Federal agencies collect data on fatalities caused by “distracted driving,” which can include anything from talking on a cellphone to eating in the car. Although there have been laboratory and observational studies that have estimated the impact of cellphone usage while driving, there has been no way to tease out exactly how many crashes are caused by texting – or cellphone use overall.
“It’s extremely hard to measure causation,” says David Teater, senior director of the National Safety Council’s transportation strategic initiatives. “I believe [that] in all states there’s a mandatory blood alcohol test after a fatality.… There is nothing even close to that with distracted driving. There’s no way of knowing for sure.”
In the past, if the police did try to investigate whether cellphone usage was involved in a crash – not a regular occurrence, since for years talking on the phone while driving was not illegal – officers would typically have to rely on the driver’s word.
Fernando Wilson, who co-authored the study with fellow University of North Texas Health Science Center assistant professor Jim Stimpson, agrees with Mr. Teater that the study’s fatality predictions are probably low.
“There is going to be a lot of underreporting,” Mr. Wilson says. “In reality it’s probably an even worse problem than what the data is showing.”
Wilson and Stimpson combined data from the US Fatality Accident Reporting System database with cellphone usage statistics, adjusting for climate, state demographics, and other factors. Using this statistical modeling, they determined that if texting had not existed, the number of distracted driving fatalities would have actually declined from 4,611 to 1,925 per year from 2001 to 2007. As it is, distracted driving deaths increased to 5,870 in 2008, according to the US Department of Transportation.