By David O. Fisher, Senior Counsel at the American Antitrust Institute

As algorithmic pricing transforms our markets, the antitrust laws are being tested in new ways. Empirical evidence demonstrates that pricing algorithms facilitate price coordination, resulting in the textbook anticompetitive harms of higher prices and lower output. But they do so through complex and novel mechanisms which allow for tacit collusion in ways that antitrust law has come to treat as unlikely. Faced with this reality, antitrust law must cast off old assumptions about tacit cartel agreements. To face the threat that algorithmic pricing poses to competition, courts must rehabilitate tacit agreement as a theory of Section 1 liability and enjoin the shared use of algorithmic pricing tools that raise market prices above the competitive level.

Although tacit agreements to fix prices are in principle illegal under Section 1 of the Sherman Act, they have become increasingly difficult to prove because of steadily heightening evidentiary standards. Lulled into a false sense of security by Chicago School thinking, courts have implicitly accepted that oligopolies are more likely to engage in interdependent pricing than to collude and that cartel agreements are by nature unstable. But contemporary economic evidence demonstrates that market concentration is associated with not only higher prices but also long-lasting, durable cartels. Pricing algorithms can make cartels even more durable by using novel methods which allow for stable agreements among disparate actors. With machine learning, they even “learn” to adopt joint profit maximization strategies and to collude on their own. These capabilities further undermine our old assumptions about tacit cartel agreements, reducing the likelihood of false positives and raising the costs of false negatives. This should prompt us to question the burdens we put on plaintiffs to prove a tacit agreement.

As part of the project of rehabilitating the theory of tacit agreement, we should also be willing to enjoin the use of algorithmic pricing that results in sustained supracompetitive prices. Long before the advent of algorithmic pricing, Judge Richard Posner and others argued that oligopoly pricing is properly understood as an illegal agreement under Section 1: one oligopolist’s price increase is an implicit offer, which the others implicitly accept by matching. Recognizing the potential for mutual benefit, the parties perform by collectively maintaining supracompetitive prices. But according to former AAG Donald Turner and others, there is no effective legal remedy for oligopoly pricing. Judges are not equipped to set and regulate “reasonable” market prices, and they cannot order firms not to react to their competitors’ prices. Courts have tenuously resolved this debate by inferring a tacit agreement only when defendants engage in communication, information sharing, or other conduct that can be enjoined. In cases of algorithmic collusion, courts can enjoin the use of shared pricing algorithms that facilitate oligopoly pricing. Viewed in this light, the secret to addressing algorithmic collusion may be as simple as a well-crafted injunction.

The antitrust laws are flexible enough to address the threats that algorithmic pricing poses to competition. But if we want to use them effectively, we must revisit old assumptions about tacit cartel agreements, and we must apply the law in a way that accounts for the realities of this new technology. Only time will tell if we are up to the task.

The American Antitrust Institute (AAI) is an independent, nonprofit organization devoted to promoting competition that protects consumers, businesses, and society. We serve the public through research, education, and advocacy on the benefits of competition and the use of antitrust enforcement as a vital component of national and international competition policy.