tim brennan, former chief economist at the Federal Communications Commission, is senior fellow at Resources for the Future and professor of public policy at the University of Maryland, Baltimore County.
Published October 24, 2019
Everyone — well, certainly regulators from California to Washington to Brussels — seems to be hot on the trail of Google, Facebook and Amazon, the big tech platforms. Concerns range from invasion of privacy to undermining democracy by spreading fake news. But at (or not far from) the top of the list is antitrust. Press reports suggest that both the Department of Justice and the Federal Trade Commission have opened investigations. Meanwhile the EU has already acted, imposing multibillion-euro fines on Google for limiting both access to non-Google applications and the ability of competitors to sell advertising that reaches viewers through Google-managed searches.
These cases against Google seem to run into a fundamental problem: the services Google offers are free. And under the standard “consumer welfare” test in U.S. antitrust, if prices don’t go up, consumers aren’t harmed. This has led frustrated critics of Google to argue that U.S. antitrust law needs to be broadened in scope — for example, by including criteria such as privacy or small business protection.
I offered a general critique of these assertions in an analysis last year in the Milken Institute Review. But there’s no denying that if the standard approach precludes an examination of Google’s practices, such critiques ring hollow. I still believe, though, there is a way to reconcile prevailing antitrust practice with a case against Google. The key is to ask whether Google chooses to offer its services for free or is driven by the marketplace to offer its services for free. To see why, take a ride with me on the “big tech antitrust” time machine.
The most auspicious day in the history of antitrust was January 8, 1982. On that day, Assistant Attorney General William Baxter announced the resolution of the antitrust cases against the two most prominent tech giants of their era. Baxter dismissed the then-13-year-old antitrust case against IBM “with prejudice” — about as negative a declaration about the merits of a case one could make. At the same time, he announced a settlement of the 8-year-old case against AT&T, which led to the most substantial change in market organization in antitrust history.
The IBM case was dropped by the Reagan administration; the AT&T case had been litigated, in the words of Mr. Baxter, “to the eyeballs.” Why the contrast? Both companies were tech giants that had monopolies in industries critical to the health of the economy. Both had been accused of bulldozing competitors from their markets. Both had been subject to excruciatingly long and invasive litigation spanning multiple White House administrations. But to Baxter, there was a world of difference.
If one tells an unregulated company like IBM that it can’t favor affiliates that sell products meshing seamlessly with IBM computers — back then, they were called “peripherals” — Big Blue could have responded by raising the price of computers to everyone, regardless of which company’s peripherals they used. This may well have been a worse outcome than letting IBM create competitive advantages for its peripheral subsidiaries.
AT&T would have had no such counter because regulation would have kept it from raising the price of its (then) monopoly local telephone service. But it could have evaded that regulation by impeding competitors to its affiliated long distance and equipment companies (remember Western Electric?) and overcharging for these products and services. Thus it could have indirectly exploited the monopoly power that regulation nominally controlled.
In a nutshell, breaking up IBM would serve no consumer welfare purpose, but breaking up AT&T would.
Fast forward to Google. Google, of course, has no price regulator telling it that it must offer its search services or its Chrome browser for free. One might then think that the IBM analogy, where antitrust wasn’t the answer, is appropriate. However, the driver in the AT&T case was not regulation as such, but the inability to raise the price of its monopoly service if it were not able to discriminate against competitors in related businesses.
Regulation is not the reason Google offers its monopoly services gratis. It may simply be too costly to charge for those services.
That may well be the case with Google. Regulation is not the reason Google offers its monopoly services gratis. It may simply be too costly to charge for those services. Charging anything markedly differs from charging no price. Requiring payment means that a seller has to monitor who is using its product and perhaps how much, implement a mechanism for collection, and enforce a firewall to prevent those who do not pay from using the service. This is clearly worth it for some online content providers — think of The Wall Street Journal as a familiar example. But for Google search, which is used billions of times a day, the costs of setting up a payment system might well be prohibitive.
If this accurately describes Google’s thinking with regard to pricing search services, then the company may only be able to capitalize on the value of search through indirect means. In particular, when people are searching for services where Google can independently charge fees — say, for advertising on YouTube or booking travel on Google Flights — the big platform company could have an incentive to bias search results to point people in that direction.
The crucial point in this scenario is that if Google were prevented from doing this, it would lack the ability to respond by charging monopoly prices for any video or travel searches. No one’s price would go up as a result of such an action. The fact that the service is free becomes not a puzzle requiring a new way of prosecuting antitrust, but the justification for applying principles used 40 years ago to break up the then-monopoly telephone company.
Before the advocates of bringing antitrust cases against big tech get too excited (and those opposed too appalled), a couple of important caveats need to be kept in mind. First, this “has to be free” justification need not apply to the other giant network enterprises. While Amazon does not charge buyers to shop around, it certainly charges a margin above the wholesale price to carry and ship products. Facebook, it seems, could charge a fee for maintaining a listing because it maintains a relatively long-lived relationship with its users. Its choice to offer this service for free seems based on marketing considerations relating to exploiting the enormous size of its network rather than the impracticality of charging for the service. But the facts could turn out otherwise.
A second caveat, and maybe a more important one, is that even where Google might have an incentive to exploit market power in search services by funneling traffic through them, there may be important savings to Google and consumer benefits as well. The savings can come from the advantages from coordinating pricing and design across multiple services. The consumer benefits, for their part, could arise from faster searches, improved quality control, better customization and other service improvements. The 1980s breakup of AT&T was premised on the conclusion that the cost savings and customer benefits from integration were small. That may not be the case for Google.
• • •
That’s why antitrust action against Google should come only after a careful assessment of both Google’s anticompetitive incentive to integrate and the benefits of allowing it to continue business-as-usual. However, that assessment does not depend on a rejection of the “consumer welfare” standard and adoption of some novel theory of what antitrust is about. Antitrust first principles continue to work — not despite, but because Google’s services are free.