As the debate over regulation of Uber continues across the country, Florida House Majority Leader Dana Young, R-Tampa, took to Twitter with a factoid about the ridesharing company’s effect on DUIs in her home city.
"DUI arrests dropped significantly in Tampa once Uber began operating here," Young tweeted June 21. "The Austin situation is disturbing."
Young was responding to a tweet by conservative focus-group leader Frank Luntz, who said the number of DUI arrests has increased in Austin, Texas, since Uber and Lyft ceased operations in May. Luntz’s tweet linked to an article that discussed the aftermath of Austin abandoning the ridesharing companies.
After talking to experts and analyzing the data, we found the relationship between DUIs and Uber isn’t as clear-cut as Young makes it seem.
The road to the truth
This isn’t the first time Young, who is running for the Florida Senate, has fought in favor of Uber, which arrived in Hillsborough County in April 2014. Lyft arrived the same month.
In 2015, the Hillsborough County legislative delegation advanced a bill (HB 509) that would regulate ride-sharing companies. Before the bill passed, Young proposed an amendment with stipulations for background checks but didn’t include fingerprinting, which is required for Hillsborough County taxicab and limousine drivers.
With Young’s interest in mind, we turned to the evidence.
Young’s communication team sent data to support the claim, but it only included two six-month periods from January-June in 2014 and 2015. Although the data from 2014 and 2015 shows a 30 percent decrease in DUI arrests, it cherrypicks the timeframe.
Kelsey Swithers, a spokeswoman for Young, said the data was the most recent information that Tampa police had as Young prepared a local bill.
We requested more recent, comprehensive data from the police department. It shows the number of DUI arrests has fluctuated since 2012.
Before we get into the numbers, here’s a few things to know: Other law enforcement organizations like the Hillsborough County Sheriff Department also make DUI arrests in Tampa, which aren’t reflected in these numbers. However, arrests by Tampa police far outnumber those from other agencies in Tampa. The data includes DUI arrests from different categories, whether it is a simple DUI stop or a serious car accident where a person was driving under the influence.
So let’s look at what the numbers show.
Graphic by Lauren Flannery
It’s hard to detect a general trend in the data. For instance, the total number of DUIs between 2012 and 2013 decreased by 135 arrests, but the total number of total DUI arrests increased by 64 between 2013 and 2014.
There was a general decline in DUI arrests after Uber came to town, with arrests staggering between 103 and 136 for most of 2015 — compared to a June 2014 high of 185. But at the beginning of 2016, arrests for DUI increased again to pre-Uber levels.
Tampa police spokesman Stephen Hegarty cautioned that a number of factors could explain the volatility in the numbers, including announcements of DUI checkpoints and heightened enforcement around holidays.
Still, Bill Bales, a professor in the College of Criminology and Criminal Justice at Florida State University, said there may be something to Young’s theory based on the preliminary evidence for 2015, even if it doesn’t control for other factors.
There were 414 fewer DUI arrests in 2015 than in 2014. It probably took around nine months before the number of Uber drivers was meaningful, Bales said, so it’s fair to use DUI stats from 2013 to 2014 as pre-Uber DUI trends.
"The fact that DUI arrests declined by 22.2 percent from 2014 to 2015, which is arguably the first full year of operation, suggests Uber has had a significant impact on DUI arrests," Bales said.
So case closed? Not exactly.
Correlation ≠ causation
Academic research rarely makes assertions like Young’s tweet, Bales said.
"The reality is we can’t control for everything," Bales said. "We rarely make the claim that X caused Y at this level. We’re not that precise."
Researchers behind a three-year study (not yet peer-reviewed) of Uber in 150 cities did conclude that Uber lowers the rate of fatal crashes and DUI arrests. The study, which did not include Tampa because it examined 2010-13 data, was just finished in May.
Sean Mulholland, co-author of the study and professor of economics at Stonehill College, said they were able to draw broad conclusions about Uber’s influence because of the large sample size.
"When looking at only one city or county, there might be other factors unique to the city that are driving the decline," Mulholland said. "But when we look at many cities, we are able to hold constant any unique factors that do not change over the years of the study."
In addition to all of the factors potentially influencing DUI stats, the culture of drinking and driving has changed over the last several decades, said Thomas Blomberg, executive director of the Center for Criminology and Public Policy Research at Florida State. The portrayal of DUI stops in media is just one way people have become less deterred from getting behind the wheel while under the influence.
Young said DUI arrests dropped dramatically when Uber came to Tampa.
Uber came to Tampa in April 2014, and DUIs did drop that month, and continued to do so in 2015.
However, experts cautioned that there are many more factors that can affect the number of DUI incidents, so attributing the drop to solely to Uber is unscientific. Plus, more comprehensive data show the number of DUI arrests has fluctuated over recent years.
We rate this claim Half True.