*Free* Small Texas-based study shows highway traffic safety messages can slightly increase the number of traffic accidents

Roadside safety messages meant to prevent highway accidents – particularly those that highlight the number of road fatalities in the area – can have a small effect in the opposite direction, instead leading to an increased number of car crashes. The study’s authors suggest this small increase results from a cognitive overload drivers experience when faced with particularly salient or negative traffic safety messages, which may seize their attention and result in distracted driving. The authors note their data also shows behavioral interventions can help if they are not too salient and are delivered when individuals’ cognitive loads are low. Automobile crashes are a leading cause of death in the United States and abroad. To reduce traffic deaths and injuries, highway agencies worldwide have turned to the use of electronic dynamic message signs (DMSs), roadside message boards that display traffic safety messages directly to the driving public. Although these behavioral interventions, designed to “seize the attention” of drivers while driving, are widely considered highly effective, their effects on traffic safety have not been thoroughly evaluated. To better understand their impacts, Jonathan Hall and Joshua Madsen used data from a traffic safety campaign in Texas, which began displaying a message showing the total number of statewide traffic fatalities on their DMS network for one week each month starting in August 2012. Using data on 880 DMSs and all crashes occurring in Texas between January 2010 and December 2017, Hall and Madsen compared the number of crashes occurring on road sections downstream of DMSs across Texas during periods when fatality messages were being displayed and when they weren’t. The number of crashes slightly increased during weeks DMSs displayed safety messages featuring fatality totals. According to the authors’ analysis, these messages resulted in a 1.35% increase in traffic crashes up to 10 kilometers downstream of the DMSs. This effect did not persist during weeks the messages were not displayed. What’s more, the authors show that this effect is greater when the reported number of deaths is larger and when road segments are more complex. In the supplementary text, the authors address seven alternatives to their cognitive overload hypothesis, including the possibilities that treated weeks are inherently more dangerous. But, in each case, they provide evidence inconsistent with each of the alternative hypotheses. “The crash data presented by Hall and Madsen clearly demonstrate a safety effect of showing fatality numbers on DMSs,” write Gerald Ullman and Susan Chrysler in a related Perspective. “However, the mechanisms for this safety effect are not clearly elucidated by the data presented in the paper. Additional analyses regarding crash types and documented causal factors in the crash reports might yield additional results.”