Traffic accident statistics on signs may actually cause more crashes

The number of crashes on Texan roads increased when electronic signs were used to display driving fatality figures

By Matthew Sparkes

Electronic signs above US highways that highlight annual traffic fatalities are intended to shock people into driving safely, but statisticians warn there is compelling evidence they actually cause accidents.

Although the messages are displayed in many US states, Jonathan Hall at the University of Toronto in Canada and his colleagues chose to examine crash data from Texas because the state uses electronic signs for fatality warnings just one week out of each month, providing ample control data for comparison.

The researchers compared the number of crashes from the two years before the signs were introduced with five years of data while they were in place. They discovered that displaying a fatality message increased the number of crashes on the 10 kilometres of road after the sign by 4.5 per cent. Their findings suggest fatality messages caused an additional 2600 crashes and 16 deaths per year across Texas.

Hall puts the effect down to distraction and the increase in a driver’s cognitive load while absorbing the information. “You see it and you’re thinking about it, and so you don’t put on your brakes quite as soon, and these little errors, 1 in 50 times, might cause a crash,” he speculates. “The perfect evidence would be a randomised control trial. I want to be clear that we don’t have that. But I think we actually have really, really compelling evidence.”

He says that the number of fatalities displayed on a sign also changes its impact as larger numbers are more shocking. In Texas, the state fatality numbers were reset each year in February and the team saw a big drop in crashes in February compared with January.

Hall says his team has written to all states that show such warning messages and asked them to collaborate on further research, but they have received little positive response. “If you’re going to do a safety campaign, it’s not that hard to say, ‘hey, let’s randomly draw five weeks from the next six months to show this sign, and analyse crashes’. But they haven’t done that, because there’s just a presumption that the signs can’t hurt,” he says.

The Texas Department of Transportation didn’t respond to a request for comment, but is understood to no longer display such warning messages above highways.

Journal reference: Science, DOI: 10.1126/science.abm3427