Search Warrants

The fact of Google’s search monopoly itself isn’t in question. Recent numbers from private information firms give Google a 90 percent market share of the search engine market in the United States and 91 percent globally, usually with even higher numbers for mobile search compared to desktop.

The numbers, however, are not necessarily evidence of nefarious conduct by Google as much as basic economics. Economists have long recognized the dynamic of network effects, the pattern of some services gaining value as more people use them. Phone service and social media platforms exhibit this pattern — as more users join these services, you can reach more people with them and enjoy more communication or online content. Unlike other goods like food or clothes, their value depends on how many other users there are, and once such a service becomes large, it attracts most future growth and tends to become dominant.

This pattern explains why so many of today’s online markets are “straight out of the box monopolies,” as the conservative Economist magazine put it. Much as the phone network became more valuable as more users connected to it, so too do functions like online search gain value with more queries. This is because as users conduct searches using the engine and then choose a result (and perhaps return to click another if unsatisfied with the first choice), Google records this information in search of the “long click” — the search result that meets the user’s need. This nominal improvement of the quality of the search engine comes from the greater volume of user queries, which trains Google’s search algorithm to improve its selection and ranking of search results.

As more users turned to the search engine, its results improved based on their queries and result choices. This meant that the first good search engine, Google, quickly gained market dominance simply due to its far-higher quality in the early days of the industry. (Anyone old enough to have conducted searches using Lycos or WebCrawler will understand this point.) It also meant that sustained high use was central to keeping result quality up, giving Google an incentive to pay for more traffic, the main issue in this suit.

Default Lines

We can now turn to the meat of the case. Since 2001, Google has given a share of its revenue to smartphone makers (original equipment manufacturers in industry jargon, or OEMs) like Apple and Samsung, as well as to browser makers like Mozilla, to get them to set Google Search as the default option on their phone handsets and web browsers.

Importantly, the deals are not known to be exclusive — another search operator could theoretically bid away the default option from Google. But the amounts involved are colossal, billions annually, with Apple being the most important: Wells Fargo estimates Google will pay the phone maker $23 billion this year for the default search position. The two companies have had a profitable if conflicted relationship since the advent of mobile tech in 2007, and the deal has been quite advantageous to both firms, cementing Google’s search monopoly and providing Apple with a reliable cash source plus ensuring a place for the crucial suite of Google apps — Search, Maps, Gmail, and YouTube. Google also made sure Apple couldn’t send searches to its Siri AI application.

Smartphone OEMs running Google’s own Android operating system are under even more obligations, since they license the operating system from the company. Google has requirements to place its search engine in the browser defaults, forbids preinstalling or promoting rival engines, and in some cases makes Search undeletable by the user.

The fact that the deals are not exclusive makes them harder to legally challenge. The government case is in part that, by paying to channel searches to itself, Google is keeping Microsoft’s Bing and Yahoo Search from getting enough traffic to improve their own products. Google’s claim that users prefer it because it offers a better search service is true, but it is also a reflection of Google’s big spending to keep the queries coming. The company is probably on thinner legal ice regarding its well-established history of downranking competing specialized “vertical” search engines that work in specific areas, like Tripadvisor or Yelp, or scraping their reviews and information and putting them directly on Google’s own search results page.

Google has other advantages, including an important lesson learned from Microsoft’s case. That trial included Bill Gates’s infamously smug and evasive video deposition being directly rebutted by extremely embarrassing emails from Microsoft executives, including the infamous alleged executive’s remark about “cutting off Netscape’s air supply.” These moments were media-ready and extremely damaging, including in the court of public opinion.

Google, on the other hand, has created a program called “communicate with care,” in which high-level execs would reliably copy a company attorney on shady emails, thereby creating attorney-client privilege and making the documents inadmissible in antitrust trials. The company also instructed its employees to avoid words like “kill” and “crush” in their emails, quite unlike Gates; in one message presented to the court by a DOJ lawyer, Google CEO Sundar Pichai asked for the chat history to be turned off in an apparently incriminating conversation.

Many of Google’s attorneys are in fact veterans of the 1990s Microsoft antitrust trial — their chief in-house legal officer and top outside litigator were both involved in prosecuting Microsoft. They represented and wrote briefs for Netscape, but are now working for the new monopolist rather than the underdog.

Further, the crucial legal fact is that Google’s position is as a default, not the only option. Google is expected to argue in court that users are free to change their search default in their browser settings. And indeed, this is easy, free, and takes only a few moments.

Yet default settings contain significant power. Very few of us in our daily mobile software ever change our default settings, due to the time needed to learn about competing options and to learn which details a user prefers. Defaults therefore have a real shaping effect on user choice, as Google itself has long been aware. “If you control the menu,” a “design ethicist” with the company has commented, “you control the choices.”

Borked

America’s anti-monopoly laws are notoriously weak, even in comparison with other capitalist countries. The United States’ antitrust law was only created late in the country’s industrial development and after decades of utterly free play by gigantic trusts like John D. Rockefeller’s Standard Oil.

Whereas the EU’s competition laws allow action to be brought based on market dominance, US antitrust law does not outlaw monopolies, but rather monopolization — in other words, your monopoly may or may not be legal depending on how you got it. Microsoft’s own decades-long global monopoly over computer operating systems was completely legal, because it arose from network effects: Microsoft’s Windows was used for the influential early IBM PCs, creating a body of users that developers of applications and games wanted to reach, and the growing body of software for Windows in turn attracted more users.

