Policymakers Shouldn’t Assume Algorithmic Pricing Is Anti-competitive

Businesses in many industries use algorithms to optimize the prices they set for their products and services. Some use algorithms to set prices dynamically, changing prices in response to shifting supply and demand, and others use algorithms to offer personalized pricing, setting prices based on what they expect an individual consumer is willing to pay. Unfortunately, some policymakers oppose algorithmic pricing, even though using data to automate pricing has the potential to optimize economic transactions.

Companies have long used dynamic pricing, from lowering prices during “happy hours” to charging more for making a long-distance call at peak hours, to reflect market conditions and consumer preferences. For example, airlines increase the price of a plane ticket around holidays in response to lowered supply or increased demand. During less popular travel times, the price for the same ticket falls. Gas stations use dynamic pricing to match or undercut nearby competitors. Hotels follow a market-sensitive pricing strategy—as more rooms are filled, the price of each subsequent room may increase. And car dealers notoriously use personalized pricing, with salespersons offering different prices to different customers as they haggle for the best.

Today, that process can be done automatically due to technology. Algorithmic pricing is a data-driven pricing strategy that automatically adjusts prices to meet market conditions. Prices are adjusted automatically to reflect demand, competitor pricing, inventory levels, and even consumer behavior. For example, if a competitor lowers its price for a specific product online, a retailer’s pricing algorithm might automatically adjust the price of the same or a similar product to match or undercut the competitor’s price. If the competitor’s price changes again, the algorithm will automatically readjust to optimize the price for the retailer. This dynamic pricing allows businesses to provide consumers with competitive pricing and maximize profits.

But some policymakers have suggested that algorithmic pricing limits price competition. In a recent hearing of the Judiciary Subcommittee on Competition Policy, Antitrust, and Consumer Rights, U.S. Senator Amy Klobuchar (D-MN) remarked that “[in the rental housing market] we have seen the widespread use of algorithmic pricing tools designed to raise prices even at the expense of higher vacancy rates.” She further stated, “I believe landlords should be competing on price. And I don’t think you see that happening when you have these algorithm-based games going on.” However, an algorithm suggesting higher prices at the expense of empty units doesn’t prevent competition based on price, even if the resulting prices are higher. The algorithm just optimizes the pricing strategy.

Sen. Klobuchar’s remarks echo comments made by Sen. Sherrod Brown (D-OH) in a letter he sent to Federal Trade Commission Chair Lina Khan last year calling for regulators to investigate “whether rent setting algorithms that analyze rent prices through the use of competitors’ private data…violate antitrust.” However, Sen. Brown’s concern is improperly aimed at algorithms. If competitors use a third-party company to collude—by improperly using competitively sensitive information from rival firms to set prices—that would violate antitrust law. However, this illegal activity could happen regardless of whether firms use an algorithm.

Economists largely agree that price adjustment algorithms based on demand and costs increase efficiency, and these strategies can benefit both sellers and buyers. Sellers can respond to market fluctuations, competitors’ pricing strategies, and consumer preferences or behaviors in real time. They also enjoy increased productivity due to automation, with more price consistency and accuracy. These benefits extend to consumers, who receive competitive prices on a larger number of products and personalized shopping experiences.

While some scholars suggest that, in some cases, dynamic pricing strategies may eventually create higher prices, using an algorithm to automate the process of determining prices is no more anti-competitive than using human labor to adjust prices. Policymakers should be careful to avoid broadly labeling the use of algorithmic pricing as anti-competitive and should instead investigate any potential antitrust violations to determine if anti-competitive or illegal actions occurred. Using algorithms to automate the process increases efficiency, creating savings that can be passed on to consumers.

If policymakers have concerns about algorithmic pricing, they should work with the private sector to establish voluntary best practices. For example, they can develop safeguards to mitigate the potential for collusion, such as rules about sharing algorithms and data with competitors, and create best practices on how to disclose to consumers when they use algorithmic pricing. They can also pursue more studies about the impact of algorithmic pricing on consumers in different markets.

While safeguarding fair competition and consumer protection is vital, it is important that policymakers do not demonize algorithmic pricing and allow firms to use algorithms to automatically adjust prices in response to changing market conditions.