Algorithms don’t raise rents, landlords do

Opinion editor’s note: Strib Voices publishes a mix of national and local commentaries online and in print each day. To contribute, click here.

•••

A friend on dating apps recently told me she is often presented with men who are lying about their age — or rather, as they tell it, being forced to lie. The culprit, they are eager to explain, is the algorithm. “I am actually 57,” reads one profile she showed me, “but lowered my age because of the algos, LOL.”

What’s disturbing about her story is not that so many men are pretending to be younger than they are (what else is new?), but that they don’t seem to know what an algorithm is. Their problem is not a sinister program preventing them from finding a younger woman. It is that many women set an age range they don’t fall into. The algorithms are just doing their job, which is optimizing to find a solution for a given set of data.

Yet algorithms have become the villain of this technological era — blamed for depriving us of love, manipulating us on social media and increasing our rent. Democratic presidential candidate Kamala Harris has issued a plan to address that last problem, proposing to crack down on real estate companies that use “algorithmic price fixing” software to raise rents.

Algorithms are used throughout the economy to set prices and manage supply. Lowering the cost of housing is a major objective of the Harris campaign, so it not surprising that it is focusing on their use in real estate. It is far from clear, however, that algorithms are to blame for rising rents.

About one-third of rental units are priced using rent-setting software that is powered by algorithms that make use of lots of data from the local and comparable markets. The alternative — that is, landlords setting rents by looking at listings of comparable units and talking to brokers — is less precise.

According to a study from two professors at the University of Pennsylvania’s Wharton School, management companies that use pricing software tend to be more responsive to market conditions: They raise rents more when the market booms, but also cut rents more and (and have fewer vacancies) when it falls. This suggests they are not colluding so much as maximizing profits.

At the same time, the study shows that landlords who use pricing software affect the behavior of landlords who don’t. In markets where more landlords use pricing algorithms, occupancy rates were lower and rents were 3% higher.

Harris supports a bill that would prohibit landlords from using the services of any company that coordinates pricing and supply information. It defines coordination as collecting data from multiple landlords, calculating a price based on that data with a “process that uses computation,” and recommending those lease terms. This would all be illegal if the lease terms were in violation of the Sherman Antitrust Act. But it is easy to foresee a situation where a landlord has to worry about being sued by his tenants anytime he uses data and a computer program to decide what rent to charge.

What the bill calls “consciously parallel pricing coordination,” most economists would call “using technology to set prices more efficiently.” The bill effectively bars landlords from using the best tools and available data to discover the right price for their units. This would indeed make landlords less profitable. But it would also make them less inclined to build more, or to lower rents when times are hard.

Besides, the main reason why rents increased was not collusion — it was a market short on supply, and that raises prices and leaves buyers with less market power. The way to restore power to tenants is not to outlaw the use of data. It is to encourage more building so consumers have more choice.

Of course algorithms can be used for nefarious purposes, in real estate or any market, and when people use them to collude or manipulate prices, they should be prosecuted. But it is not clear this has happened in real estate. Algorithms are often vilified, but they are a valuable innovation that can help markets become more efficient and offer more choice.

In either the housing market or the dating market, algorithms are not to blame. All they do is bring forth the underlying data that reflects a mismatch between demand and supply — whether the product is reasonably priced apartments or honest bachelors.