Since no arms were visibly twisted to create the Windows monopoly, it was never seriously legally challenged. But in 1995, Microsoft set out to crush start-up web browser Netscape by bundling its own weak-sauce browser Internet Explorer into its Windows 95 update, putting its product on everyone’s computer worldwide. It also used its major clout to keep PC manufacturers from including rival browsers under threat of loss of access to Windows, and paid large amounts to AOL, Apple, and others to use Explorer.

These actions crossed a line that even the US government couldn’t accept. By using its existing, legal monopoly to then take control of another market, Microsoft committed “monopolization,” for which it was ultimately convicted in 1998, declared legally to be a monopolist, and ordered broken up. Only the dubious 2000 presidential election saved Microsoft, when George W. Bush’s administration directed the DOJ to drop the breakup remedy and instead settle for weak “behavioral” restraints: requiring the company to end its exclusive agreements with PC OEMs.

The system was further weakened by Reagan-era changes to antitrust law led by figures associated with the archconservative University of Chicago, but above all by Yale law professor and later Supreme Court nominee Robert Bork. Bork held that big business was unfairly restrained and maligned in the then-receding New Deal era, and he claimed that enormous market power should not be enough on its own to challenge a corporate merger or order a breakup. Instead, this should only happen if it could be proved the company had raised consumer prices.

This view had the effect of dramatically decreasing the level of antitrust vigilance and unleashed a merger orgy throughout the 1980s and ’90s, creating the landscape of hyperconcentrated markets and gigantic global companies that we live in today. Since then, wages have grown slowly, while profits and the incomes of the richest households have shot up.

Bork’s legacy is especially powerful in online markets, including search, because there the product is provided free to the consumer. The reality of network effects — testified to by the company’s own officers — and the resulting tremendous scale of the companies are not enough to bring government scrutiny under Bork’s standards, since the consumer is clearly not paying a high price. The company’s monumental data hoard, its near-complete control of many unbelievably important markets, and its far-reaching influence over government policy have no relevance to this standard.

This obviously creates a major advantage for companies like Google, Facebook, and others that, despite great market power, aren’t actively raising prices and indeed aren’t charging any at all. Even Microsoft could not rely on this defense, since it charged for Windows software.

Late in his career, Bork actually wrote a preemptive defense of Google for the prestigious Journal of Competition Law and Economics, saying the mere “possession of monopoly power in the relevant market” isn’t enough for government action, and the law prohibits only “the willful acquisition and maintenance of that power” — i.e., monopolization. Bork is better known for his later failure to be confirmed to the US Supreme Court, but in the long-lasting effects of his legal doctrine, he has enjoyed quite the revenge.

Bad Breakup

Among the possible trial outcomes is a breakup, what is legally referred to as a “forced divestiture” — a legal mandate to sell off chunks of a company, as with AT&T in 1984 or Standard Oil in 1911. This is a classically American remedy, as it involves taking the lightest approach to interfering with the sacred market — turning a monopoly into an oligopoly, a small number of giant firms dominating a market instead of just one. Many pieces of Standard Oil later remerged (ExxonMobil is a product of the former Standard Oil of New Jersey and Standard Oil of New York), as did AT&T. But it remains a major penalty for giant corporate empires.

Alphabet could be seen as an especially attractive candidate for a breakup, for the simple reason that apart from the original Search unit, all the company’s other business lines were added through acquisitions (including the predecessor of Google Maps in 2004, the Android mobile operating system in 2005, YouTube in 2006, and the crucial ad server DoubleClick in 2007). This makes it easy to see how divestment lines could be drawn, again much as in the case against Microsoft; there a breakup was contemplated to separate the monopoly’s Windows operating system business from its Office applications unit.

But it doesn’t look likely that Alphabet will be remade in this manner. Breakups are considered a somewhat drastic outcome of antitrust, and indeed no large corporation has been ordered broken up since AT&T in 1982. Since this case only partially involves advantages the company gained from its conglomerated units, it is more likely that the court will merely mandate changes in how Google does business.

The most punishing likely outcome for the company would be a consent decree to desist from the commercial revenue-sharing deals with the OEMs, removing their incentive to set Google Search as the default on phone handsets and their preloaded browsers globally. A requirement not to downrank competing engines like Tripadvisor is another possibility.

But these sorts of behavioral requirements have an unimpressive history, including with Microsoft. As part of its settlement with the EU in 2009, Microsoft was ordered to include “ballot screens” for users to choose browsers, but at times the company stopped complying when it thought no one was looking. Microsoft was fined $733 million for the offense in 2013, about 1 percent of its revenue for that fiscal year. But even that kind of slap on the wrist will depend on a federal judge deciding Google harmed consumers despite never raising prices they pay directly. The shadow of Bork and his fellow corporate cronies hangs over the case.

Of course, this is only one trial. Google just agreed to settle a suit with a group of US states over its app store Google Play, although it remains in litigation with store users like Fortnite-maker Epic Games and Match Group, owner of dating apps Tinder and OkCupid. EU legal actions continue, and Google is also being separately sued by the US DOJ over its advertising business, the real plum, expecting to go to trial next March.

But even an unlikely breakup wouldn’t fix the major problems in these markets, with one or perhaps two gigantic entities ruling over most of our online lives. For something as toweringly important as the internet’s main source and channel of information — the “master switch,” as Tim Wu calls it — public ownership and some form of democratic control should be on the table.