Empty U.S. roads should be safer — but they’re not

With fewer vehicles on the road, drivers may be putting pedal to the metal

Almost three months into the coronavirus crisis finds U.S. highways with just a fraction of their pre-pandemic traffic. It would be natural to assume that should make driving safer, at least until the country is fully opened. But if anything, the roads may have gotten more dangerous.

As NPR recently reported, the National Safety Council has found that fatalities per mile driven were up 14 per cent in March from a year earlier.

So what’s going on? Part of the problem could be that people are more stressed, especially when they go outside. Stress is distracting and might affect reflexes and judgment. Plus some people are probably driving in masks that reduce their peripheral vision.

Those might be understandable reasons. But it also looks as if drivers may be driving more recklessly. In particular, rushhour driving speeds are up. And speeding has increased across the country, including in very high-speed categories, such as those caught driving at 160 km/h or more.

Unfortunately, that’s probably in large part a result of our behavioural foibles.

With fewer cars on the roads, there’s less traffic congestion, which makes it physically possible to drive faster. And when people can drive faster, they do.

Plus with fewer cars and pedestrians out and about, drivers probably feel they have more room for error and are less likely to have to react to an unexpected swerve by another driver or a texting jaywalker. More people driving alone because of social-distancing guideline could also account for the rise in risky driving; some drivers exercise less care when they don’t have passengers (although on the other hand, passengers can themselves be distracting).

And that’s not all: When driving, we often match our behaviour to perceptual feedback from other cars. That means there’s a herd effect; when a few motorists speed others drive faster as well.

Taking all of this together, we get a behavioural cocktail that leads to less-than-ideal actions. But how should we break this feedback loop? Police could step up patrolling, raising the probability that dangerous drivers would get caught. That said, this would only alter behaviour if the change were clearly communicated to the public.

People are also more likely to take actions — such as safe driving — that contribute to the public good when others observe them. But with so few other people on the road, who’s going to see drivers’ behaviour?

One simple nudge that might help would be to put up more roadside signs that tell drivers their speeds. At minimum, this would make high speeds visible to drivers and at least to some other motorists.

We could also give people signals about how much faster they’re driving than the average, perhaps by sending out letters to drivers based on data from speed cameras.

In any event, people’s behaviour is unlikely to adjust without some sort of policy change. So the question shouldn’t be whether to intervene, but which route to take — and how fast.