Ride-sharing vs. drunken driving
Drunken driving deaths are down in the Pittsburgh area. Should we be thanking ride-sharing companies?
By Brian O’Neill, Pittsburgh Post-Gazette Columnist
April 7, 2019
Is Uber saving lives?
There was a heinous killing of a 21-year-old college student in Columbia, S.C., last weekend, after she’d evidently gotten into a car that she mistook for the Uber she had called. Such calculated evil has shaken some who use ride sharing, and the likely fallout is more precaution: ride-hailers double-checking the driver’s name and car model against those on their apps before opening the car door.
That’s only wise. But it’s human nature to change behavior because of relatively unlikely encounters, if the report is dramatic enough. Surveys suggest Americans feel pretty safe when they’re behind the wheel, but less safe in planes where they don’t have control, and all the data showing that planes crash much less often than cars won’t change those minds.
“The great mythological symbol of America — the lone-wolf individual, out there challenging the world and all comers — makes Americans devalue some kinds of risks and elevate others,” Kirk Johnson of The New York Times wrote in a July 2000 report on the things Americans choose to fear.
It’s no stretch to say that Uber, Lyft and other ride-sharing services are built on millions of individual decisions and thousands of lone wolves deciding they want to make some money with their cars. And data have begun to dribble in showing that maybe, just maybe, that’s why we’re seeing fewer drunken-driving deaths in Pittsburgh.
Statisticians always warn against making too much of small sample sizes, but recent trends are encouraging. ValuePenguin, a personal finance research website owned by LendingTree, was looking into auto insurance data in Pennsylvania when it came across numbers that seemed worth a stand-alone report: Drunken-driving fatalities dropped by 17 percent from the three-year period ending in 2014 to the period ending in 2017 — and they dropped 83 percent in Allegheny County.
Lyft and Uber arrived in Pittsburgh in 2014.
No change of this magnitude ever comes down to a single factor. It’s the nature of such statistics that one terrible crash can skew the numbers. It could be this good news is a blip on the screen. It also should be noted that urban areas always have had lower rates of traffic deaths than the countryside. (Speed limits are lower and mandatory stops are more frequent.) But the increasing likelihood of drinkers, particularly urban/suburban drinkers, leaving the driving to someone else after a night on the town can’t be ignored here.
According to ValuePenguin, the number of drunken-driving fatalities per capita in metropolitan Pittsburgh’s most rural county (Armstrong) is five times as high as its most urban county (Allegheny). And since Uber and Lyft arrived in Pittsburgh, the gap has widened.
Drunken-driving deaths are down 83 percent in Allegheny, 92 percent in Beaver, 32 percent in Washington, 25 percent in Westmoreland and 22 percent in Fayette. Meantime, Butler County stayed even while such fatalities rose 14 percent in Armstrong.
Again, the smaller a county’s population, the more likely one or two bad accidents skew the numbers. ValuePenguin looked at 60 of Pennsylvania’s 67 counties, tossing out the smallest ones, and within that group Beaver and Allegheny counties had the second and third lowest rates of drunken-driving fatalities, respectively.
An Uber spokesman said it operates across the state. He would not offer any comparisons of ride sharing among the seven counties in metro Pittsburgh, but said there are hundreds of thousands of Uber rides a week in Allegheny County, and the number of rides rose 50 percent between 2016 and the end of 2018.
In short, there is evidence that ride sharing is saving lives, but it’s all circumstantial. It could also be people are drinking less or finding bars within walking distance.
Whatever the reasons for the downturn in drunken-driving deaths, this data set came too early to gauge the impact of a stricter ignition interlock law that went into effect in Pennsylvania in August 2017. Under that law, people with a first-time DUI conviction and a blood-alcohol content of at least 0.10 are required to blow into a tube that measures the presence of alcohol before turning the ignition key. If the device detects too much booze, the engine won’t start.
That should prevent accidents, too. In a few years, if the number of drunken-driving deaths keeps falling, we won’t know if it’s attributable to ride sharing or ignition denial, but we’re not likely to care.