Pricing Algorithms Are Not Inherently Anticompetitive
A federal judge recently blocked New York’s ban on landlords setting prices based on algorithmic recommendations. She was right to do so. New York’s law risked conflating ordinary efforts to maximize profit with unlawful conspiracies to suppress competition. The two are not the same.
Under General Business Law § 340-b(3), a landlord who sets rent "based on recommendations" from software that analyzes more than one landlord's data has entered an "unlawful agreement." The law effectively treats landlords’ use of the same third-party pricing software as evidence of an unlawful agreement, even without proof of a direct agreement among them.
In RealPage v. James, the case challenging New York’s law,U.S. District Judge Valerie Caproni blocked enforcement of subsection 3 but left the neighboring subsection 2 intact. Subsection 2 targets agreements among competitors not to compete, conduct already prohibited under federal antitrust law. As Caproni noted, it "does not make new conduct unlawful." Subsection 3 went further by imposing a new restriction on the use of pricing software, rather than targeting the underlying anticompetitive conduct.
Long before pricing algorithms, landlords, like other businesses, used competitors’ prices to inform their own pricing decisions. They studied competitors' listings and hired consultants to analyze market conditions. "There is nothing untoward about considering prices charged by competitors when deciding how to price one's own product," Caproni wrote. Yet subsection 3 risks treating the use of pricing software as unlawful even when it draws primarily on public information and landlords retain independent control over their pricing decisions.
Caproni concluded that the statute "prohibits normal commercial conduct just because it is facilitated by software." Her ruling rested on the First Amendment. She recognized the state’s interest in protecting competition and addressing housing-market concerns but concluded that the law was not sufficiently tailored to those interests. The broader point is that a state cannot justify an overbroad restriction simply by pointing to the possibility of anticompetitive conduct.
Pricing software can facilitate collusion, but the relevant question is how firms use it.The federal case against RealPage illustrates the distinction: The government alleged that the company used nonpublic competitor data and other practices to facilitate coordination among landlords.
In August 2024, the Justice Department sued RealPage, alleging that landlords supplied the company with confidential lease data every night, including rents after discounts, and that RealPage used the pooled data to recommend rents to their competitors. Landlords also traded information directly through "weekly phone calls, emails, and in-person visits." The goal was no secret. One RealPage executive explained that, with enough landlords on board, they would "likely move in unison versus against each other." These allegations concern far more than the use of publicly available listings to make independent pricing decisions.
Federal antitrust enforcers recognize that benchmarking—comparing a firm’s practices and performance with those of other companies—is not inherently anticompetitive. FTC Commissioner Thomas Rosch explained in 2011, “Information about prices, costs, capacity and availability can benefit both producers and consumers, by allowing markets to function more efficiently, intelligently, and competitively, at prices and output levels that more accurately reflect changes in supply and demand.”
The Supreme Court drew a similar distinction in Maple Flooring Manufacturers’ Association v. United States (1925). It plainly states, “the natural effect of the acquisition of wider and more scientific knowledge of business conditions, on the minds of the individuals engaged in commerce and its consequent effect in stabilizing production and price, can hardly be deemed a restraint of commerce…or in any respect unlawful.” Continuing, and especially relevant to pricing algorithms, the Court specifies that a “[r]estraint upon free competition begins when improper use is made of that information through any concerted action which operates to restrain the freedom of action of those who buy and sell.”
Pricing algorithms can provide firms with the same kind of improved market information that the Court discussed in Maple Flooring. Their use does not, by itself, establish that competing firms have agreed to restrain competition. The DOJ acknowledged as much in response to public comments regarding its settlement with RealPage, stating, “The complaint does not allege that adoption and use of the same revenue management software by competing landlords, standing alone, is necessarily anticompetitive.”
Instead, the DOJ’s lawsuit against RealPage alleged that the company violated Section 1 of the Sherman Act by soliciting “current, forward-looking, granular, and highly competitively sensitive information” from landlords and using this information, in tandem with “user group meetings and its revenue management certification program,” to align pricing. That alleged conduct differs from the information exchange upheld in Maple Flooring, which involved historical data on completed transactions, broad dissemination of reports, and no agreement to coordinate prices or output.
In November 2025, the DOJ announced a proposed settlement requiring RealPage to “cease having its software use competitors’ nonpublic, competitively sensitive information to determine rental prices” and to “remove or redesign features that limited price decreases or aligned pricing between competing users of the software,” among other provisions. The proposed settlement illustrates the distinction between targeting specific practices that can facilitate coordination and categorically prohibiting algorithmic pricing.
The goal for policymakers should be to preserve the potential benefits of pricing algorithms while preventing coordination. As former Assistant Attorney General Makan Delrahim noted in 2018, pricing algorithms are “an important part of the digital economy and can account for great efficiencies that benefit consumers.”
Empirical research underscores why the effects of pricing algorithms require a more nuanced assessment. A 2026 working paper published by economists at the University of Pennsylvania and Boston College found “robust evidence that the algorithm helps building managers set prices that are more responsive to market conditions,” resulting in “lowered rents and increased occupancy compared to non-adopters in the same submarket and building class” during the Great Recession. The authors also found that “prices are more consistent with a model of joint profit maximization among adopters of the same software” and estimate the resulting markup difference among adopters to be $24.90 per month per unit. These findings suggest that pricing algorithms can generate efficiencies.
In summary, the competitive effects of pricing algorithms are highly fact-specific. Regulators should distinguish tools that help businesses respond independently to market conditions from practices that facilitate coordination through nonpublic information sharing, restrictions on price competition, or other anticompetitive conduct. New York’s blanket restriction on algorithmic pricing fails to make that distinction. Rather than prohibiting a technology with legitimate and potentially valuable uses, policymakers should target the specific practices that undermine competition.
