Charles Platiau / Reuters

The battlefield is getting crowded. European antitrust enforcers have been fighting America’s tech giants for years. In the U.S., both the Justice Department and Federal Trade Commission are headed for the sound of the guns. And now the states have entered the fray. On September 9, 2019, 48 state attorneys general, led by Texas Attorney General Ken Paxton, launched an investigation into Google’s potential violations of antitrust laws.

The revolution in high tech is confronting liberal democracies with some of the most difficult public-policy challenges since the Industrial Revolution, and the dawn of artificial intelligence promises far more to come. Antitrust enforcement will remain an important tool—particularly for protection against certain horizontal restraints on trade, such as price-fixing cartels. But in the current global free-for-all of antitrust enforcement against American’s tech giants, how much good can state officials in the U.S. reasonably do? The answer is: for the public, not much, but for themselves, quite a lot.

A compelling antitrust case against Google might focus on one or two specific business practices that might constitute a restraint on trade under Section 1 of the Sherman Act. But an oceanwide fishing expedition for evidence of potential monopolization under Section 2 of the act will quickly become mired in a complicated economic analysis of supply chains that are changing by the day. And as the states’ subpoena makes clear, they’re pursuing the latter approach—focusing on Google’s behavior in the market for online advertising.

A state case for showing that Google is guilty of monopolization will have to establish that Google has “monopoly power in the relevant market,” and that it willfully acquired or maintained that power “as distinguished from growth or development as a consequence of a superior product, business acumen, or historic accident,” under the precedent established by the Supreme Court in U.S. v. Grinnell Corp. in 1966.

A market analysis produced by Plum Consulting, in the U.K., that’s largely critical of Google argues that defining the relevant market for online advertising is difficult for several reasons. First, these markets are multisided: Is the relevant market the market for advertisers, or is it the market for publishers that stand on the other side of Google’s intermediation? Second, there are lots of different market segments, and they all overlap. Third, and most important, the market structure itself is changing at rapid speed.

Google has achieved astonishing success in both the web-browser and search-engine sectors. The firm controls well over 90 percent of global search-engine traffic, with much of the rest divided between Microsoft’s Bing and Verizon’s Yahoo. But nothing guarantees that Google will be able to maintain this position indefinitely, or even for very long. Already in the U.S., Amazon, with its treasure trove of high-quality data, has overtaken both Microsoft and Verizon in online advertising, and could rapidly become a compelling alternative to Google.

Has Google engaged in anticompetitive practices to achieve or maintain its monopoly position? It could theoretically do so in several ways, according to the Plum report, including by using network effects, economies of scale and scope, vertical integration, and data access to exclude competitors. But in each of these cases, as the Plum report concedes, the potential for abuse exists alongside enormous benefits to the public. For example, the public prefers a network with more advertisers and publishers on it, an example of a clearly beneficial network effect. Meanwhile publishers and advertisers are able to reach their target markets more easily and accurately, which means much greater bang for the buck. And paradoxically, the reduced transmission of raw data among different firms means that privacy and security may be more easily and accountably maintained by a Google than by multiple firms in a more fractured market.

As the famous antitrust cases against IBM and AT&T demonstrated, investigations of the dreaded monopoly power take little account of the technology industry’s actual history, which in the modern era has been one of relentless disruptive innovation. The answer to Microsoft’s supposedly fearsome “tying arrangement” of Internet Explorer to Windows was not antitrust litigation, but rather a disruptive innovation in the combination of search and browser, as Google soon showed. Netscape must be kicking itself.

Antitrust law consistently produces the very losses that it imagines it’s protecting the public from. Consider this classic example from Robert Bork’s 1978 book, The Antitrust Paradox. The basic danger of monopolies (and cartels) is that they may use their market power to reduce output and raise prices, thereby capturing profits well above competitive levels and imposing unfair losses on the public. Now suppose that through efficiency and innovation, a supplier of widgets has achieved 90 percent market share, and antitrust enforcers impose a settlement whereby no competitor in that market will be allowed to achieve more than 50 percent market share. Suppose the settlement is implemented, and the same efficiencies and innovations set in, and one of the competitors starts to approach market dominance. To avoid crossing the 50 percent threshold, it would have to reduce output. As a result, in a market cartelized by government action, prices must rise. Antitrust enforcement thereby creates, in the place of an imaginary monopolizer, a very real price-fixing cartel, and a price-fixing cartel of the worst kind: one whose otherwise-ephemeral cartel discipline is enforced by the government’s own antitrust enforcers.

Those problems are daunting enough at the federal level, where antitrust enforcers at the FTC and Justice Department have acquired a respectable level of expertise. State antitrust enforcers, on the other hand, face all those problems with far less expertise—and infinitely greater incentives for opportunism.

State officials routinely use antitrust enforcement to shield their favored constituents from interstate competition. That is the long and dreary history of state attorney general investigations of proposed mergers inside their jurisdictions. They may not have any idea whether the proposed merger will benefit the public in the long run. But they know what their major campaign donors and key constituents want.

It’s no accident that California and Alabama are the only two states that didn’t join the Texas investigation. Alabama has bent over backwards to attract companies like Google, which recently built a large new data center there. As for California, if it wasn’t home to Google’s headquarters, it would almost certainly be leading this latest charge. California is one of the states most adept at shielding in-state producers from competition. The 1943 antitrust case of Parker v. Brown, which created an ill-advised exception to the Sherman Act for monopolies and cartels protected by state governments, upheld a California law that inflated its raisin producers’ profits by foisting higher prices on consumers outside the state. (At the time, California produced 95 percent of the country’s raisins.)

The second and even more licentious problem of state antitrust enforcement is that state AGs sometimes use litigation to shake down private industry for enormous amounts of cash, and the losses get passed on to consumers. The classic example is the 1990s tobacco settlement. In that case, the major tobacco companies agreed to restrict tobacco production, raise cigarette prices, and pass hundreds of billions of dollars to state coffers, so long as the states agreed to protect the newly minted cartel. The settlement, the George Mason University professor Michael Greve has argued, created “a national de facto tax, in excess of a quarter trillion dollars, that no legislator, state or federal, ever voted on,” a tax that has been paid for almost entirely by consumers. The tobacco companies, meanwhile, made out like bandits.

The states’ move against Google could benefit state officials enormously. Whether it will benefit anyone else is the paradox of Paxton’s crusade.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.