The lawsuit that the U.S. Justice Department (DOJ) filed last August against the real-estate software provider RealPage, as well as a report issued last month by the White House Council of Economic Advisers (CEA) claiming to find $3.8 billion in consumer harms in the rental-housing market arising from the use of algorithmic-pricing tools, both stand as stark examples of the kinds of fundamental misunderstandings of both artificial-intelligence (AI) tools and market dynamics that the incoming administration will need to address.
Initiatives like these, taken in the waning days of the Biden administration, threaten to establish dangerous precedents that could chill innovation, while failing to address genuine challenges in the housing market. Their timing is particularly concerning, as courts and regulators are just beginning to grapple with how to approach AI across multiple sectors.
The RealPage case—to which the DOJ earlier this month added six large landlords as co-defendants—advances the novel and problematic theory that the mere use of common pricing software constitutes tacit collusion among landlords. If accepted more broadly, this approach would cast doubt on the use of a wide range of beneficial technologies. Meanwhile, the CEA report’s simplistic methodology and assumptions about causation provide a shaky foundation for policy interventions in housing markets.
As the Trump administration begins, there is an opportunity to chart a different course that better balances legitimate oversight with the benefits of technological innovation. To do so would require moving beyond speculative theories of harm to focus on actual anticompetitive conduct, while preserving the efficiency gains that sophisticated market analysis can provide. (Indeed, one early sign of potential movement in that direction is that the CEA report appears to have been removed this week from the White House website.)
The incoming administration’s approach to these issues will set important precedents for how we regulate AI and other emerging technologies in the years ahead.
The Biden Administration’s Flawed Approach to AI in Housing Markets
The CEA Report’s Faulty Assumptions
The Biden CEA’s estimate that algorithmic-pricing tools cost renters $3.8 billion in 2023 rested on deeply problematic methodological assumptions. Most notably, the report relied on linear interpolation of demand elasticities—an approach that assumes direct, proportional relationships between software adoption and price changes. This methodology fails to account for crucial real-world complexities like market-saturation points, regional variation in housing supply, and local economic conditions that affect rental prices.
In a nutshell, the report contended that, without these algorithms, landlords would engage in more aggressive price competition. But this premise is deeply flawed. Prior to algorithmic pricing, rental markets were characterized by information asymmetries in which many landlords were forced to rely on incomplete local information. Pricing algorithms, by contrast, serve to democratize market intelligence, providing smaller property owners access to sophisticated pricing strategies previously available only to large and well-resourced corporations.
Consider the traditional alternative. Without algorithmic guidance, most landlords set prices by rudimentary methods: they would compare against a few nearby properties, use outdated historical data, or rely on gut instinct. Algorithms like RealPage introduced data-driven rationality into a historically opaque market.
The CEA report acknowledges this potential benefit, but effectively dismisses it by emphasizing the potential anticompetitive effects. But the council’s own methodology depends on multiple “simplifying assumptions,” as the report concedes that its estimates are merely approximate.
Even more concerning is how the report treats correlation as causation. The analysis assumes that higher rents in markets where there was greater adoption of pricing software must be caused by the software itself, rather than considering that markets with more sophisticated property managers might naturally see both greater penetration of pricing technology and higher rents that are due to underlying market conditions. This basic analytical error undermines the credibility of the report’s conclusions.
DOJ’s Novel Theory of Harm in RealPage
The DOJ’s case against RealPage compounds the analytical errors evident in the CEA report by advancing a novel and problematic theory of harm. The lawsuit essentially argues that, when multiple property managers use the same common third-party pricing software, this constitutes tacit collusion. This theory ignores established antitrust precedent, under which a charge of collusion requires evidence of actual agreements to coordinate pricing, not just parallel conduct or the use of similar business tools.
More fundamentally, the DOJ’s approach reveals a profound misunderstanding of how AI functions in market contexts. Modern AI-pricing tools don’t simply tell all users to charge the same price; they analyze vast troves of market data to help individual property managers make more informed pricing decisions based on their specific circumstances.
The crude methods to which smaller landlords had to resort before the rise of algorithmic pricing created more opportunities for mispricing and market inefficiencies that can harm both property owners and renters. AI tools help to reduce these inefficiencies by providing more accurate market information to all participants.
Lower Prices Aren’t Always Better
Perhaps most troublingly, both the White House report and DOJ case appear to assume that lower prices are always better. This ignores basic economic principles about the role of prices in sending market signals and allocating resources efficiently. In some cases, artificially suppressed rents can actually harm market function by discouraging new housing construction or investments in maintaining existing properties.
