Famous Statistic Blaming Car Crashes on Human Error Is Wrong

You have probably heard that almost all car crashes are due to human error. You may have even heard an exact, scientific-sounding number attached to it: 94 percent, according to the National Highway Traffic Safety Administration (NHTSA). And it’s also likely this number made sense to you, like it did to me when I first heard it, because most drivers are impatient and easily distractible. 

But persistent advocacy from safety experts has brought to light that this statistic is made up. And, in fact, there is no good evidence to support the claim that “human error” is the cause of most crashes. Their argument has been so convincing that the head of the National Transportation Safety Board, Jennifer Homendy, has officially called on NHTSA to remove the 94 percent statistic from their website, deeming it reflective of a “culture that accepts” 40,000 people dying in fatal car crashes a year. 

In the alphabet soup of federal transportation agencies, NHTSA is an agency within the Department of Transportation which oversees and enforces vehicle safety, whereas NTSB is a standalone investigative unit that conducts forensic investigations into fatal crashes—mostly planes, trains, boats and helicopters—and issues recommendations to the agencies like NHTSA. 

The 94 percent number comes from a 2015 NHTSA memo that surveyed crash reports from 2005 to 2007. It found that in 94 percent of cases a human action was "the last failure in the causal chain of events leading up to the crash" but that it is “not intended to be interpreted as the cause of the crash.” This makes sense, as it is simply the way car crashes work. Humans are driving, so the last action before a crash is probably the human doing something. But that means absolutely nothing for what actually caused the crash. To make such a determination, it would require a thorough investigation by qualified inspectors like NTSB or from a local police department. Very few crashes undergo such thorough investigations, making it easy and convenient to blame “human error” for everything, when other factors like road design, excessive speed limits, or weather conditions may have been more important.

Nevertheless, the 94 percent statistic has been misrepresented in all kinds of dangerous ways. On NHTSA’s own website, the stat appears on a page about the potential benefits of automated vehicles: “Automated vehicles’ potential to save lives and reduce injuries is rooted in one critical and tragic fact: 94% of serious crashes are due to human error.” It is, of course, not a fact at all, but nevertheless widely deployed by AV boosters as the main argument in favor of AVs and a main driving force behind AV hype.

Homendy has been on this anti-94-percent train, so to speak, for a while now. Back in October, she tweeted “Stop with the 94%!” before linking to a Streetsblog article that debunked the statistic. 

More recently, on January 14, Homendy posted a Twitter thread explaining why the misleading statistic matters. “This leads the public to believe there's nothing anyone can do about it, so who cares? When, in fact, the public should be outraged about all the death/serious injuries on our roads & all the action that has not been taken to prevent it…It also provides cover for all those planners, designers, engineers, auto manufacturers, employers, federal, state, & local leaders, and so many others who share in the responsibility of ensuring that if one part of the road system fails, the others will still protect people.” 

Indeed, the US is one of the few developed countries where vehicle fatalities have gotten worse in the last decade, not better. While there are many reasons for this, from an obsession with larger vehicles to a dependence on unsafe road design, the complacency with traffic fatalities as an inevitable fact of life is certainly a contributor.

“If we just focused on drivers, no one else would be held accountable for taking action,” Homendy tweeted. “Everyone must take action to save lives.”