This spring, Amazon did something very un-Amazonish.

Across many popular product categories, from batteries to baby food, the retail behemoth quietly discon­tinued aggressive promotions for its private-label brands, which compete with--and in truth are often near-clones of--independent merchants' products. It was an uncharacteristic retreat for a company that generally loves nothing more than using every weapon in its substantial arsenal to annihilate its rivals.

Call it the Elizabeth Warren effect. Amid growing public wariness of the biggest tech companies and their outsize role in our economy and public life, the Democratic presidential candidate is far from the only politician in her party demanding stiffer regulation. Even President Trump has blasted Amazon as a "no-tax monopoly," and his Federal Trade Commission is spinning up a tech task force "to ensure consumers benefit from free and fair competition." Still, it's Warren's proposal to break up Google, Apple, Amazon, and Facebook--on the grounds that "they have hurt small businesses and stifled innovation"--that has captured imaginations and newspaper headlines. In Mountain View and Menlo Park, they're suddenly sweating through their hoodies.

Hence Amazon's about-face. "They're trying to get out in front of the issue and defuse any regulatory solution," says economist Hal Singer of Georgetown University and Economists Incorporated. To Singer, whose work has influenced policy proposals from senators including Mark Warner and Al Franken, it was a hopeful sign.

Another came from Facebook CEO Mark Zuckerberg, who, after years of lobbying against government oversight, recently called for legislators to establish ground rules for "harmful content" so companies like his don't have to figure them out on their own. Concessions like these suggest we can keep the digital giants from squelching competitors without resorting to extreme measures.

That's good, because there are plenty of reasons to be skeptical about the idea of breaking up big tech. For one thing, such trustbusting doesn't address the real problem: platform owners abusing their absolute dominion over what have become, for many, indispensable venues for doing business.

"When you're dealing with natural monopolies, you have to take away their ability to exploit their control over a marketplace," says Barry Lynn of the Open Markets Institute. Splitting off Amazon's e-commerce operation from its cloud services division wouldn't do much to help merchants on the former who see themselves as victims of the company's self-dealing, or enterprise software startups on the latter who suspect Amazon is using their data to copy their products.

Amazon controls 85 percent of U.S. e-commerce sales for the categories of arts and crafts/party supplies, and 83 percent of U.S. e-commerce sales for the category of household essentials.Source: Jumpshot

Preventing dominant incumbents from buying up newcomers--another favored fix--might create more competition, but it also might depress startup formation, since founders often see acquisition as their most reliable exit strategy. Antitrust is full of such question marks, and the academic literature underpinning the case for trust-busting is riddled with inconsistency, says Geoffrey Manne of the International Center for Law & Economics. "The evidence is weak," he says. "Noticing there's an increase in concentration and asserting it correlates with some outcome you're worried about doesn't mean there's a causal relationship."

Even those who generally agree with Warren's diagnosis of the problem see her call for breakups as more rallying cry than policy prescription. "You have to get everything down to a bumper sticker," says Singer. This may focus minds on the matter, but it distracts from competing remedies--which are straightforward, historically validated, and politically plausible.

 inline image

One is enforced nondiscrimination. This is essentially a legal framework preventing platform owners like Google and Amazon from using their control to advantage their own interests, by, say, favoring their own offerings. The principle that a company whose network represents an essential public service has an obligation not to discriminate is a principle with deep roots in American law, dating back to the establishment of railroad and telegraph lines.

"The only network monopolies we've not applied these rules to are the platform monopolies," says Lynn, "because they've spent a hell of a lot of money to ensure it wouldn't happen." (Lynn and his team were with the think tank New America until, he contends, Google, a major funder, pressured its leadership to cut ties with him. Google has denied this.) Indeed, the tech giants have already endorsed the principle of nondiscrimination. They've lobbied the FCC on behalf of so-called net neutrality rules that prevented internet service providers from offering faster connectivity to companies willing to pay more.

Nondiscrimination has been a useful tool in regulating the cable television industry, and Amazon's willingness to back down on promoting its own goods at the expense of outside sellers suggests the tech giants can live--and continue to innovate--when that becomes the norm. "Yes, their profits will be lower than in a world where they're free to discriminate with impunity," says Singer, "but not materially lower." His idea: federal legislation establishing a nonpartisan "net tribunal" that would adjudicate claims of digital self-dealing.

Breaking up big tech companies is the nuclear option: Possessing the option provides useful leverage. But if you're in a position that requires you to seriously consider using it, something has gone badly wrong. Why not try fixing what's broken before pushing the button?