A proliferation of Airbnbs, or similar short-term rentals, in a neighborhood contributes to higher rates of crime in the area, according to a new study by two Northeastern researchers.

The relationship is likely because the highly transient housing “pokes holes in the social fabric of the neighborhood,” says Dan O’Brien, associate professor of public policy and urban affairs who, with his colleague Babak Heydari, associate professor of engineering, recently published a comprehensive study of Airbnb listings and crime rates in neighborhoods throughout Boston.

They found that it was the proportion of buildings with at least one home-sharing listing—and not the volume of tourists cycling through such units—that had the greatest (indeed, only) measurable effect on crime in the neighborhood. Their research was published Wednesday in PLOS One, a peer-reviewed scientific journal published by the Public Library of Science.

“What seems to be the problem is that Airbnb is taking households off the social network of the neighborhood and eroding its natural capacity to manage crime,” says O’Brien, who also studies criminology and criminal justice at Northeastern.

The researchers hope their study can help guide local and regional policy decisions about short-term rental regulations.

O’Brien and Heydari compiled 911-call data and Airbnb listings and reviews from 2011 to 2018, a period of rising concern about crime and during which listings on the online home-sharing platform more than doubled in Boston. They found that certain violent crimes, including fights, robberies, and reports of someone wielding a knife, tended to increase in a neighborhood a year or more after the number of Airbnbs increased.

The lag, Heydari says, is evidence that it’s not the immediate presence of rowdy tourists or criminals taking advantage of newcomers that’s driving an increase in crime. Such effects would be seen in the same year that listings increased, not afterward.

“What we’re seeing is evidence of a slower process, one that becomes significant over the years,” he says. “It’s another support that changing the social fabric of the neighborhood is what’s undergirding these results.”

In reply to a request for comment, a spokesperson for Airbnb referred a News@Northeastern reporter to a written response posted to the company’s website. The missive takes issue with the study’s methodology and conclusions and questions whether the researchers controlled for other factors such as new housing construction and other economic conditions.

To measure the presence of Airbnb in the city, O’Brien and Heydari used a combination of geographic available data scraped from website listings and the “host since” feature on the company’s website, which logs the year a host joins the site. They organized the data by longitude and latitude and matched it with census tracts to approximate neighborhoods in the city.  

This gave the researchers a sense of the number of Airbnb units per neighborhood, but not how they were distributed—after all, a dozen units contained to one condominium complex may have a different overall effect than a dozen units in a dozen different houses along a street.

In order to determine this distribution, which they describe as the penetration of Airbnb units in each neighborhood, the researchers divided the number of unique addresses with listings by the number of parcels in the census tract. The calculation resulted in an approximate measure of the proportion of buildings with at least one listing.

Heydari and O’Brien relied on data about when a user “joined” Airbnb because the platform does not make more specific data available, but the researchers acknowledge that their proxy isn’t always a perfect representation of when units cropped up in buildings.

“This serves to highlight the need for Airbnb—and other social networks and web platforms—to be more transparent in their data,” Heydari says.

The researchers also accounted for other, unrelated changes to the neighborhoods that might have had an effect on crime rates, including rental transience and sociodemographic changes, Heydari says.

And while their data are limited to Boston, the motivating theory that a community of neighbors who all know and look out for each other is a strong crime deterrent is one that can be translated to cities and communities across the country, the researchers say.

“What we’re showing here is that when we tease out the other factors, we see that the penetration of Airbnb has a role in a gradual increase in crime,” Heydari says. “That effect is not specific to Boston, although other cities and neighborhoods may contain factors that change the degree to which it presents.”

For media inquiries, please contact Jessica Hair at j.hair@northeastern.edu or 617-373-5718.