Airbnb May Affect Rents But New York City Study Doesn't Tell Us

By Adam Millsap of Mercatus.

"A new study from the New York City Comptroller’s office claims that Airbnb has contributed to higher rents in the city. The study has been picked up by several new outlets who have repeated this claim, even though the study doesn’t establish a causal relationship between more Airbnb units and higher average rents.

The comptroller office’s study uses Airbnb listing data to calculate the number of Airbnb units for 55 neighborhoods in New York City. The key variable of interest is the proportion of all housing units in a neighborhood that are listed on Airbnb. The authors also control for a bunch of other neighborhood characteristics using data from the American Community Survey. Using this data, they estimate the relationship between changes in average monthly rent and changes in the proportion of Airbnb units in a neighborhood by year from 2009 to 2016.

They find that average rent and the proportion of Airbnb units in a neighborhood move together, which is not particularly surprising. What they don’t find, despite claims to the contrary, is that a higher proportion of Airbnb units causes average rents to increase.


The comptroller office’s theory is that a higher proportion of Airbnb units available for overnight stays means there are less units available for long-term renters, and this decrease in supply of long-term units leads to higher rents. This theory is consistent with basic supply and demand, but estimating the effect statistically is tricky since the relationship between prices and quantities of related products runs both ways.

For example, a decrease in supply of long-term rental units due to an increase in the supply of Airbnb units would increase average rents if nothing else changed. Alternatively, an increase in rents could induce more people to rent out their apartments—or extra rooms in their apartments—via Airbnb in order to take advantage of the higher prices. In fact, in the latter scenario some renters may rent out extra rooms to earn additional income to offset their higher rent.

In the first case, more Airbnb units would be causing rents to increase. But in the second case, higher rents due to some other factor would be inducing more people to rent out their units on Airbnb. Both stories result in higher rents and more Airbnb units, but the causes are completely different. If all we see is that average rent rises along with the number of Airbnb units, we have no idea which is causing which.


Statistically estimating the effects of prices on quantities and vice versa is difficult because prices and quantities are determined simultaneously, meaning they each cause the other. Economists have come up with advanced techniques to deal with this issue, but they aren’t used in the comptroller’s study.
Because of this, all this new study can say is that average rents and the proportion of Airbnb units in New York City neighborhoods were positively related to each other from 2009 to 2016. The study can’t say which one caused the other and doing so is misleading."