These conceptual and methodological flaws matter, because they threaten to establish dangerous precedents for how we regulate emerging technologies. If sharing data with a common third-party software provider is treated as equivalent to collusion, it could chill the development of valuable market-intelligence tools across many industries. Similarly, if regulatory analysis continues to rely on simplistic assumptions about market dynamics, we risk promulgating policies that harm the very consumers they aim to protect.
The Economic Flaws in the DOJ’s Approach
The debate over algorithmic-pricing tools needs to be grounded in how these technologies actually function in real-estate markets.
Modern pricing software typically integrates multiple data streams—including local market conditions, seasonal patterns, property-specific attributes, and broader economic indicators—to suggest optimal pricing strategies. This process mirrors how human property managers have always made pricing decisions, but with greater analytical precision and deeper market insights.
Importantly, these systems don’t dictate prices. They provide recommendations that property managers can accept, modify, or reject, based on their business judgment and local knowledge.
While the DOJ’s complaint has attracted headlines and fed into the public’s worst suspicions about the behavior of landlords, the RealPage case will hinge on real-world data that has not yet been subject to sufficient analysis. For example, the DOJ’s complaint alleges:
This dynamic exists not only in markets with growing demand, but also so-called ‘down markets,’ where demand is decreasing. … AIRM and YieldStar resist price decreases in down markets as much as possible while achieving targeted occupancy rates. RealPage told one prospective AIRM client that the combination of ‘AI and the robust data in the RealPage ecosystem’ would allow the landlord to ‘avoid the race to the bottom in down markets.
In contrast, New York University’s Arpit Gupta notes:
The academic research on RealPage emphasizes that they help the market reach equilibrium faster. That entails rent increases in tight markets; but also faster rent decreases during down markets.
If even facts as fundamental as price behavior in a down market is subject to dispute, then—as they say on TV—“the jury is still out” on RealPage.
The benefits of pricing algorithms to smaller market participants are particularly noteworthy. Previously, only the largest property management companies could afford sophisticated market analysis, maintaining teams of analysts to track market trends and optimize pricing strategies. Thanks to algorithmic pricing, smaller landlords now have access to analysis of similar quality, even as many continue to check their own gut instinct. As the DOJ’s complaint quotes an email from a property owner:
I also rely on comparing available units to adj[usted] leases needed, to forecast leases, to gut check the pricing recs. These data points are always a factor in my pricing decisions.
It seems clear that these tools serve to enhance market efficiency, rather than harm it. When property managers have better information about market conditions, they can more quickly adjust to changes in supply and demand. This responsiveness actually helps to prevent the kind of sustained price increases that can occur when markets adjust slowly to changing conditions. For instance, properties using sophisticated pricing tools can show faster response to seasonal demand shifts, preventing the price spikes that occur when managers are slow to adjust to facts on the ground.
Data-driven pricing tools serve several crucial market functions beyond simple price setting. They can help property managers to better predict the timing of maintenance and renovation based on market conditions; optimize lease lengths and renewal timing; identify emerging market trends that inform investment decisions; and reduce vacancy rates through more accurate pricing.
These functions contribute to more efficient market operation, benefiting both property owners and renters. Lower vacancy rates, for example, allow property managers to spread fixed costs across more units, potentially moderating pricing pressures. Better maintenance timing can help to preserve property quality while managing costs.
It’s also important to consider the alternative. Without sophisticated pricing analysis, property managers often default to crude rules of thumb—like automatic annual increases or simple competitor matching—that can actually create more market distortions than data-driven approaches. Modern pricing tools can help managers make more nuanced decisions that better reflect true market conditions.
When both large and small property managers have access to sophisticated market analysis, competition can focus more on genuine differentiation in property quality and service, rather than the information asymmetries that historically redound to the benefit of the largest competitors. This shift ultimately benefits renters by fostering more transparent and efficient markets.
The Legal Flaws in DOJ’s Approach
The theory of harm put forward by the DOJ in the RealPage case risks fundamentally reshaping antitrust enforcement in the digital age. If courts were to treat the use of common pricing software as tantamount to collusion, that precedent could chill innovation far beyond the real-estate sector.
Information sharing can enhance market efficiency by helping firms to make better-informed independent decisions. Collusion, by contrast, requires an agreement—explicit or tacit—to coordinate pricing or other competitive behavior. The DOJ’s RealPage complaint effectively collapses this distinction.
