Rent-Setting Algorithms Could Exacerbate Rising Rents: Report

The software uses an algorithm, with inputs from a variety of market-specific data sets, to price apartment rents in order to remove the guesswork from the process. But subjects interviewed by ProPublica believe the algorithm could be artificially inflating rents and recommending hikes that most landlords wouldn’t attempt otherwise.

“As a property manager, very few of us would be willing to actually raise rents double digits within a single month by doing it manually,” Andrew Bowen, vice president corporate sales for RealPage, was quoted as saying at a Nashville, Tenn., convention last year.

The data firm, however, disputes the claims of the investigation.

“RealPage is aware of the article. It is misleading and inaccurate, and we disagree with its conclusions,” RealPage said in a statement to Commercial Observer. “RealPage completely agrees there is a decades-long severe affordable housing shortage. … Neither RealPage nor any housing providers can move markets up or down. Each asset (bedroom type actually) is optimized separately without considering any information about the inventory availability at other communities or the strategies employed by other owners.”

One client of RealPage is Greystar, one of the largest landlords in the country, which owns dozens of multifamily assets across the tri-state region alone. The firm uses the software to price available units and outperform competitors in its markets by up to 4.8 percent, according to ProPublica.

Greystar did not immediately respond to a request for comment.

A fact sheet from RealPage says that the software was developed in the early 2000s and is not dissimilar to price-setting software used by airlines, hotels, grocery stores and others. YieldStar was, after all, developed in part by Jeffrey Roper, a former Alaska Airlines director of revenue management.

RealPage believes that it is not only helping its clients, but also helping renters who can see rent prices at other apartments in an area and, as “consumers,” make appropriate financial decisions. The algorithm also provides pricing based on lease terms and move-in dates.

The goal of the software, according to the company, is not “rent maximization” but “revenue optimization.” What’s the difference? The latter finds a balance between filling vacant units at a price that is both profitable for landlords and acceptable to tenants, the RealPage fact sheet says.

The software doesn’t always seem right to the humans seeking its advice — about 10 to 20 percent of property managers reject the system’s recommendations. 

One key difference between the human and algorithmic approaches is that landlords tend to try to minimize vacancy, assuming that turnover and long vacancies are detrimental. With the help of the algorithm, however, some landlords have found that keeping the building emptier but rents higher brings in higher returns — which, of course, works against the tenant. 

One firm with up to 40,000 units under its watch found that while operating at 97 to 98 percent occupancy was considered ideal before YieldStar, it could actually raise rents if it kept occupancy closer to 94 percent.

Another concern is that if a critical mass of landlords are using these algorithms, that may be akin to collusion or price-setting and could trigger antitrust issues. In fact, the Department of Justice got involved when RealPage purchased a competitor in 2017, but ultimately allowed the acquisition to go through.  

“Within any given market, revenue management works independently at each community, so some communities may see decreases and others increases,” RealPage said in its statement. “To be clear, RealPage Revenue Management does not consider or have visibility into competitor availability.”

Realpage is currently used by landlords to price 19.5 million apartments in the country, according to ProPublica.

Mark Hallum can be reached at mhallum@commercialobserver.com.