Uber along with  Mothers Against Drunk Driving released a study earlier this year taking credit for a decline in drunken driving-related car crashes among drivers under 30. It was called out for not producing enough evidence to make the connection.

But now comes a study from researchers at Philadelphia’s Temple University that suggests the entry of Uber’s low-cost ride service, Uber X, into 14 California counties led to a 3.6 to 5.6 percent decline in drunken driving deaths.

Brad Greenwood, an assistant professor at Temple University’s Fox School of Business, got the idea for the study after he’d had a little too much to drink at a friend’s wedding and decided to take Uber home. His research was conducted independent of Uber, although Greenwood says he’s discussed the findings with company officials.

Greenwood and his colleague Sunil Wattal, also an associate professor at Temple University, analyzed data from the California Highway Patrol’s Statewide Integrated Traffic Records System from 2009 to 2014.

That data told them the number of crashes that happened in each county along with “blood alcohol content of the driver (i.e. if alcohol was involved), the number of parties involved, weather, speed, and other environmental factors,” according to the study.

Rather than look at overall trends for drunken driving-related deaths, the researchers were able to observe statistics for the different times Uber entered different markets and compare them with each other.

“If there’s an overall drop off, we wouldn’t see an effect that’s just correlated with Uber,” Greenwood explains. “We’d just see the overall change in the trend, which would be completely uncorrelated with Uber’s entry.”

Greenwood says he was surprised to find that people seem less inclined to take Uber during a period of surge pricing, or generally if the cost is too high. From the study:

While results suggest that introducing Uber X into a city has a significant dampening effect on the number of alcohol related driving deaths, the introduction of Uber Black does not. All else equal this suggests several key pieces of information. First, it suggests that previous within city investigations of the effect of Uber entry may have been overstated. Second, it suggests that a coupling of cost and availability is the key driving force behind the decrease in DUI related deaths, indicating that patrons are unwilling to pay a price premium for the Uber Black service, even in the short term. Economically, these results suggest an average decrease in DUIs related homicides of 3.6% in locations treated by Uber X in the state of California.

But Greenwood says this is also a bit of a gray area, because Uber has refused to release data on surge pricing, so researchers can’t tell when it was in effect.

“I think the company is cognizant of the fact that [surge pricing] may have this negative effect on the rate of drunk driving,” he says.

Natasha Thomas, the program director for the Bay Area chapter of Mothers Against Drunk Driving, says she expects to see more studies that produce similar results.

“We definitely do believe that Uber, or any other ride-sharing company, has helped us in our fight against drunk driving,” she says. “They offer reliable, right-to-the-minute services to help people get home safe instead of getting behind their wheels intoxicated.”

Greenwood says a lot of the debate about ride services and the “sharing economy” is based on speculation. His hope is that policymakers take the findings of studies, like his, to try and quantify the potential benefits of companies like Uber.

Image credit:  Carolyn Coles | Flickr