In 2020, the COVID-19 pandemic shifted traffic patterns and reduced traffic around the U.S. as schools moved to virtual learning, businesses considered nonessential closed, and many employees worked remotely. People were spending more time at home and making fewer trips in their vehicles.
Highway safety professionals assumed that crashes and fatalities would fall along with the traffic volumes, since traffic volumes are the primary factor in most crash prediction models — more cars on the road generally correlates with more crashes. Despite the decrease in traffic volumes over the course of 2020 and total miles driven dropping by 13%, many states experienced a spike in road fatalities.
According to the National Safety Council, more than 42,000 people lost their lives and another 4.8 million were seriously injured in motor vehicle crashes in 2020. This represents an 8% increase in road accident fatalities and a devastating 24% increase in fatality rate (the number of fatalities per 100 million vehicle miles traveled) compared to 2019. In all, 2020 saw the most traffic fatalities since 2006, losing years of progress in reducing the most serious crashes.
These numbers are alarming and raise questions about what contributed to such an increase in fatalities. Anecdotally, there have been reports of increased speeding — both in terms of the number of people speeding and the speeds they are driving — and increased substance abuse among drivers involved in accidents. It is not hard to imagine that drivers have been more distracted than usual this past year as well.
Even with the best designs and safety features in place, highway safety still relies heavily on the roadway users themselves. The crash prediction models in the Highway Safety Manual are derived from traffic volumes and the features of the roadway itself. The models can be calibrated to local conditions to account for typical driver behavior within a state or region, but they generally assume the driver behavior in that area is fixed. When driver behavior changes drastically, as it has during the pandemic, crash prediction methods become less reliable.
Higher speeds result in more severe crashes, so focusing on strategies to reduce speeds is key to decreasing highway fatalities. As traffic volumes begin to return to normal levels, the additional vehicles on the road will help regulate speed. But this may be a good time to consider design policies that promote designing for lower speeds.
Wider lanes, gentler curves, longer sight distances and wider clear zones can make a road feel more comfortable, but these elements also make it feel safer to drive at higher speeds. When the feel of the road does not match the context of the road or the posted speed, drivers get mixed signals about what is safe and appropriate. Without other traffic or constant enforcement encouraging adherence to the speed limit, the roadway geometrics and roadside conditions become the most obvious cues for safe driving speeds. Self-enforcing roads use geometry and traffic calming techniques to give drivers cues about the speed appropriate for the purpose and context of the roadway. Design policies should encourage design speeds to be chosen deliberately and to not substantially exceed the minimum design criteria.
As the data continue to come in from 2020 and beyond, researchers will work to pinpoint the reasons for the spike in fatal crashes during the pandemic. Their findings will help refine safety models to make them more adaptable to unforeseen changes in traffic patterns and driver behavior in the future. In the short term, we can go back to the basics for reducing fatalities by focusing on proven strategies for lowering speeds, reducing distraction and encouraging sober driving.
Technology can be used to inform best practices on designing and building safe transportation infrastructure.
