There were only two drunken driving arrests last New Year’s Eve in San Francisco, the lowest since 2009, according to crime statistics from the San Francisco Police Department provided to the Ferenstein Wire.

The recent data comes on the heels of a new study revealing that the introduction of Uber’s low-cost service, UberX, reduces drunk driving deaths all over California.

Temple University’s Brad Greenwood and Sunil Wattal published a new paper  which finds that (not surprisingly) cheap taxi-like options make it easier for people to make the safer decision and call an Uber or Lyft rather than driving home themselves.

If the benefits of Uber in California were extended to the entire country, ride-sharing would save billions of dollars and hundreds of lives.

“Economically, results indicate that the entrance of Uber X results in a 3.6%-5.6% decrease in the rate of motor vehicle homicides per quarter in the state of California. With more than 13k deaths occurring nationally each year due to alcohol related car crashes at a cost of 37 billion dollars, results indicate that a complete implementation of Uber X would create a public welfare net of over 1.3 billion to American taxpayers and save roughly 500 lives annually,” wrote the team.

The overwhelming benefits of ride-sharing apps makes it harder for local politicians to keep Uber out of their state or city with regulations. Last month, New York City Mayor Bill de Blasio faced a backlash when he proposed to cap the number of ride-sharing cars that could be on the road in NYC. Eventually, he conceded, but only after Uber put forth a massive public campaign and employed costly political operatives.

This latest evidence makes the case that Uber isn’t just a battle with local taxis, but a battle against drunk drivers.

Previous attempts at estimating the number of drunk driving deaths that Uber had reduced were fraught with controversy. When Uber teamed up with Mothers Against Drunk Driving (MADD) for a study, Uber ended up claiming that the conclusions of the research were much more convincing than MADD was willing to admit. There was just too much noise in the data.

This latest study uses more sophisticated statistical analysis and leverages a common economic trick to estimate the effects of Uber. The decision to enter a market is often political: Uber will start hiring in a city if it knows it won’t face overbearing regulation. That decision is often unrelated to other reasons why there might be reductions in drunk-driving accidents, such as a “city’s population, bar scene, and tougher enforcement,” note the authors.

California is uniquely important in the study because the state has been part of UberX’s rollout since it’s beginnings in 2012, which means we can study what happens when the company can saturate a city with enough of its drivers to change consumer behavior. The state is also large enough that economists can estimate the differences in how Uber affects cities within the same state as it slowly enters adjacent markets.

Today, Uber and Lyft drivers are all over San Francisco. As a resident, I can say that using ride-sharing apps in the city has just become part of the culture.

Last New Year’s Eve in San Francisco, there were so many Uber and Lyft drivers on the road that users didn’t experience the dreaded “surge pricing”, which, in the past, could mean paying over $100 for a short ride home. There was little excuse for anyone to drive on NYE with a plethora of cheap alternatives.

The data on San Francisco drunk-driving arrests during New Year’s Eve makes it difficult to know whether Uber was the cause of the reduction in drunk driving, or whether the decline since 2011 would have happened anyway. This latest study makes the case that Uber was likely a significant, if not the primary, factor.

Evidently, Uber, Lyft, and other disruptive transportation services are saving lives.