The arrival of ride-sharing services has resulted in an increase fatal traffic accidents in U.S. cities, a draft paper released Thursday shows.

{mosads}The research, produced in collaboration between the University of Chicago and Rice University, shows that the ride-share apps are associated with a two- to three-percent increase in the number of fatalities for both vehicle occupants and pedestrians.

In 2010, the total number of traffic fatalities reached its lowest since 1949 at 32,885, but since then those numbers have slowly begun to rise.

The yet-to-be-published research compares publicly available national traffic safety data to dates that Uber and Lyft started operating in each analyzed city. It then looks at the accident rate per vehicle mile traveled in the city.

The researchers noted that they did not compare costs with the “many benefits that accrue from the presence of ride-sharing in a city.”

“These include improved mobility for the disabled and for minorities, flexible job opportunities that are especially valuable to those otherwise at high risk of unemployment, and customer convenience and resulting consumer surplus,” they wrote.

They also did not consider the cost of non-fatal accidents, because “data is not readily available.”

Representatives from Lyft did not immediately respond to The Hill’s request for comment.

Uber said its data science team reviewed the report and “found it to be flawed.” 

“Uber has contributed to safety in many ways and we take our responsibility to help keep people safe seriously. That’s why we’ve rolled out features to reduce the risk of drowsy driving and notify drivers of their speed on the road and will continue to do more,” a spokesperson said.