Dismantling Google is a terrible idea

The parallels draw themselves. In 1999 America’s government prevailed in a high-profile antitrust suit against a tech giant it alleged was abusing a monopoly. The case then turned on the “power of the default” in internet browsers: Uncle Sam said Microsoft was forcing computer-makers to distribute its browser along with its Windows software. It resulted in proposals to break Microsoft up (though the firm won on appeal and remained whole).

Tech-watchers could be forgiven for feeling a sense of déjà vu. In August trustbusters scored their first major victory against big tech in a quarter of a century when Amit Mehta, a judge in the District of Columbia, ruled that Google was a monopolist in online search. Using the power of the default, he argued, Google had blocked rivals and raised prices for its ads beyond free-market rates. On October 8th the Department of Justice is due to file proposed remedies for this abuse of monopoly power. These may include a proposal to dismantle the tech giant—perhaps by hiving off Chrome, its browser, or Android, its operating system for mobile handsets.

This would be foolhardy. It is not at all clear that it would solve the central issue presented in the case. And besides, even though Google has long enjoyed the vast profits associated with its vice-like grip on search, it may not continue to do so. New generative artificial-intelligence (ai) tools, such as Chatgpt and Claude, are quickly gaining market share.

Google is the web’s most-used search engine, handling about 90% of queries in America. That dominance, Mr Mehta ruled, has been cemented through “default search agreements”. Open up Safari on an iPhone or Mozilla Firefox on a laptop and type a query into the search bar, and it will be Google that returns the results. For the privilege of doing so Google shares some of the advertising revenue its search engine generates. These payments amounted to $26bn in 2021. Some $20bn went to Apple alone.

For firms to pay to be first in the queue for potential customers is hardly an outlandish idea. Cereal-makers pay supermarkets to be “eye-level” on shelves; publishers pay booksellers for spots on their coveted “front table”. The trouble with default search agreements is that they do not just make one option more prominent. They take choice away altogether.

Hiving off Chrome or Android would not fix this problem, as long as Google were still allowed to pay their eventual owners to be the default search engine. So the court should target default arrangements directly. It could limit Google to being able to pay to be one of a range of choices of search engine, a fix that European regulators have already put in place. Absent the fat cheque Google pays to be the default, Apple and other deep-pocketed tech firms might focus on building search engines of their own.

An order to force Google to make public some of the technology that enables its search engine to work, such as its index of web pages and search-query logs, could make it easier for rivals to try. The trial revealed that it costs an estimated $20bn to build a search engine, plus $3bn-4bn per year in annual research and development. Reducing those costs would let smaller companies compete, too.

Another reason to avoid a remedy as drastic as a breakup is that technology moves far faster than any legal system can. Add in the appeals process and any action against Google is still years away. Yet already there is emerging evidence that Google’s grip on search is slackening as generative-AI tools gain ground. A survey by Evercore, a bank, found that Chatgpt is the “go-to search engine” for 8% of Americans. Innovation dramatically weakened Microsoft’s dominance a quarter-century ago, too. The firm was swiftly left behind as mobile technology took off.

Antitrust intervention may have sped the firm’s decline. That is why it is important for regulators to look forward as much as back. If Google is left unchecked, the danger is that its incumbent position impedes competition. With its huge proprietary datasets, Google may one day build better ai tools than its rivals. Buoyed by its monopoly profits, it offers its current AI tools free, unlike newer competitors which must charge subscriptions to help cover their costs. If Google were indeed blocking future rivals, then limiting its ability to use its search engine to distribute its ai products might stop it from exploiting one monopoly to acquire another. A breakup, however politically appealing it may be to some, is not the answer. ■