Crash causes, countermeasures, and safety policy implications.

There are interrelationships between crash causes, countermeasures, and policy implications, but they are not necessarily direct and obvious. Part of the problem is the definition of a cause. The seminal 1979 Indiana University "Study of Accident Causes" has cemented some false assumptions that must be overcome to yield an effective crash countermeasures policy. The taxonomy of crash causes and the prevalence of different causes are determined by the investigators, who are biased in different ways. The prevalent notion that approximately 90 percent of the crashes are due to human errors or failures is due to a threshold bias, and the implied notion that 90 percent of the countermeasures should be directed at changing these behaviors is based on an erroneous assumption that the cure must be directly linked to the stated cause. A more balanced approach to the definition of a cause and to the search for crash countermeasures is needed, and the safe system approach appears to be a most promising one.