Section 1 of the Sherman Antitrust Act requires evidence of an agreement to restrain trade—not merely parallel conduct or the use of similar business tools. Courts have consistently held that the mere exchange of market information, without more, does not constitute an antitrust violation. For example, in 1925’s Maple Flooring v. United States, the U.S. Supreme Court recognized that market participants can share industry statistics and other information through trade associations without running afoul of antitrust law.
The Court also made clear in 1975’s United States v. Citizens & Southern Nat’l Bank that “the dissemination of price information is not itself a per se violation of the Sherman Act.” In fact, as noted in 1978’s United States v. United States Gypsum Co., the “exchange of price data and other information among competitors does not invariably have anticompetitive effects; indeed such practices in certain circumstances increase economic efficiency and render markets more, rather than less, competitive.”
The standards for evaluating information-sharing cases are well-established. Courts consider factors like:
- Whether there is evidence that the parallel conduct is the result of an agreement, rather than “chance, coincidence, independent responses to common stimuli, or mere interdependence unaided by an advance understanding among the parties”—Bell Atlantic Corp. v. Twombly (2007);
- Whether alleged price increases are likely the result of parallel pricing, rather than “parallel, market-following behavior”—In re Commodity Exchange, Inc. Gold Futures & Options Trading Litig. (S.D. N.Y. 2018);
- Whether the information exchanged is confidential pricing information that would lend itself to collusion; and
- Whether the defendants have market power that would allow them to effectuate a conspiracy.
Here, the DOJ hasn’t shown there was any agreement among competitors; the allegation is simply that they all used RealPage’s algorithm. But the RealPage software analyzes historical and current market data—not the future pricing intentions of rental property owners. Moreover, the DOJ would have a difficult time proving that rent increases are due to use of the algorithm, rather than generalized inflation, restrictions on building more housing, and other supply and demand factors specific to particular housing markets.
Landlords typically operate in markets with numerous competitors and low levels of concentration. And they largely process information that would be publicly available—albeit more difficult to compile and analyze—without sophisticated software. Perfect competition would require market participants to have good information about supply, demand, and prices. Tools that help market participants to better understand these fundamentals therefore serve to enhance rather than harm competition.
The DOJ’s theory also ignores the important economic distinction between price discovery and price fixing. Algorithmic-pricing tools help firms discover market-clearing prices by analyzing vast troves of data. This is fundamentally different from agreements to fix prices at artificial levels. Indeed, more accurate price discovery typically leads to more competitive markets, not less.
If adopted by courts, the DOJ’s theory could have far-reaching consequences in any situation where competitors use common third-party tools to analyze market data, as this could potentially be characterized as unlawful information sharing. It would also create significant uncertainty for businesses and likely deter the adoption of efficiency-enhancing technologies.
A more sound approach would focus on traditional antitrust analysis: examining whether specific uses of technology actually harm competition through agreements to restrain trade. This would allow courts to address genuine competitive concerns, while preserving the benefits of technological innovation in market analysis and pricing.
A Better Path Forward: Recommendations for the New Administration
As the new Trump administration wades into these issues, there is a crucial opportunity to course correct on AI regulation in housing markets and beyond. The flaws in the Biden administration’s approach to algorithmic-pricing tools highlight the need for a more nuanced, evidence-based framework that better balances innovation and competition concerns.
Two concrete steps would help to chart a better course. First, the DOJ should reassess its position in the RealPage litigation, and abandon the novel theory of harm the complaint seeks to advance. Second, regulators should develop clear guidelines to evaluate the competitive effects of information-sharing technologies. Such guidelines should recognize that tools that enhance market transparency and price discovery typically promote competition. They should also acknowledge that the mere use of common third-party software does not constitute unlawful coordination.
Competition policy should focus on actual market problems, rather than speculative harms. In housing markets, this means addressing structural barriers to supply like restrictive zoning and permitting delays. These factors, not pricing technology, are the primary drivers of housing-affordability challenges.
The new administration has the opportunity to demonstrate how thoughtful regulation can promote both innovation and competition. This would require moving toward a more sophisticated understanding of how modern analytical tools can enhance market efficiency. The alternative—continuing down the path of treating sophisticated market analysis as inherently suspect—risks sacrificing significant economic benefits in pursuit of phantom concerns. The time is right to reset this balance and establish a more coherent framework for evaluating competition in technology-enabled markets.
