News Release

Study finds Airbnb units expand market but reduce long-term rentals, including affordable housing

Authors suggest new tax to alleviate resulting inequities

Peer-Reviewed Publication

Carnegie Mellon University

Marketplaces for short-term accommodations have emerged as a way for landlords to promote their properties to short-term renters. This can lead some landlords to switch from long-term rentals and affect the supply and affordability of rental housing. Despite recent government regulations to address this concern, it is unclear how many and what types of properties are switching. A new study combined data from Airbnb—the most popular platform for short-term rentals—and the U.S. Census to estimate a structural model of property owners’ decisions and evaluate relevant regulations. The study found that the presence of Airbnb units in a community caused a mild decrease in the long-term rental supply (i.e., switchers), including affordable housing, which harmed local renters. But it also expanded the rental housing market, which may benefit low-income landlords. The study’s authors propose a new tax to reduce switching while maintain rental market expansion, and to alleviate socially inequitable outcomes.

The study, by researchers at Carnegie Mellon University (CMU) and LG CNS (an information technology consulting firm), is forthcoming in Management Science.

City regulators have launched various policies regarding short-term rentals, especially in cities where affordable housing is a concern. Some regulations limit the number of days a property can be listed, while others charge a transient occupancy tax on the listing price, similar to a hotel occupancy tax. By 2020, many U.S. cities had imposed regulations on Airbnb, but it is unclear how this platform and these changes have affected the rental housing market.

“The benefits of renting for the landlords can be directly observed from the prices and occupancy rates in the long-term market and on Airbnb,” notes Kannan Srinivasan, Professor of Management, Marketing, and Information Systems at CMU’s Tepper School of Business, who coauthored the study. “But the costs of renting and how they differ by demographics, properties, and cities are unknown. Our study identified the underlying renting costs, including both tangible and intangible costs.”

The study estimated a structural model of property owners’ hosting decisions (e.g., whether to rent on the long-term rental market or Airbnb, and if opting for Airbnb, how many days to rent). They used data from two sources to construct a comprehensive list of potentially available properties in selected areas: 1) every property listed on Airbnb in nine representative metropolitan areas in 2015 and 2017, collected by AirDNA, a third-party company specializing in data collection and analysis, and 2) the 2015 and 2017 American Housing Survey, a comprehensive, longitudinal, national housing survey administered by the U.S. Census Bureau.

Examining three sets of covariates—property characteristics, host demographics, and

market characteristics—the researchers modeled hosts’ revenue-cost tradeoffs to identify landlords who switched to short-term rentals. By constructing a structural model, they simulated a counterfactual scenario without Airbnb and compared it with the scenario when Airbnb was present. Their model also allowed them to evaluate the effectiveness of rental regulations.

            The study found that the presence of Airbnb in a community slightly reduced the long-term rental supply but also created a market expansion effect, with outcomes varying across metropolitan areas. Cities where Airbnb is more popular (e.g., Miami, New York, San Francisco) experienced a larger reduction of the local rental supply but did not necessarily have a larger percentage of landlords who switched from long-term rentals to Airbnb rentals.

            Affordable housing was the major source of both negative and positive impacts of Airbnb: The presence of Airbnb units caused a larger reduction in rental supply, especially among affordable units, which harmed local renters. But Airbnb units also created a larger market expansion effect for affordable housing, which benefited local hosts who owned affordable units and may have been less economically advantaged. The study concludes that policymakers need to strike a balance between local renters’ affordable housing concerns and local hosts’ income needs.

            After assessing the most commonly used regulations, the study’s authors suggest that imposing a linear tax on Airbnb landlords is more desirable than limiting the number of days a property can be listed. In particular, they propose a new convex tax that imposes a higher tax on expensive units while leaving less expensive units less taxed. The new tax could reduce switching while maintaining market expansion, as well as alleviate social inequality. In practice, there have been continuing concerns that Airbnb has raised concerns that it exacerbates income disparity since the platform’s gains are disproportionately skewed to those with more wealth. The new tax could reduce the fraction of total host profits earned by economically advantaged hosts and help maintain social equality.

“To the best of our knowledge, this is the first study to systematically and formally model hosts’ decisions and recover the underlying tradeoffs,” says Hui Li, Associate Professor of Marketing at CMU’s Tepper School of Business, who led the study. “This framework allowed us to conduct counterfactual analyses to identify actual switchers and examine policy impacts.”

            Among the study’s limitations, the authors note that the measures they examined do not capture every aspect of policy effects and recommend that other potential effects (e.g., effects on renters, long-term effects on new home purchases and construction) should be considered in future studies.


Summarized from an article in Management Science, Market Shifts in the Sharing Economy: The Impact of Airbnb on Housing Rentals by Li, H (Carnegie Mellon University), Kim, Y (LG CNS), and Srinivasan, K (Carnegie Mellon University). Copyright 2021. All rights reserved.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.