Artificial intelligence has been blamed for job cuts, freaky art and plagiarism. But allegations that algorithms illegally raised rents cannot be true, a report by industry researchers argues.
In October, ProPublica reported that software from RealPage was driving up apartment rents across the country. Lawsuits followed, claiming that the software — which suggests rents for landlords by predicting availability and demand — had essentially created a cartel of multifamily operators. The Department of Justice launched an investigation.
Now, a group of industry insiders, two of them quoted in the ProPublica article, are pushing back.
Revenue management software pioneers Jeffrey Roper and Donald Davidoff — both of whom currently serve as CEOs of software firms that specialize in the multifamily sector of commercial real estate — joined with industry consultant Dom Beveridge on a report rebutting the nonprofit news site’s story.
“Not only is it untrue that there is a multifamily cartel of any kind, it isn’t even possible,” the authors write.
In the ProPublica article, Davidoff and Roper — who served as chief scientist for RealPage from 2004 to 2012 — discuss the history of revenue management software. But they claim the story excluded their explanations of how the technology works were excluded, and now want to set the record straight.
Airlines, hotel owners, rental car companies and rail operators have used revenue management software since the 1980s. Apartment owners have used it since at least the 1990s, but the article has stoked fresh anger among tenants and regulators.
They are concerned that when the technology is applied to apartments, it allows landlords to share rent information and boost rents above natural levels, fueling the country’s affordability crisis.
Several property managers told ProPublica that the software led them to seek higher rates than they would have without it. In the aftermath, at least two class-action lawsuits were filed against Richardson, Texas–based RealPage.
The new report argues that the software is designed to encourage competition, not suppress it, and that the allegations “provide an unhelpful distraction from the much more pressing matter of housing affordability.”
In a statement, RealPage noted that it played no role in the whitepaper and that “we applaud the thorough work it represents and hope it will be a catalyst for similarly constructive analysis and discussion that will rebut the many inaccuracies and distorted narratives about our industry.”
A spokesperson for ProPublica said the outlet stands by its reporting.
The rebuttal begins by examining how a cartel can occur and whether the multifamily market could sustain one. Supply of a product, in this case rentals, needs to be consolidated among relatively few competitors with common financial objectives and a way to coordinate pricing decisions.
ProPublica had looked at Belltown, a trendy Seattle neighborhood with just over 9,000 market-rate apartments. The publication found that 70 percent of the apartments are managed by 10 management companies.
“There’s an ‘emperor has no clothes’ thing about reading that statistic,” Beveridge said. “Think about grocery stores, or car dealerships, or any other kind of local service. Can you think of anywhere where 70 percent of the supply comes from some number greater than 10 different companies?”
ProPublica reported that at one Belltown building using RealPage’s YieldStar revenue management software, rents have increased 42 percent since 2012, compared with 33 percent at similar downtown buildings. (ProPublica did not disclose if any of those buildings used YieldStar.)
Beveridge said studying Belltown as a distinct apartment market doesn’t accurately reflect how renters search for apartments.
“You’re not only looking on the same block,” he said. “You’re probably looking at multiple different neighborhoods quite far away from each other when you’re considering moving.”
Nationally, ProPublica found consolidation: The share of apartments in the U.S. controlled by the 50 largest property managers has grown annually for the last 14 years. By 2021, they oversaw about 4.2 million units.
The rebuttal attacked that argument by noting that the country’s largest apartment owners — as opposed to managers — had seen market share decline 17 percent in a decade to about 2.4 million units. Meanwhile, the top property managers increased their share by 58 percent.
ProPublica called managers more relevant to the issue.
“Our reporting focused on property managers — not property owners — because managers are the ones who are using revenue software to price apartments, and they have indeed become more consolidated over time,” the ProPublica spokesperson said. “Renters are simply encountering fewer apartment operators when looking for a place to live in some urban markets.”
A former antitrust prosecutor told ProPublica that RealPage’s online forum and invitation-only meetings “could raise an antitrust red flag.”
Still, the rebuttal makes the case that multifamily operators, with control over about 20 percent of the apartment market, are too fragmented to be a cartel.
It also notes that financial objectives vary among rental properties. Some owners need to fill apartments fast, even if it means cutting rents. Others have the time to wait.
“There’s no scenario at all where somebody gets to price high all the time,” Beveridge said. “This idea that there’s some causality of revenue management that’s making the prices go up is absolutely untrue.”
A new project in lease-up needs cash flow and activity, so owners aim to sign renters quickly, even if it means offering concessions. A value-add project, on the other hand, may be more willing to wait for higher-paying renters to show investors that the renovations yielded returns.
If owners and operators have different objectives, the rebuttal reasons, they should have different definitions of an ideal price.
The authors focus on “exposure,” or the amount of vacant or soon-to-be-vacant units.
“Competitors within the same submarket usually experience different levels of exposure at a given point in time,” the authors write. “Because exposure varies a lot, it makes no sense for competing properties to follow the same pricing strategies.”
With all of the trade-offs and conflicting variables governing pricing dynamics, no revenue management software can create a single pricing regime, the authors write.
Opportunity to coordinate
The question being decided in court and at the Department of Justice is whether landlords and operators using the same software, which crunches data from other buildings when suggesting prices, constitute a cartel.
It is unclear that using the software means landlords agreed to manipulate pricing. In fighting the lawsuits, lawyers will also note that landlords shared pricing information long before algorithms came along.
In the end, the report’s authors call for more housing construction and mock the notion that landlords should charge less than tenants are willing to pay.
“It is tempting to blame landlords for housing affordability,” they write. “But it is fundamentally disingenuous to suggest that the solution to a deeply structural issue like housing affordability is to demand greater altruism from investors.”