Is There an Economic Justification for the Attacks on Rent Algorithms?

By Jay Ezrielev[i]

Government agencies and private plaintiffs have a novel explanation for high rents: pricing algorithms embedded in rental revenue management software. This software makes individualized rent recommendations to landlords and rental property managers (collectively “landlords”). RealPage, the largest rental revenue management software company, has been a target of several recent antitrust lawsuits claiming that the company’s software facilitates rent price-fixing. Private plaintiffs, the U.S. Department of Justice (“DOJ”), and state Attorneys General have sued RealPage and its clients under a novel antitrust theory centered on pricing algorithms and pooling of competitor data. The plaintiffs allege that RealPage’s clients have conspired to share their nonpublic data with RealPage’s software, and that the software facilitates this conspiracy by using the data to recommend supracompetitive rents.

However, the plaintiffs offer no economic evidence to indicate that revenue management software or the aggregation of nonpublic data caused higher rents. Instead, they have simply asked the courts to assume that the use of pooled competitor nonpublic data is inherently anticompetitive, while providing no economic basis for this proposition. Our economy benefits tremendously from analyzing aggregated nonpublic data. This includes analyzing nonpublic data to gain a better understanding of local market conditions. There are significant efficiencies in discovering pricing that effectively balances supply and demand in rental markets. This is what revenue management software does. Landlords’ use of this software is likely to increase rental occupancy, lower average costs, and enhance incentives to develop rental housing. Indeed, as I discuss below, rather than preventing supracompetitive rents, limiting the use of revenue management software may lead to an increase in rents.

I. Antitrust Assault on Rent Algorithms

Back in July of 2024, the Biden administration accused landlords of using revenue management software to fix rents. Then, in August of 2024, the DOJ along with eight states sued RealPage for operating an “unlawful scheme to decrease competition among landlords in apartment pricing.” In describing the case Merrick Garland, the Attorney General under President Biden, said that “everybody knows the rent is too damn high and we allege this is one of the reasons why.” Since then, the DOJ filed an amended complaint that added two states as plaintiffs and landlords as defendants. The Attorneys General of Arizona, District of Columbia, and Maryland have brought similar lawsuits.

Private class action lawsuits against RealPage and Yardi, another rental revenue management company, have made similar allegations. Courts took different approaches in their rulings on the motions to dismiss in the two class actions. In In re: RealPage, Inc. Rental Software Antitrust Litigation (No. II), the U.S. district court for the Middle District of Tennessee found that the plaintiffs’ allegations do not warrant the per se standard but allowed the case to proceed under the rule of reason. However, in Duffy v. Yardi Sys., Inc., the U.S. district court for the Western District of Washington ruled that the plaintiffs’ allegations justify the per se standard. Prior to its lawsuit against RealPage, the DOJ filed a Statement of Interest supporting the per se standard in the RealPage class action litigation. The DOJ and the U.S. Federal Trade Commission also jointly filed a Statement of Interest in support of the Yardi class action.

In addition to the antitrust lawsuits against RealPage and Yardi, San Francisco, Philadelphia, and Minneapolis have adopted ordinances prohibiting the use of rent algorithms. Others are considering similar prohibitions, including the states of New York and Colorado. There is also proposed federal legislation to ban rent algorithms. The prohibitions focus on algorithms’ use of nonpublic competitor rental data.

II. A Little Perspective

Considering the broad campaign to ban rent algorithms, it is remarkable how little evidence links algorithms to higher rents. It is important to have some perspective. Between Q4 2011 and Q4 2024, home values grew twice as fast as rents. (See Figure 1.) During this period, rents grew about 1.2 percent per year in real (inflation adjusted) terms, suggesting that rents are not the primary driver of high housing costs. Since 2011, there has also been a significant rise in rental occupancy. According to the U.S. Census Bureau data, the 5 year average U.S. rental occupancy rate increased from 90.0 percent in Q4 2011 to 93.7 percent in Q4 2024. (See Figure 2.) The increase in the rental occupancy rate is tantamount to an expansion of rental output, where the 3.7 percentage point increase in the rental occupancy rate translates to 1.8 million additional rentals (based on 49 million total rental units in 2024). Rising rental occupancy also reduces landlords’ average cost per rental. Higher rental output and lower costs both imply lower rents.

What caused the rise in the rental occupancy rate? Occupancy reflects a balance between supply and demand in rental markets, and rents play a critical role in maintaining that balance. Rents that are too high lead to excess supply and low occupancy rates, while rents that are too low lead to shortages, making it difficult for renters to find available units. Low rents also weaken incentives to undertake improvements and develop new rental housing.

There are several reasons why rents may deviate from the efficient levels. Landlords may overprice or underprice rents because they misjudge demand. Revenue management software seeks to reduce such pricing errors, which in turn improves market function and enhances efficiency, benefiting both renters and landlords. Indeed, revenue management software may have contributed to the rising occupancy rates by helping landlords avoid overpricing, thus reducing vacancies.

Furthermore, rents may deviate from efficient levels because of rent control regulations. Such regulations may create shortages and weaken incentives to invest in rental housing. Landlords may also set supracompetitive rents through coordinated action. This is what the DOJ and states allege in their complaint against RealPage. Supracompetitive rents, however, lead to excess supply and falling occupancy rates, the opposite of what we observe in rental markets. 

