Larger numbers of Airbnb properties are associated with more robberies, burglaries, theft and violence in London, research suggests.
The study, led by Cambridge University, indicated that a 10% increase in active Airbnb rentals in London would correspond to an additional 1,000 robberies per year across the city, above 2018 levels when the study data ends.
Researchers used crime statistics from the Home Office and Greater London Authority.
They analysed these alongside data from AirDNA, a site that scrapes Airbnb to provide figures, trends and approximate geolocations for the short-term letting market.
It suggested a "positive association" between the number of properties listed as Airbnb rentals and police-reported robberies and violent crimes in London neighbourhoods between 2015 and 2018.
Urban sociologists say the rapid pace at which crime rises in conjunction with new rentals suggests that the link is related more to opportunities for crime, rather than loss of cohesion within communities - although both are likely contributing factors.
Dr Charles Lanfear from Cambridge's Institute of Criminology, co-author of the study, said: "We tested for the most plausible alternative explanations, from changes in police patrols to tourist hotspots and even football matches.
"Nothing changed the core finding that Airbnb rentals are related to higher crime rates in London neighbourhoods.
"While Airbnb offers benefits to tourists and hosts in terms of ease and financial reward, there may be social consequences to turning large swathes of city neighbourhoods into hotels with little regulation."
Founded in 2008, Airbnb is a giant of the digital economy, with more than five million property hosts on the platform in some 100,000 cities worldwide, the study observed.
London is one of the most popular Airbnb markets in the world, with an estimated 4.5 million guests staying in a London Airbnb during the period covered by the study.
Dr Lanfear and his University of Pennsylvania co-author Professor David Kirk mapped data from 13 calendar quarters - January 2015 to March 2018 - onto "Lower Layer Super Output Areas", or LSOAs.
These are designated areas of a few streets containing around 2,000 residents, used primarily for UK census purposes.
There are 4,835 LSOAs in London, and all were included in the study.
Crime statistics for six categories - robbery, burglary, theft, antisocial behaviour, any violence, and bodily harm - were mapped onto LSOAs populated with AirDNA data.
The researchers analysed all forms of Airbnb lets, but found the link between active Airbnbs and crime is primarily down to entire properties for rent, rather than spare or shared rooms.
The association between active Airbnb rentals and crime was most significant for robbery and burglary, followed by theft and any violence.
No link was found for antisocial behaviour and bodily harm.
On average across London, an additional Airbnb property was associated with a 2% increase in the robbery rate within an LSOA.
This association was 1% for thefts, 0.9% for burglaries, and 0.5% for violence.
Dr Lanfear said: "Crime seems to go up as soon as Airbnbs appear, and stays elevated for as long as they are active."
He continued: "A single Airbnb rental can create different types of criminal opportunity.
"An Airbnb rental can provide an easy potential victim such as a tourist unfamiliar with the area, or a property that is regularly vacant and so easier to burgle.
"A very temporary occupant may be more likely to cause criminal damage.
"Offenders may learn to return to areas with more Airbnbs to find unguarded targets.
"More dedicated Airbnb properties may mean fewer long-term residents with a personal stake in the area who are willing to report potential criminal activity."
The study noted that Airbnb has taken steps to prevent crime, including some background checks as well as requirements for extended bookings on occasions popular for one-night parties, such as New Year's Eve.
Airbnb said that measures such as its party ban were not introduced until 2020, after 2018 when the study data ended.
An Airbnb spokesperson said: "This report is misleading as it fails to account for a general rise in crime rates and seasonal travel peaks in London, which have a far greater impact on crime than the cited, decade old data.
"Listings rented for more than 90 nights a year on Airbnb account for just 0.17% of homes in London.
"Airbnb creates new income for families, businesses and communities, boosting London's economy by £1.5 billion last year alone while supporting over 16,800 jobs."
Bram Gallagher, director of economics and forecasting at AirDNA, said: "The findings show only a small correlation between short-term rentals (STRs) and some types of crime, similar to other tourist-related activities like restaurants or hotels in aggregate.
"Importantly, the report dispels concerns that STRs undermine community cohesion, finding no evidence to support this.
"Overall, the data reinforces that STRs function much like other traditional tourist accommodations rather than creating new types of risk, while bringing economic benefits of tourism to local areas."
The research is published in the journal Criminology.
