebook. Some aspects of online platforms favor centralization more than in the average business. The value of a social network to an individual user depends on some degree to knowing everyone else is using it, too. On the other hand, monopolists have argued throughout American history that their stranglehold over sectors of the economy was merely an example of market efficiency.

“The operating companies in telephone used network effect arguments for political purposes in the 1880s and 1890s,” said Richard John. “The idea that a bigger network is necessarily going to be a better network is just a claim, like any other claim.”

Take Yang’s comment about no one wanting to use the fourth best search engine. Google, which represents more than 90 percent of the search market, may indeed still be the best search engine. But there’s nothing about its size that guarantees quality. A key premise of antitrust is that in the absence of real competition, monopolists no longer feel pressure to keep improving their products. And with Google, complaints have emerged in recent years over the growing share of results pages devoted to ads and Google-produced content such as “answer boxes” and “featured snippets”—design changes that seem optimized for the company’s ad revenue more than for the quality of the results.

The network effect argument also overlooks the fact that the dominant tech companies operate in dozens of separate industries. Amazon isn’t just a retail platform, and Apple doesn’t just make phones. When critics of Big Tech argue for the giants to be broken up, they’re not usually talking about splitting Google search into four pieces. They’re talking about blocking and reversing the kinds of acquisitions that built these impenetrable empires in the first place. Sometimes those acquisitions help the giants maintain their dominance in an existing field, like when Facebook bought Instagram or Google took over Waze, a challenger to Google Maps. Other times, they allow the big platforms to expand into other businesses, as with Amazon’s purchase of Whole Foods or the Twitch gaming platform. None of that has anything to do with network effects.

Another clue that the internet isn’t some unique area of monopolization is the fact that consolidation has been happening for decades across all kinds of decidedly analog industries, including meatpacking, beer, and hospitals. And evidence keeps piling up that this has all kinds of bad effects on things like income inequality and innovation.

According to Wu, the goal of antitrust law should be simple: eliminating barriers to entry.

“If there are those monopoly profits out there, there should be a big lure for people to get in,” said Wu. “If no one’s getting in, there’s a barrier. The ongoing mission of the law should be to identify and break down those barriers.”

The fact that presidential candidates like Warren and Yang are talking about antitrust at all is a major development. But the discussion is unlikely to be very illuminating until it moves beyond whether or not to “break up Big Tech.”


More Great WIRED Stories

Read More