Figure 1.

Figure 2.

III. What is Wrong With the DOJ Complaint

Notwithstanding a lack of empirical evidence, the DOJ complaint alleges that RealPage’s revenue management software caused supracompetitive rents by harming the “competitive process.” The DOJ alleges that the software harmed the “competitive process” by collecting competitively sensitive, nonpublic data from RealPage’s clients to give landlords an informational advantage vis-à-vis renters. According to the DOJ complaint, RealPage and its clients engaged in “unlawful information-sharing” that enabled the clients to use “competitors’ nonpublic data to predict with more certainty the highest price that the market will bear for a particular unit.” However, contrary to the complaint’s assertions, having more information about demand when deciding what to charge for rent benefits renters.

Information about rental demand is critical for landlords because of the significant risk they face in setting rents. A rental unit can rent out instantly, or it can stay vacant for many months, depending on the asking rent. For landlords, determining what to charge for rent is a guessing game, and they often get it wrong by either overpricing or underpricing. Revenue management software uses data to help landlords minimize such errors. Correcting underpricing errors means increasing rents. The DOJ complaint focuses on this effect, but it is only half the story because correcting overpricing means lowering rents.

Underpricing and overpricing corrections have different implications for rental output. Raising rents to correct underpricing would not significantly reduce the number of rentals because underpricing occurs when landlords mistakenly believe that increasing rents will cause a significant loss of tenants. Thus, underpricing occurs when demand is inelastic. On the other hand, lowering rents to correct overpricing would significantly increase the number of rentals because overpricing occurs when landlords mistakenly believe that increasing rents will not cause a significant loss of tenants. Thus, overpricing occurs when demand is elastic. Overall, minimizing pricing errors is likely to increase both the number of rentals and the occupancy rate. The rise in the average occupancy rate that we observe in rental markets is consistent with increasing use of revenue management software to correct pricing errors.

Correcting pricing errors also increases landlords’ return on investment, thus enhancing the incentives to develop rental housing. Overall, greater use of revenue management software is likely to increase rental occupancy, lower average costs, and enhance incentives to develop rental housing. All these effects are likely to reduce average rents.

The DOJ complaint also raises the concern that RealPage’s revenue management software facilitates rent collusion. However, successful collusion requires a mechanism to prevent deviations from collusion. If enough participants deviate, the collusion unravels. RealPage’s revenue management software lacks such a mechanism. Landlords can simply “deviate” by rejecting recommendations. The complaint does not allege an agreement to accept recommendations. The software does not reveal recommendations or acceptance decisions to other landlords. Indeed, according to the DOJ complaint, RealPage’s revenue management software users reject recommendations more than 50 percent of the time.

We should also be skeptical of a nationwide price-fixing conspiracy involving RealPage’s revenue management software because only a small percentage of rental units use this software. RealPage offers three revenue management software packages: AI Revenue Management (“AIRM”), YieldStar, and Lease Rent Options (“LRO”). The DOJ complaint claims that only AIRM and YieldStar violate antitrust law. The complaint does not claim that LRO is illegal presumably because the complaint does not allege that LRO pools competitively sensitive, nonpublic data across competing landlords. According to RealPage, in 2023, fewer than 7 percent of all U.S. rental units used AIRM or YieldStar and fewer than 4 percent used LRO.

IV. The Evidence?

Scrutiny of RealPage’s revenue management software began following the publication of a ProPublica article in October 2022. The article presented no evidence linking the software to higher rents but merely provided anecdotes of high rent increases during the 2022 inflationary period when other sectors of the economy experienced comparable price increases. In December 2024, the Biden Administration White House Council of Economic Advisers put out an analysis claiming that landlords’ use of RealPage’s revenue management software increased rents by $3.8 billion in 2023. However, this analysis is highly problematic because it relies on assumptions that lack empirical support.

Likewise, a recent unpublished study by Sophie Calder-Wang and Gi Heung Kim that found that RealPage’s revenue management software is associated with higher rents. The study found that using this software increases rents on average by $25 per unit per month. However, the study’s significant data and methodological flaws render it unreliable. For example, the study uses a flawed methodology for identifying landlords’ adoption of revenue management software. The study relies on a 2011 survey of landlords, public marketing announcements, and Google searches to determine whether a landlord is using revenue management software. If the study identifies a landlord as a user of the software at a point in time, the study assumes that the landlord uses revenue management for all its rental units in all subsequent periods. There is no empirical basis for this assumption. Moreover, the study’s reliance on a static competition model ignores important dynamic effects of revenue management. It is worth noting that the Calder-Wang Kim study did not find a significant difference in rent premiums between AIRM/YieldStar and LRO, even though the DOJ complaint does not allege that LRO violates antitrust law. Overall, there is a lack of reliable evidence of any harm to renters from RealPage’s revenue management software.

V. Conclusion

Banning revenue management software or the use of pooled nonpublic data in rent algorithms will not make rental housing more affordable but may raise rents instead. Addressing housing affordability calls for a serious conversation about policy, and blaming algorithms is a distraction from that conversation.

Click here for a PDF version of this article

[i] Jay Ezrielev is the Founder of Elevecon.