More than 90% of crashes are because of driver error – except that's not what the data says

The NHTSA (National Highway Traffic Safety Administration) paper that includes this statistic is often abused because it’s easy to remember: the “critical reason” a crash occurred was assigned to the driver in 94% of crashes. Read carelessly, that sounds like “drivers caused 94% of crashes.” Read properly, it says something materially different. NHTSA defines the “critical reason” as the immediate reason for the critical pre-crash event, often the last failure in the causal chain. It then states, explicitly, that this is not to be interpreted as the cause of the crash or as assigning fault to the driver, vehicle, or environment.

That should end the argument. The claim that the paper proves more than 90% of crashes are the driver's fault isn't a bold conclusion; it's a misquote. The report even repeats the warning when discussing the 94% driver figure: in none of those crashes was the assignment intended to blame the driver for causing the crash.

The misunderstanding is understandable, but it doesn't hold up once you read the definition. It happens because the survey title includes “Causation,” the table heading says “critical reason attributed to drivers,” and the driver categories include words such as recognition error, decision error, and performance error. Those phrases sound like blame. But technical language doesn't override the definitions it comes with. “Critical reason” is a coding variable in an exploratory crash investigation. “Fault” is a legal, ethical, and causal conclusion. They are not the same thing.

The distinction matters. A driver who fails to see a pedestrian may be the last actor in the chain. But why did they fail to see? Was the crossing poorly lit? Was the road designed for speeds too high for the land-use context? Was the vehicle’s A-pillar obscuring the view? Was the signal phasing hostile to pedestrians? Was the driver workload increased by roadside complexity? A final human action can sit at the end of a long system failure. Calling that “driver fault” erases the system.

Wider road-safety practice has moved in exactly the opposite direction from the blame-the-driver reading. The Safe System Approach, which the US has adopted, says people inevitably make mistakes, but the transport system can be designed and operated to accommodate those mistakes and prevent death or serious injury. It anticipates human error and makes roads forgiving, with responsibility shared among users, designers, operators, vehicle makers, enforcement, and post-crash care.

So yes, people may misinterpret the NHTSA article because the 94% number is simple and seductive. But the correct reading is plain. The paper reports where researchers assigned the immediate pre-crash “critical reason” in a defined sample. It does not prove that drivers are at fault in more than 90% of crashes. To claim otherwise overlooks what the paper explicitly says.

If you would like to know more about the Safe System Approach and how to make a positive impact on road safety, get in touch with our team or me.