Matt Hersh, a lecturer in the Department of Statistics and Data Sciences at The University of Texas at Austin, recently applied some statistical science to the contentious debate over the impact of ridesharing in Austin.
Ride-sharing services like Uber and Lyft had cited statistics suggesting that the companies have led to a decrease in the number of drunken-driving accidents and arrests, in the lead up to a vote on Saturday, May 7th related to ride-sharing regulations in Austin.
Hersh examined the data in question using statistical approaches that controlled for seasonal differences and other variances in the data and found that the claims didn't stand up. In fact, there was no significant effect on drunken driving. He told KXAN, "There is no relationship, no correlation between Uber and Lyft and DWI accidents or DWI arrests."
Several local news outlets cited Hersh's work:
- Austin American-Statesman – "Police revise drunken driving crash stats key to Prop 1 campaign"
- Austin Chronicle – "Public Notice: Learning to Share?"
- KXAN – "Groups face off before crucial Prop 1 vote on Saturday"
- Time Warner Cable – "Austin's Prop. 1 Opponents Question DWI Decline"
Comments